Anybody for some liquid metal cooling?

[Illicit Tweakin]

http://www.nanocoolers.com/products_cooling.php

This looks interesting; Of course it's in dire need for some mods.
I'm sure a few people here could help out in that department.

[Nubius]

lol, crazy, the things they are coming up with these days.

[ShaftedTwice]

Yeah, thats pretty crazy..

[saratoga]

They don't actually say how good a conductor it is, or how good the heat capacity is (metals tend to have terrible heat capacity).

I think the idea is not that its a high performance coolant (though it may be), but rather that its electrically very conductive, and so you can use a solidstate pump with it instead of one with a moving impellor.

[Illicit Tweakin]

http://nwc.serverpipeline.com/news/54200844

NanoCoolers Inc. has developed cooling technology targeted initially at servers, where power densities can exceed 100 watts/cm2, and at notebook PCs, to cool both the central processor and the graphics processor.


Besides fans, the computer industry has relied on heat pipes, which cycle water, for cooling. But heat pipes tend to dry out over time, said CEO Jim Moore. NanoCoolers uses a nonflammable form of "liquid metal," a mixture of gallium and indium that Moore called 65 times more efficient than water. Further, the NanoCoolers pump design consumes only 300 to 400 milliwatts, using Lorentz electromagnetic forces to move the liquid. "The properties of the liquid permit a lower-power pump," said Moore.

[Illicit Tweakin]

The pump design has no moving parts. The liquid metal is electricly charged inducing a magnetic field, like the windings on a motor. The metal flows through a magnetic field like an electric motor.

[Nubius]

Sooo liquid metal in tygon tubing with a new electromagnetic pump in the future? :D

[MaxxxRacer]

pretty interesting stuff...

[STEvil]

http://64.233.167.104/search?q=cache:CkrgYDHjA8oJ:www.nuc.berkeley.edu/thyd/ne161/rahmed/coolants.html+specific+heat+capacity+%22liquid+met al%22&hl=en

http://hyperphysics.phy-astr.gsu.edu/hbase/tables/thrcn.html - thermal conductivities of some common metals.

I had another link (its burried in a water cooling thread I think) that had the thermal conductivity of liquid metal in it.. but I didnt really think about it at the time... gonna dig through my posts and see if I can find it.

EDIT

Thread: http://www.xtremesystems.org/forums/showthread.php?t=54974&highlight=liquid+metal

Post (by antipop): http://www.xtremesystems.org/forums/showthread.php?p=741312#post741312

Unfortunately didnt have much of a clue what "liquid metal" meant at the time given that mercury is a liquid metal, and that many metals once headed become liquid.. :stick:

[kamaleon]

if magnetic fields are involved (liek a magnent) wouldnt that be 'dangerous' to hard drives?

[STEvil]

not particularly.

[shadowing]

I recall that if you want to actually risk damage to your hard drive by magnets, you would need magnets powerful enough to pulll the iron from your blood just to even erase it. So in a way, no.

[kamaleon]

ahhhh... i know better now ;)

[Nubius]

lol pull the iron from your blood.....Magneto can do it! :D

[Butcher_]

If magnetism were that dangerous to HDDs they'd self erase when they spun up. ;)
You don't really want to be sticking them to car yard electro-magnets, but anything in your house is not going to do much damage. :p

[slavik]

sounds like a new way to format ^^


If magnetism were that dangerous to HDDs they'd self erase when they spun up. ;)
You don't really want to be sticking them to car yard electro-magnets, but anything in your house is not going to do much damage. :p

[cavemanxs]

Erasing the HDD isn't what I'd worry about - it's the potential for inducing current in the rest of the circuitry that scares me :eek: But I'm sure the clever people have come up with a solution to this - are these things out there yet, and if so has anyone seen them in action - this could be Huge :slobber:

[aMp]

So how real is this? Every time a "revolutionary" new cooling tech comes along, I get suspicious. I remember reading a lot of gushing press about these guys (http://www.cooligy.com/) almost two years ago, all of it primed by some skillful PR. They promised a product in six months, but have since gone off the radar -- haven't updated their press page since 2003!

