Overclock.net › Forums › Cooling › Cooling Experiments › Liquid Mercury Custom Heatsinks. (Don't roll your eyes, genuine could work Idea)
New Posts  All Forums:Forum Nav:

Liquid Mercury Custom Heatsinks. (Don't roll your eyes, genuine could work Idea)

post #1 of 25
Thread Starter 
Okay,
Before you Flame about how mercury is poisonous blah blah, I know, I am tired of people saying about using mercury instead of thermal paste etc... but then one day while reading a thread here I was thinking, "Mercury would never work ... unless...." and it stuck me.

How about this:

A custom Made Heatsink, a large number of Hollow Pure copper Pipes or "Stems" Rising vertically from the base, with the thinnest possible thickness without melting or distorting under heat. The base itself, Hollow too, also extremely thin, which should fit ontop of, and AROUND the CPU chips, Each of the pipes is connected with the hollow base, and is filled from the top with liquid mercury, and then sealed, with a Fan ontop of the pipes blowing air downwards within the pipes, giving extremely effective cooling.

Depending on the nature of mercury under heat, it would even be possible to have no metal between the Mercury and the chip, as you could use something similar to kneaded eraseraround the chips, and then sink the heatsink into it, forming a seal stopping mercury from escaping, and having it in direct contact with the chips.

Tell me if this is possible, As this would be the most effective method of cooling (barring Dry ice and that lot, as this wouldn't need supervision, just fit the heatsink and you are fine until you need to replace the CPU)

Probably just my mind thinking things up and missing an obvious flaw, But yeah
450

~Dan
Red Machine
(9 items)
 
  
CPUMotherboardGraphicsRAM
AMD Phenom II X6 1605T GA-970A-UD3 AMD Radeon HD 7800 Series GEIL  
RAMHard DriveHard DriveCooling
GEIL  Seagate Barracuda 250GB 7200 16 MB cache Seagate Barracuda 500GB 7200 16mb Cache Asus AXE Square Gold CPU Cooler 
OS
Windows 7 Ultimate x64 
  hide details  
Reply
Red Machine
(9 items)
 
  
CPUMotherboardGraphicsRAM
AMD Phenom II X6 1605T GA-970A-UD3 AMD Radeon HD 7800 Series GEIL  
RAMHard DriveHard DriveCooling
GEIL  Seagate Barracuda 250GB 7200 16 MB cache Seagate Barracuda 500GB 7200 16mb Cache Asus AXE Square Gold CPU Cooler 
OS
Windows 7 Ultimate x64 
  hide details  
Reply
post #2 of 25
The flaw is thermal expansion. Your theory is good, but unfortunately since both the mercury and any air inside the copper heatsink will want room to expand, it will leak out the sides, or in the case of a sealed environment, increase pressure. This would then require a perfect seal, and pressure testing, which also means increasing the thickness of the copper walls, and now we are back to square one. The usefulness of mercury is wasted when the copper becomes thicker because any benefit it had is negated by the loss in thermal efficiency due to the heat having to transfer between the two metals to finally reach the atmosphere.
    
CPUMotherboardGraphicsRAM
Core i7 920 D0 4.2ghz HT (1.3625v) Asus R3E 2xGTX 460 (non SLi, no overclock) 6x2gb G.skill @ 6-8-6-24-1T 
Hard DriveOptical DriveOSMonitor
WD-VR 300GBx1, 2xWD 1tb,2x60gb Agility Some crappy combo burner... Arch x64 3xDell U2410f rev A02 
KeyboardPowerCaseMouse
X-Armor U9BL TT Toughpower 1200w (NTB more efficient) Mountain Mods Pinnacle 24 CYO Roccat Kone (R.I.P. A4Tech x7) 
Mouse Pad
Steelpad Experience I-1 
  hide details  
Reply
    
