Copper Foam

dante6491

dante6491

dante6491

Soldato
Joined
30 Jun 2006
Posts
6,244
Location
Horsham
I thought this may be of some interest to the cooling enthusiasts here.

I was at Conservative party conference this week and they have a company called Versarien exhibiting. http://www.versarien.com/

Basically they have made a fairly cost effective way of creating copper foam. They had a couple of test machines running with their own waterblocks (which are hopefully coming to the market soon). The idea being the porous structure allows the thermal paste to better transfer heat because of greater surface area. Obviously if it is tested and works and priced alright it will be rather interesting.

I had a quick search and we were talking about this back in 2009. http://forums.overclockers.co.uk/showthread.php?t=18043719&highlight=copper+foam

Discuss.
 
dante6491 said:
The idea being the porous structure allows the thermal paste to better transfer heat because of greater surface area.
Click to expand...

I dont really get this tbh. A porous waterblock would have lower thermal mass as there is less metal. Additionally if there is corrosion it would eat through proous material a lot faster than a solid block.

The advantages I can see are:
- faster and possibly cheaper manufacturing
- lighter blocks so less sag

But im not completely sold on the idea tbh
 
panyan said:
I dont really get this tbh. A porous waterblock would have lower thermal mass as there is less metal. Additionally if there is corrosion it would eat through proous material a lot faster than a solid block.

The advantages I can see are:
- faster and possibly cheaper manufacturing
- lighter blocks so less sag

But im not completely sold on the idea tbh
Click to expand...

^
This


Copper is a much better conductor than all the thermal pastes available to us, a Flat block to go on a flat IHS seems to be the logical choice. I can see it as a more effective alternative to transfer the heat to the coolant though, inside the block. Instead of copper fins, a carefully made copper 'foam' with consistent size holes offers more surface area. Overall there will be higher restriction to the loop but with better heat transfer, and it isn't like our pumps these days cant take it (unless you creative WCers are still using mini aquarium pumps). Maybe this foam would be better used in between the copper fins of a radiator instead of super thin sheets of aluminium zig-zaging up and down.

Will be interesting to see some private modders have a go with this, they are usually the ones to get the ball rolling.

Any volunteers?
 
dante6491 said:
I can imagine it going IHS->TIM->Copper Foam->Solid copper fins.
Click to expand...

I can see where you are coming from, but you do kind of get a bottle neck with heat transfer by having more of lower conducting material (TIM in this case). To cover the extra surface area of the foam, you will need more TIM. As you know the reason why TIM works well is because though it has a much lower heat transfer coefficient, it will connect the block to the IHS where it normally wouldnt touch and therefor have air in between which transfers heat even worse! Ideally, you would want your block to touch the IHS directly and have a thin layer of TIM in parts where it cant, that is why we only put a very thin layer or a tiny blob.

If you go IHS>TIM>copperfoam>Copper block, You have much less copper directly touching the CPU and the rest of the copper further away with more grease in between.

As conducting materials go these days to better connect IHS to block, Cool Labs liquid metal is the stuff to go for. I personally prefer the peace of mind of a non electrically conductive time though.
 
Copper foam represents a new generation of performance for air coolers but I honestly don't remember what the verdict was for water cooling, interesting that a cost effective way of producing it has been discovered as cost was the only thing holding it back before :(
 
Thank you OP, that's well worth knowing.

You don't want a layer of copper foam between IHS and waterblock (or heatsink). Ideally you want solder at the interface, then one of the gallium alloys, then normal paste. Foamed copper is unlikely to get a better contact than any of the above.

Where you want copper foam is in the waterblock. Current best practice is a set of slots cut in the base which water is directed over. Instead, you can direct the water through the maze of holes within a layer of foam - using the enormous surface area within the foam to transfer heat.

Considering their manufacturing method, which I think is rather inventive, it'll be entirely possible to have a solid, not-foamed base for mechanical strength with a layer of foam for water to pass through attached. As in, like a single block of copper - there's no need for an interface between the solid copper and the foam.

Versarien-graphic-2.jpg


Thankfully it looks like they're doing just that. The copper-water transfer is probably the biggest temperature jump in modern water cooling, so one hopes they make a success of it.
 
You must log in or register to reply here.
Back
Top Bottom