-aMp-

[saratoga]

Erasing the HDD isn't what I'd worry about - it's the potential for inducing current in the rest of the circuitry that scares me :eek: But I'm sure the clever people have come up with a solution to this - are these things out there yet, and if so has anyone seen them in action - this could be Huge :slobber:

Since its a DC pump, the magnetic field is nonvaring and cannot induce current (in a stantionary object anyway).

[Illicit Tweakin]

So how real is this? Every time a "revolutionary" new cooling tech comes along, I get suspicious. I remember reading a lot of gushing press about these guys (http://www.cooligy.com/) almost two years ago, all of it primed by some skillful PR. They promised a product in six months, but have since gone off the radar -- haven't updated their press page since 2003!

-aMp-
As far as transferring large amounts of heat, quietly in a small form factor, this is great. Servers and laptops will benefit the most.

[DBlue135]

There would be eddy currents as a constant field couldn't accelerate the coolent, but these currents would only be present near or in the pump.

If it's a DC pump I'm actually pretty sure that you'd have to induce a (localized) current in the fluid. An AC pump would do the job much better. Even better than would be a series of sequentially firing electromagnets.

[Turok]

So what you guys are saying is that this solution powns a WC solution?
And Im not so sure if a magnetic field is good for the surrounding hardware.
Wont that make things like CPU's more unstable?

[DBlue135]

Does a normal pump screw up your hardware? They have rather powerfull magnets in them to spin an impeller. This set up could probably get away with weaker magnets and still have acceptable flow/heat transfer.

[saratoga]

So what you guys are saying is that this solution powns a WC solution?


If powns means "sucks compared to" then, yes, thats what I'm saying.


And Im not so sure if a magnetic field is good for the surrounding hardware.
Wont that make things like CPU's more unstable?

Read the thread.


If it's a DC pump I'm actually pretty sure that you'd have to induce a (localized) current in the fluid.

Yeah thats pretty much what they say:



When a DC current is applied to the EM pump via the electrodes, the pump propels the liquid through the closed loop system, transporting the heat from the heat source exchanger to the ambient heat exchanger within the module. The force within the pump which produces the flow is known as the Lorentz force. The Lorentz force states that when current is applied to a conductor in the presence of a magnetic field, there is a mechanical force on the conductor in the direction of the right-hand rule (current x magnetic field). In our EM pump design, a magnetic field is generated within the fluid channel by an external magnet. Electrodes are positioned on each side of the fluid channel perpendicular to the magnetic field to provide the current that will flow through the liquid metal. The magnetic field in the pump is enhanced by the use of a yoke, which also decreases the magnetic fields external to the structure. As DC current flows through the liquid metal from one electrode to the other, the Lorentz Force generated on the liquid metal (the conductor) forces the liquid metal down the fluid channel. Since the liquid loop is a fully filled system, the fluid upstream flows into the channel and the process continually repeats itself, creating constant fluid flow. By controlling the DC current to the EM Pump, one can control the rate of fluid flow within the loop, making the cooling system scalable within its functional limits.

[DBlue135]

Originally Posted by The Link in the First post
Dude, relax. I was responding to your comment that there would be no induced current because it was a non inverting magnetic field. . .

Since its a DC pump, the magnetic field is nonvaring and cannot induce current (in a stantionary object anyway). No need to attack me. . .

[saratoga]

Dude, relax. I was responding to your comment that there would be no induced current because it was a non inverting magnetic field. . .


Eh? Don't be so defensive. I was actually intending that as a mild compliment since you figured it out just by thinking the problem through.


No need to attack me. . .

Given that I posted that in response to someone else, before you ever posted in this thread, and without actually being aware of your existance, I'm not sure why you think thats an attack on you :)

Or an attack on anyone actually since its just a simple statement of fact.