Woah! Forget water or air....

[Darkmatterx]

I can't believe nobody thought of this sooner to be honest...

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

Depending on how much it costs, this may be my next cooling setup.

 

[shawnlizzle =]]

definitly looks promising, a way better heat conductor and still in its liquid form... however, this looks just like advanced heat pipes with liquid metals in them. this will definitly perform better than normal hsf fans, but definitly will not replace water/air systems.

<font color=red>gforce mx100/200 @ 230/440 =]</font color=red>

 

[phial]

<A HREF="http://www.nanocoolers.com/products_cooling.php" target="_new">http://www.nanocoolers.com/products_cooling.php</A>


HOW TO MAKE A CLICKABLE LINK

TYPE THE FOLLOWING WITHOUT THE STARS: [URL*] paste your link here [/URL*]

THANK YOU FOR YOUR TIME

-------
<b>It's a man's obligation to stick his boneration in a women's separation; this sort of penetration will increase the population of the younger generation.</b>

 

[Mephistopheles]

As for water vs. liquid metal cooling:

Physically, water has a strong advantage over metals: <b>water has tremendously high specific heat</b>- just over 10 times the specific heat of copper, for instance, and about 5 times the specific heat of aluminum. This means that dissipating 100W through water will heat it only very slightly: it will heat up 10 times less than an equal mass of copper.

However, unfortunately, while water does have 10 times the specific heat that copper has, it also has a several hundred-fold (~800-1000!) disadvantage at thermal conductivity. This basically means that while water will have a huge thermal inertia, it will also not do a good job at heat conduction. It can <i>store heat</i> very easily, but it's just not very fast at <i>transporting heat</i>, unless, of course, you move the damned water itself around like in watercooling systems and keep the <i>temperature difference</i> between a large area interface of water and radiator, for instance.

The use of a metallic liquid is therefore very justified to counter this effect. Sapphire has been reporting that they're using a gallium-alloy-based liquid coolant. If you check the thermal characteristics of gallium, it's pretty clear that it would make for a suitable liquid metal coolant. Gallium is one of four metals which are typically liquid at near room temperature - with the other three being mercury, caesium and rubidium. In addition to that, gallium has nearly the same heat capacity as copper, and it has 80~100 times the thermal conductivity of water! In a closed loop, a gallium coolant will therefore have <b>0.1x the specific heat of water,</b> while at the same time being able to <b>conduct heat 80~100 times better than water</b>.

These numbers seem to suggest that a gallium-alloy coolant would do much better than water. You might be inclined to think that it heats up ten times "faster", but that's not the whole picture: it would also be capable of coping with that heat and transferring it over to a heatsink much, much faster. A radiator for this sort of coolant would only have to have <b>80~100 times less area</b> to be as effective at heat dissipation as an equivalent water radiator.

(AFAIK... anyone please correct me if I'm wrong)

 

[Ned_Flanders]

<A HREF="http://www.nanocoolers.com/products_cooling.php" target="_new">for the lazy</A>

______________
Welcome to my Shed of Pleasure

 

[phial]

You got owned, newb. Check above.

-------
<b>It's a man's obligation to stick his boneration in a women's separation; this sort of penetration will increase the population of the younger generation.</b>

 

[Crashman]

Doesn't sound too significant. Nuclear reactors use special cooling methods to deal with high heat, so that reference is plugged.

Water and heatpipes are just methods to move the heat to a larger radiator. BOTH are nothing more than air coolers. If you had a heatsink with the same surface area and airflow as a 120mm radiator, it would perform as well as the 120mm radiator.

<font color=blue>Only a place as big as the internet could be home to a hero as big as Crashman!</font color=blue>
<font color=red>Only a place as big as the internet could be home to an ego as large as Crashman's!</font color=red>

 

[BWMerlin]

Like you said it takes more energy to heat water but it also take more to cool it down (I think this was referred to as energy activation levels in chemistry). Therefore yes water will cool it down, but it stores heat and then the problem becomes how do you get rid of the heat. With metals they heat up quicker but also cool down quicker. So using a liquid metal has its advantage in that it can move the heat around faster (metals are good conductors of energy) and get rid of it quicker as well. If you were to use the same size radiator that you use with your water cooling with a liquid meatal you will find that it should run significantly cooler and have better heat transfer.

 

[Mephistopheles]

Mercury (Hg) is unfortunately highly toxic, which doesn't make it a very good idea. On top of that, its thermal characteristics are very lackluster: it has a fifth of gallium's thermal conductivity and around a third of its specific heat.

But you could, for instance, use a <A HREF="http://www.absoluteastronomy.com/encyclopedia/G/Ga/Galinstan.htm" target="_new">galinstan</A> cooling solution - it has 25% better thermal conductivity than gallium and freezes only at -20C - so it's liquid in room temperature. Galinstan must be used with caution, because it oxidates easily, but it is <i>not toxic</i>, which makes it a very good candidate.

Who knows what they're using? I just read the article, and I didn't find any reference at all as to what the hell they're using as a coolant...

 

[Mephistopheles]

That's basically what I was thinking, agreed. I don't know about this being called energy activation levels, but then again, I'm a physicist, not a chemist. But I get the feeling that, even if it were called that, the name would not be appropriate - it's just thermal conductivity, there's nothing being activated... AFAIK...

 

[Mephistopheles]

Oh, and as a side note, I'm very impressed by the fact that liquid metal coolant <i>pumps</i> have no moving parts, just electrodes and magnets. This basically means highly improved durability and <i>silent operation!</i> Almost too good to be true. Add to that the fact that a liquid metal, whichever it is, will have about 100 times better thermal conductivity than water, which means that a radiator needs only to have 100 times less surface area to have the same dissipation abilities of a comparable water radiator, and you can start to feel the potential of liquid metal cooling!

Silent, compact, efficient. Very impressive.

Edit: Sapphire seems to know what we know too: their <A HREF="http://www.sapphiretech.com/vga/blizzard.asp" target="_new">product description</A> matches what we've been discussing:

Built upon a liquid metal technology that is 65X more thermally conductive than water and requires no moving parts

I'm very, very curious as to what this could do in the CPU cooling camp. Probably more expensive than water, but probably superior too! 3.0Ghz on A64 X2 anyone?

I wonder how much time until we start seeing commercially available CPU coolers using liquid metal?... <P ID="edit"><FONT SIZE=-1><EM>Edited by Mephistopheles on 05/22/05 02:31 PM.</EM></FONT></P>

 

[Mephistopheles]

I finally looked it up and confirmed that the liquid metal coolant is indeed <A HREF="http://www.nanocoolers.com/newsroom_detail.php?news_id=16" target="_new">predominatly gallium-based</A>:

NanoCoolers' initial product, which it hopes to begin selling early next year, is a cooling system for laptops and other computers that's based on a small electromagnetic pump circulating what the company bills as a nontoxic liquid metal.
With its patents still pending, NanoCoolers has declined to release details of the metal, except to say it's a mixture comprised predominately of gallium.