joes copper foam heatsink

just a few notes on the copper foam heatsink article found here:

http://www.overclockers.com/articles1274/

he mentions several times that cooling capacity diminishes quickly when the amount of foam increases, but fails to mention that its because of the relativly low pressure design of the fans used.

i would be intereted in a newer test with say 1" foam and a blower, which would have the power to really push some air through that much.

sorry if my first post in this section is a little lacking in thought/detail, i promise the article is a quick read. i thought it might be more interesting to post here than emailing joe about it.

As much as I like the concept the real issue is the source and not the method. MFGRs need to get the damn thermals under control. If they did we wouldnt need this sort of cooling.

Specifically certain Intel chips. There aren't many AMD's that run over 60C. Stock anyways. This would be really good for prescotts and AMD's dual cores. I wonder how well a Venice would overclock with this.

Sentential said:

As much as I like the concept the real issue is the source and not the method. MFGRs need to get the damn thermals under control. If they did we wouldnt need this sort of cooling.

Click to expand...

I totally agree, but hopefully this will be taken care of with the die shrink coming up.

Or maybe heatpipes will do the job of conducting the heat around.

The problem that I see it getting the heat into the foam. AFAIK not much of the foam is touching the hot part, and it would be hard to transfer the heat to it.

Maybe we can use a little bit of solder to get the foam to come in good contact with the base?

Veeeery interesting. From that little test he did there, it actually seemed to do a good job of cooling. Put some time into some good R&D and you may have a new king of heatsinks. Some of the top heatsinks are where they are because of technology. (ie- look at where heatpipes went...)

thats not what they mean, lee.

lets say theres 100% transfer from the base to the foam. how far do you think that heat is going to transfer through those tiny little strands? not far, a thicker pad wouldnt help because the heat would never go past the first quarter inch.

your first idea was exactly waht i was thinking, small heatpipes acting as veigns for the heat to go through, passing through the copper sponge. that would provide good cooling with very little airflow.
perfect for small basic terminals.

The main problem I see with the copper foam heat sink is how to clean the dust out of it. I think that eventually the dust particles will clog the foam up and keep the air from flowing through it very well. All of the heatsinks I have seen are able to be blown out with a can of air, I doubt this would be the case with the copper foam.

Stack a delta, a dust filter and a finger guard on top of that and you may be onto a winner!

I like the heatpipes idea...

A heatsink designed with foam fins and a solid base would probably yield very good cooling properties.

The thermal conductivity of the foam is only about 70W m/k so the extention/penetration of heat is only about 1/8 of an inch. The primary advantages are the elimination of the Boundary layer due to the small pore size and the large surface area.

Fin eff. is about 97% which makes it so effective even with small working areas. Does not work well with air blowing over it partially due to the conductivity and also roughness actually increases the boundary layer if formulated like a typical FHS. Flow has to be theough the foam.

It would work well as a boiler for Phase change or as a micro channel water block, although fouling due to corrosion would have to be dealt with.

I saw one abstract for a Darpa research project where somebody claimed 10X performance from using the hollow struts in the cu foam as miniture heat pipes

Nifty stuff - they use it for filtration, EMI shielding, fuel cells etc.

lee1026 said:

Maybe we can use a little bit of solder to get the foam to come in good contact with the base?

Click to expand...

Most solder is a VERY poor thermal conductor. Which is why many copper heatsinks
don't perform as well as you might expect. Thermal resistance between the base & fins.

Seven7Thirty30 said:

A heatsink designed with foam fins and a solid base would probably yield very good cooling properties.

Click to expand...

Only if the fins were VERY short.

Plus you have the problem of joining them in a way which wouldn't significantly effect thermal conductivity.

MILLTHERM said:

The thermal conductivity of the foam is only about 70W m/k so the extention/penetration of heat is only about 1/8 of an inch. The primary advantages are the elimination of the Boundary layer due to the small pore size and the large surface area.

Fin eff. is about 97% which makes it so effective even with small working areas. Does not work well with air blowing over it partially due to the conductivity and also roughness actually increases the boundary layer if formulated like a typical FHS. Flow has to be theough the foam.

It would work well as a boiler for Phase change or as a micro channel water block, although fouling due to corrosion would have to be dealt with.

I saw one abstract for a Darpa research project where somebody claimed 10X performance from using the hollow struts in the cu foam as miniture heat pipes

Nifty stuff - they use it for filtration, EMI shielding, fuel cells etc.

Click to expand...

Probably works great for those applications.
But for a large heatsink on a big heatsource? Let's think about it.
Instead of traveling directly from point A to point B as in a solid copper block
the heat must travel around any number of air pockets by way of extremely
small conduits of varying size. Might even hit a dead end.
The second image in that article shows what I mean.

Most solder is a VERY poor thermal conductor. Which is why many copper heatsinks
don't perform as well as you might expect. Thermal resistance between the base & fins.

Click to expand...

well. then can we use molten Copper as a bond?

lee1026 said:

well. then can we use molten Copper as a bond?

Click to expand...

Sure. If you can heat it to the molten state without also melting the objects to
be joined. Which could be a real trick if you intend to join two copper objects.
You could try forge welding but that can be problematic with a highly conductive
metal like copper. And not something that lends itself well to high production processes.

The above is why I use pure Silver for my heatsink projects. It melts at a slightly
lower temp than copper, 220f to be exact. That temp window is too narrow
for heatsink manufactuers so they use solder.

Wouldnt a good thermal epoxy be fine?

I wonder how much water pressure can flow through it before the foam it self breaks up?

I like the idea of using heatpipes to carry the heat from the cpu to the coper foam. And silver solder would probably make a big difference in heat transfer to the foam.

I doubt this will ever develop into anything though.

Vacuum Brazing, laser weld would work fine

Now what if you used this stuff inside a Water block?

attach some of this to a very thin copper plate.

with the plate being extremly thin, the foam will heat up quickly.

but with enough water flow you could probably get GREAT results.

Because of the fact that there would be sooo much surface area for the water, and more surface area = more ways for heat to reach the water faster.

Kinda the same idea as having alot of turbulance inside of a water block to force the water to tumble around more and touch more of the block to absorb more heat.

i'm sure it could be done and i would bet money that it would do fairly well if someone can get it to bond with a copper plate fairly well so there is good heat transfer between the foam and plate. and makre sure the water can flow though it well.

Like...have the block FULL of this foam stuff. And use side inlets. intead of top mounted ones. That way the water has to flow THROUGH the foam, and can't just go from one top inlet over and through the top of the foam to the Top outlet.

I'm not sure how well that would work...but i bet it would work at least FAIRLY well since the water is forced though the center of the foam and out of the center of the foam on the other side....

anyone think this idea could work??