SUMMARY: Metallic foam structures have interesting properties as a heatsink material.
I happened to come across a potentially new heatsink material – copper foam. This is derived from a very interesting fabrication process which turns various metals into what looks like metallic sponge. Under a microscope, you can see its sponge-like quality:
The porous nature of the material allows the relatively easy passage of air/fluids through it (how easy depends on the density of the pores, although the more dense, the higher the pressure drop). For a given size, metallic foams can be 5 to 10 times larger in surface area compared to a similar sized flat plate. The attraction for electronic cooling is its ability to achieve high heat-transfer within a compact footprint¹. For example, some studies have shown that compared to a pin fin array, foam is three times more efficient.
One of the issues with copper foam is its wicking nature – if you solder foam to a plate, you may wind up with substantially less surface area – it does act as a sponge and will wick solder very quickly, filling its pores with solder and thereby reducing its effectiveness. There are techniques which can minimize this effect, such as flame spraying, which must be used to attain maximum heat transfer.
The pictures below show an example of a dense copper foam:
The sponge-like structure is evident and a side view shows its continuous structure:
As an example of this technology for use in a heatsink, a layer of copper foam was bonded to a copper plate using thermal epoxy:
The copper plate has a number of holes to enhance airflow through the foam (lateral flow through a large cross section falls off quickly due to its dense nature):
The height of this assembly with a 10 mm fan is less than 1″ thick.
For this test, I used two different fans – one a 70 x 10 mm fan spinning at about 4200 rpm, and one with a faster (and noisier) 70 x 25 mm fan spinning at 5500 rpm. Both fans were mounted directly upon the copper foam – it was found that effectiveness falls off very quickly with distance. The copper foam heatsink was tested on an Acorp 4S845A motherboard with a modified P4 1500 to read CPU case temps.
|70 x 25 mm, 5514 rpm, 69 dBA|
|70 x 10 mm, 4160 rpm, 56 dBA|
¹Die Temp as measured by the P4 on-die diode, per MBM.
Interpreting C/W: For every watt (CPUw) that the CPU
consumes, the HSF will limit the CPU’s temperature rise to (C/W x CPUw)
plus the temperature at the HSF’s fan inlet. For example, at an ambient temp of 25 C, a C/W of 0.25 with a CPU radiating 50 watts means that CPU temp will increase 50 x 0.25 = 12.5 C over ambient temp, or 37.5 C. The lower the C/W, the better.
Considering the un-optimized nature of this heatsink, its performance is nonetheless revealing – it out-performs many 1U heatsinks utilizing more conventional construction methods, such as very dense copper fin arrays. A better foam-to-copper bond should yield increased performance. I should also point out that thicker foam loses its effectiveness quickly, as the pressure drops increase quite rapidly with thickness.
Copper foam can be an effective heatsink material but does require hands-on experience to attain maximum performance within given footprint constraints.
¹ Some examples HERE.