CPUMotherboardGraphicsRAM
Core i7 920 D0 4.2ghz HT (1.3625v) Asus R3E 2xGTX 460 (non SLi, no overclock) 6x2gb G.skill @ 6-8-6-24-1T 
Hard DriveOptical DriveOSMonitor
WD-VR 300GBx1, 2xWD 1tb,2x60gb Agility Some crappy combo burner... Arch x64 3xDell U2410f rev A02 
KeyboardPowerCaseMouse
X-Armor U9BL TT Toughpower 1200w (NTB more efficient) Mountain Mods Pinnacle 24 CYO Roccat Kone (R.I.P. A4Tech x7) 
Mouse Pad
Steelpad Experience I-1 
  hide details  
Reply
post #3 of 25
Thread Starter 
Hmm,

Yeah I didn't Take that into account.... How about leaving the Tops of the Copper pipes open? or with a Filter on the end to let air out but prevent dust and other impurities from entering?as long as the pipes are not filled to the top, it would still have the same effect, if using a foam filter or similar is not really ideal, then you could use very thin perforated copper or even a foil, which could then be removed easily for filling/refilling and even Draining in case of problems...

hehe that was actually an improvement, espescially with the pierced foil or alluminum, as that would create no barrior between the mercury and the air at the ends of the pipe... as we know from convection currents this would help cool the mercury faster... Meaning better performance!

~Dan
Red Machine
(9 items)
 
  
CPUMotherboardGraphicsRAM
AMD Phenom II X6 1605T GA-970A-UD3 AMD Radeon HD 7800 Series GEIL  
RAMHard DriveHard DriveCooling
GEIL  Seagate Barracuda 250GB 7200 16 MB cache Seagate Barracuda 500GB 7200 16mb Cache Asus AXE Square Gold CPU Cooler 
OS
Windows 7 Ultimate x64 
  hide details  
Reply
Red Machine
(9 items)
 
  
CPUMotherboardGraphicsRAM
AMD Phenom II X6 1605T GA-970A-UD3 AMD Radeon HD 7800 Series GEIL  
RAMHard DriveHard DriveCooling
GEIL  Seagate Barracuda 250GB 7200 16 MB cache Seagate Barracuda 500GB 7200 16mb Cache Asus AXE Square Gold CPU Cooler 
OS
Windows 7 Ultimate x64 
  hide details  
Reply
post #4 of 25
I think you may be on to something there, the only problem being that you would have to ensure that there wasn't too much or too little mercury inside the heatsink "reservoir" and that the system was always kept horizontal so the heatsink wouldn't leak out the top. That, and mercury evaporates (it is a liquid) albeit slowly (~9000 hours for .5g at 68*F). And once again, mercury gas is heavily toxic (I believe much more so than its liquid counterpart).
Quote:
Originally Posted by wikipedia 
Mercury can be inhaled and absorbed through the skin and mucous membranes, so containers of mercury are securely sealed to avoid spills and evaporation. Heating of mercury, or of compounds of mercury that may decompose when heated, is always carried out with adequate ventilation in order to avoid exposure to mercury vapor. The most toxic forms of mercury are its organic compounds, such as dimethylmercury and methylmercury. However, inorganic compounds, such as cinnabar are also highly toxic by ingestion or inhalation.[76] Mercury can cause both chronic and acute poisoning.

Also, I believe mercury amalgamates with copper, which means the two substances will combine over time...
    
CPUMotherboardGraphicsRAM
Core i7 920 D0 4.2ghz HT (1.3625v) Asus R3E 2xGTX 460 (non SLi, no overclock) 6x2gb G.skill @ 6-8-6-24-1T 
Hard DriveOptical DriveOSMonitor
WD-VR 300GBx1, 2xWD 1tb,2x60gb Agility Some crappy combo burner... Arch x64 3xDell U2410f rev A02 
KeyboardPowerCaseMouse
X-Armor U9BL TT Toughpower 1200w (NTB more efficient) Mountain Mods Pinnacle 24 CYO Roccat Kone (R.I.P. A4Tech x7) 
Mouse Pad
Steelpad Experience I-1 
  hide details  
Reply
    
