Liquid center heat spreader

[Diggrr]

Here's an idea for any and all to pick apart....list pros and cons please.

I've been reading alot about heat pipes as a very efficient heat transport/spreader (specifically the spreader plate style) though one made for a cpu seems a rarity except in notebooks.

Would a liquid core heat spreader be more efficient than a solid copper plate? No phase change, no airspace, just completely liquid filled.

I've seen thermographic pictures of heatsinks in action, and it looks as though the center is doing all the work. If one could make a spreader less complicated than a heatpipe (i.e. makeable at home), and more efficient than a solid plate of copper, the standard hsf perched atop it would thus be more efficient as there is little draft from the center of a fan.

If so, which liquid?

Any thoughts?

[UserName]

This (http://www.thermacore.com/thermabase.htm) has a good read.

The only advantage a liquid would have is that is can convect, i.e. move. So the liquid can get hot and rise to be cooled. As it is doing so it can be replaced by cooler liquid.

Heat actually moves through a liquid much slower (10 to 100 times) than a solid.

A direct die water cooling setup would really be doing what you say

Note
You can make a heat pipe at home.

Take a copper pipe (at least a foot long) and solder a cap on the bottom. Solder a reducer to a ¼ inch pipe and a little piece of ¼ inch pipe. Now poor in about half a cup of water. Heat the bottom until the water is boiling. Remove from heat. Quickly crimp and solder the ¼ inch pipe at the top. And let it cool.

That is a heat pipe.

Now grab the bottom with your hand. If everything went well it will start to boil ant the top will get warm~hot to the touch.

[Diggrr]

Yeah, I've seen several ways to make homebuilt heat pipes. One guy I ran into even uses propane and butane to make heat pipes that cool components in his electric car. He said propane is r290 refrigerant and butane is r600...huh, didn't know that.

Anyway, I have a yy-mini cube case, and going verticle is a problem. As I don't have access to sintered metal tubing (read..poor) or a quality wicking material, I'd be restricted to verticle use with a home built heat pipe. My board puts the socket at the top, so I have a grand 1" clearance over that.

I was just wondering if there's a liquid available that can beat copper as a heat spreader. I already have the copper plate (liquid vessel) designed.

Thanks.

[Aesik]

I'm afraid the answer is no as to whether or not there is a liquid that can conduct heat as good as copper. There really isn't even anything close that would work or we'd be using that liquid in our water(lquid)-cooled systems ;)

[kevin_bouchard]

I'm afraid the answer is no as to whether or not there is a liquid that can conduct heat as good as copper. There really isn't even anything close that would work or we'd be using that liquid in our water(lquid)-cooled systems
what about mercury? what is its heat transfer abilities? I would think that its atleast better than water,but this is only a guess. Also availability is also another issue.

[sfa ok]

Originally posted by kevin_bouchard

what about mercury? what is its heat transfer abilities? I would think that its atleast better than water,but this is only a guess. Also availability is also another issue.

It's very toxic, and very dense. You can float bolts in the stuff.

http://forums.overclockers.ws/vb/showthread.php?s=&threadid=14286

[Aesik]

Mercury is about 8.3 W.m/K, or only about 1/45 that of copper.

[Pepsi]

I think Mercury would work; but it's a little toxic, and I think the stuff would expand a lot once you got a little heat on it ...like a thermometer???

[Diggrr]

Well, thanks for the replys guys.

I could get mercury free, the shop I'm layed off from gets broken mercury switches all the time, and last I looked, there was a gallon + of it. Get this, they pay someone to dispose of it.

But since it's only 1/45 th as good as copper, that's definately a no go.

Guess I'll continue cojitatin' over a different idea. Thanks again for the replys.

[rossd]

UserName posted a recipe for homemade heat pipe in this thread that sounds interesting, but I don't completely understand it.

I don't understand how you can get the pressure low enough to boil the water at body temperature, just by allowing the pipe to cool from 100C.

The pressure difference between steam at 100C and 37C would be only 17%, assuming an ideal gas. If you look at the phase diagram for water, you need a much bigger reduction in pressure to make water boil at 37C.

So, how does it work ?

[kevin_bouchard]

But since it's only 1/45 th as good as copper, that's definately a no go.
so would that be making it more conductive than water? I think it would be neat to see it as an alternative to water for watercooling but a leak would be deadly, to the pc and the user :(

[Aesik]

Don't even think about trying to put mercury through any kind of pumping system! I can think of more than a dozen reasons in as many seconds.

[Szech]

If you could somehow get the liquid to move around in the center, I think it would work well. I'm reminded of the stirrers I set up at work. Drop a little bead in the beaker, and magnets stir it around. I suppose you could set up something similar on a much smaller scale, but magnets + computers = no.

[Diggrr]

Oh no. I'm not. If I used it at all (and I'm not) it would be in encased completely within the sealed block.

I have access to mercury, but that in no way means I'd use it for cooling or even in my computer. I have no deathwish for me or my system.:D

[dreadlord79]

I work in a Data Center for Sprint PCS and asked on of the techs about using mercury for cooling. He said the IBM used to use it on their servers for cooling. The reason that they stopped was because it was to volatile when changing the heat exchangers on the servers if anything went wrong. The tech did tell me that they worked like a charm and there were very few problems with them. You would have to put so copper pipe, 1 in, 1 out to your block. Plus, you would have to make a cork-screw like design so that as it expands, the Mercury has some place to go. Since the system would be creating a lot of heat, it would flow all by itself since the hot would rise by itself pushing the cooler mercury back into your reservoir.

[UserName]

About the home made heat pipe.

I think if you do the calculations at about 400 degrees you’ll see how it works. We would bring it to a roiling boil with a propane torch and heat the whole pipe.