CPUMotherboardGraphicsRAM
Core i7 920 D0 4.2ghz HT (1.3625v) Asus R3E 2xGTX 460 (non SLi, no overclock) 6x2gb G.skill @ 6-8-6-24-1T 
Hard DriveOptical DriveOSMonitor
WD-VR 300GBx1, 2xWD 1tb,2x60gb Agility Some crappy combo burner... Arch x64 3xDell U2410f rev A02 
KeyboardPowerCaseMouse
X-Armor U9BL TT Toughpower 1200w (NTB more efficient) Mountain Mods Pinnacle 24 CYO Roccat Kone (R.I.P. A4Tech x7) 
Mouse Pad
Steelpad Experience I-1 
  hide details  
Reply
post #5 of 25
Thread Starter 
Right,

Okay so Lets address these issues,
Firstly Mercury Gas is Toxic and Poisonous if inhaled:

Fans, The dispersion of the Mercury gas through even the one fan (which is mounted where the Gas would logically evaporate) would mean that inhilation is not really a factor that should be worried about, let alone after subsequent Dispersion through case fans.

As for the Evaporation, yes this problem is partially Solved by the Perforated foil tops, you could refill this when needed, and as you say the rate of evaporation is so low, that you would not need to check it frequently... almost rarely ^^

The only real problem that we see here is the amalgamation, but that could be solved, either use a different metal (silver for better Performance, at a high cost, or an aluminium alloy for similar performance)

~Dan
Red Machine
(9 items)
 
  
CPUMotherboardGraphicsRAM
AMD Phenom II X6 1605T GA-970A-UD3 AMD Radeon HD 7800 Series GEIL  
RAMHard DriveHard DriveCooling
GEIL  Seagate Barracuda 250GB 7200 16 MB cache Seagate Barracuda 500GB 7200 16mb Cache Asus AXE Square Gold CPU Cooler 
OS
Windows 7 Ultimate x64 
  hide details  
Reply
Red Machine
(9 items)
 
  
CPUMotherboardGraphicsRAM
AMD Phenom II X6 1605T GA-970A-UD3 AMD Radeon HD 7800 Series GEIL  
RAMHard DriveHard DriveCooling
GEIL  Seagate Barracuda 250GB 7200 16 MB cache Seagate Barracuda 500GB 7200 16mb Cache Asus AXE Square Gold CPU Cooler 
OS
Windows 7 Ultimate x64 
  hide details  
Reply
post #6 of 25
Mercury amalgamates with most metals:
"An amalgam is a substance formed by the reaction of mercury with another metal. Almost all metals can form amalgams with mercury, notable exceptions being iron and platinum. Silver-mercury amalgams are important in dentistry, and gold-mercury amalgam is used in the extraction of gold from ore."

See the wikipedia pages:
http://en.wikipedia.org/wiki/Amalgam_(chemistry)
http://en.wikipedia.org/wiki/Mercury_(element)

That leaves platinum as the prime (and very expensive) candidate. Not to mention the inherent cost of mercury, the hoops you'd have to jump through to obtain the raw material in the first place (or the breaking of many thermostat bulbs.)

Possible? Probably.
Practical? Probably not.

Also, the rate of evaporation was taken at 68*F (20C), presuming you are attaching this to a processor die, you can assume the evaporation rate to increase exponentially. Especially under load - since most processors idle (given a moderately low ambient temperature) around 25-30*C (77-86*F). Furthermore, as the volume of the mercury decreases, the rate of evaporation increases. The ~9k hour estimation is based on a 1cm droplet of mercury amassing .5g, over time that volume would decrease, and the rate of evaporation would actually increase, in reality it would be safe to assume a .5g droplet of mercury at 68*F, to be between 6000 and 7000 hours. Granted we would be working with a much greater quantity of mercury in a much larger volume; however also with a much higher temperature, which has the most direct effect on the rate of evaporation.
    
CPUMotherboardGraphicsRAM
Core i7 920 D0 4.2ghz HT (1.3625v) Asus R3E 2xGTX 460 (non SLi, no overclock) 6x2gb G.skill @ 6-8-6-24-1T 
Hard DriveOptical DriveOSMonitor
WD-VR 300GBx1, 2xWD 1tb,2x60gb Agility Some crappy combo burner... Arch x64 3xDell U2410f rev A02 
KeyboardPowerCaseMouse
X-Armor U9BL TT Toughpower 1200w (NTB more efficient) Mountain Mods Pinnacle 24 CYO Roccat Kone (R.I.P. A4Tech x7) 
Mouse Pad
Steelpad Experience I-1 
  hide details  
Reply
    
CPUMotherboardGraphicsRAM
Core i7 920 D0 4.2ghz HT (1.3625v) Asus R3E 2xGTX 460 (non SLi, no overclock) 6x2gb G.skill @ 6-8-6-24-1T 
Hard DriveOptical DriveOSMonitor
WD-VR 300GBx1, 2xWD 1tb,2x60gb Agility Some crappy combo burner... Arch x64 3xDell U2410f rev A02 
KeyboardPowerCaseMouse
X-Armor U9BL TT Toughpower 1200w (NTB more efficient) Mountain Mods Pinnacle 24 CYO Roccat Kone (R.I.P. A4Tech x7) 
Mouse Pad
Steelpad Experience I-1 
  hide details  
Reply
post #7 of 25
There is a liquid metal cpu cooler back in '08, I've read the article somewhere, might wanna check that, it isn't mercury though, since it wouldn't be environmental for sure.
But I always thought about using mercury though, they would make an awesome coolant in a non phase changing situation smile.gif
Erin
(15 items)
 
  
CPUMotherboardGraphicsRAM
i7-2600k @ 5 Ghz Gigabyte Z68X-UD3P-B3 Sapphire HD 6950 2GB 8GB Kingston HyperX 
Hard DriveHard DriveCoolingOS
Seagate Barracuda 500GB Crucial M4 64 GB Noctua NH-D14 Win 7 Ultimate 64bit 
MonitorKeyboardPowerCase
Acer 23-inch Steelseries 6Gv2 Corsair TXv2 750W Cooler Master 690 II 
MouseMouse Pad
Razer DeathAdder Razer Goliathus 
  hide details  
Reply
Erin
(15 items)
 
  
CPUMotherboardGraphicsRAM
i7-2600k @ 5 Ghz Gigabyte Z68X-UD3P-B3 Sapphire HD 6950 2GB 8GB Kingston HyperX 
Hard DriveHard DriveCoolingOS
Seagate Barracuda 500GB Crucial M4 64 GB Noctua NH-D14 Win 7 Ultimate 64bit 
MonitorKeyboardPowerCase
Acer 23-inch Steelseries 6Gv2 Corsair TXv2 750W Cooler Master 690 II 
MouseMouse Pad
Razer DeathAdder Razer Goliathus 
  hide details  
Reply
post #8 of 25
I suppose the idea is more like a mercury-filled thermometer.

To make this more interesting, perhaps one of those vertical pipes can be constructed of thin-walled glass. On the glass surface, markings can be engraved to indicate temperature (just like a thermometer)biggrin.gif

450

PS The liquid metal cooler referred above is Danamics' Superleggera:-
http://www.guru3d.com/article/danamics-lmx-superleggera-review/
Edited by windfire - 11/13/11 at 12:44am
post #9 of 25
The biggest issue is safety.

Mercury isn't something that you just play around with. As was stated above, mercury does evaporate, and this is the real health hazard. It doesn't take much mercury in the air to cause health issues. It does accumulate in the body, and I don't think that there are many members here who would be willing to die to knock a couple of degrees off of their load CPU temps. wink.gif

There is also no way that you could sell this to anyone else. thumb.gif

Then the cost issues as well. Mercury isn't cheap, cleaning it up after a spill isn't either. The cost / benefit really isn't there.
Edited by AtomicFrost - 11/13/11 at 12:53am
YetiKube
(19 items)
 
  
CPUMotherboardGraphicsRAM
(1) 2500K ASRock Extreme3 Gen3 1 MSI 460 Hawk 1GB * Used to be SLI, RIP Hawk ... Mushkin 996995 DDR3 1600 (2 Sticks) 
Hard DriveHard DriveHard DriveOptical Drive
(1) OCZ Solid 3 SSD (1) 150GB Western Digital Raptor (1) 1TB Western Digital Black Caviar (1) Samsung SATA DVD Burner 
CoolingOSMonitorKeyboard
Antec Kuhler H2O 620 Windows 7 Professional Samsung 24" 1920 x 1200 Filco 114 key (Blue Cherry) 
PowerCaseMouseMouse Pad
Corsair HX850W - 850W Mountain Mods - U2 UFO Orginal - Gloss Black Logitech G400 / Razer Deathadder BE Razer Goliathus Extended / Artisan Hien VE (Sof... 
AudioOtherOther
Creative Sounds Blaster X-FI Titanium HD Asus USB N-53 Wireless USB adapter  12 Case Fans 
  hide details  
Reply
YetiKube
(19 items)
 
  
CPUMotherboardGraphicsRAM
(1) 2500K ASRock Extreme3 Gen3 1 MSI 460 Hawk 1GB * Used to be SLI, RIP Hawk ... Mushkin 996995 DDR3 1600 (2 Sticks) 
Hard DriveHard DriveHard DriveOptical Drive
(1) OCZ Solid 3 SSD (1) 150GB Western Digital Raptor (1) 1TB Western Digital Black Caviar (1) Samsung SATA DVD Burner 
CoolingOSMonitorKeyboard
Antec Kuhler H2O 620 Windows 7 Professional Samsung 24" 1920 x 1200 Filco 114 key (Blue Cherry) 
PowerCaseMouseMouse Pad
Corsair HX850W - 850W Mountain Mods - U2 UFO Orginal - Gloss Black Logitech G400 / Razer Deathadder BE Razer Goliathus Extended / Artisan Hien VE (Sof... 
AudioOtherOther
Creative Sounds Blaster X-FI Titanium HD Asus USB N-53 Wireless USB adapter  12 Case Fans 
  hide details  
Reply
post #10 of 25
Quote:
Originally Posted by AtomicFrost View Post

The biggest issue is safety.
Mercury isn't something that you just play around with. As was stated above, mercury does evaporate, and this is the real health hazard. It doesn't take much mercury in the air to cause health issues. It does accumulate in the body, and I don't think that there are many members here who would be willing to die to knock a couple of degrees off of their load CPU temps. wink.gif
Then the cost issues as well. Mercury isn't cheap, cleaning it up after a spill isn't either. The cost / benefit really isn't there.
This.
Not just anywhere in the body, but more specifically, it likes to react with certain chemicals found heavily in the body and turn into something that hangs out in your brain, lungs and kidneys. And, mercury poisoning (at least anything beyond a minor case) is not treatable, and not much research is being done to find treatment, since it is (for the most part) a completely avoidable ailment.

However, for the sake of experimental discussion, it is probably possible. Just really not.... worth it.
    
CPUMotherboardGraphicsRAM
Core i7 920 D0 4.2ghz HT (1.3625v) Asus R3E 2xGTX 460 (non SLi, no overclock) 6x2gb G.skill @ 6-8-6-24-1T 
Hard DriveOptical DriveOSMonitor
WD-VR 300GBx1, 2xWD 1tb,2x60gb Agility Some crappy combo burner... Arch x64 3xDell U2410f rev A02 
KeyboardPowerCaseMouse
X-Armor U9BL TT Toughpower 1200w (NTB more efficient) Mountain Mods Pinnacle 24 CYO Roccat Kone (R.I.P. A4Tech x7) 
Mouse Pad
Steelpad Experience I-1 
  hide details  
Reply
    
CPUMotherboardGraphicsRAM
Core i7 920 D0 4.2ghz HT (1.3625v) Asus R3E 2xGTX 460 (non SLi, no overclock) 6x2gb G.skill @ 6-8-6-24-1T 
Hard DriveOptical DriveOSMonitor
WD-VR 300GBx1, 2xWD 1tb,2x60gb Agility Some crappy combo burner... Arch x64 3xDell U2410f rev A02 
KeyboardPowerCaseMouse
X-Armor U9BL TT Toughpower 1200w (NTB more efficient) Mountain Mods Pinnacle 24 CYO Roccat Kone (R.I.P. A4Tech x7) 
Mouse Pad
Steelpad Experience I-1 
  hide details  
Reply
New Posts  All Forums:Forum Nav:
  Return Home
  Back to Forum: Cooling Experiments
Overclock.net › Forums › Cooling › Cooling Experiments › Liquid Mercury Custom Heatsinks. (Don't roll your eyes, genuine could work Idea)