Hey all -
I was just thinking that a open-cell metal foam might be a great way to achieve efficient cooling of IC's. Does anyone around here know if someone is looking into this?
For those that don't know, metal foams are basically like sponges made out of metal (typically Aluminum right now). The advantages of this type of structure is a HUGE surface (read - radiation) area, low weight, and fairly cheap manufacturing costs. Through processing controls, it is possible to achieve very high (90%) open porosity in these foams, which would provide for low-air resistance.
I know that metal foams have received some attention for cooling applications, but didn't know if anyone was playing around with them specifically for IC's. For example, check out:
http://www.metalfoam.net/commercial.html
It's been a while since my Heat and Mass Transfer course, so I'm not sure what layout would be most effective. I was thinking of either a sandwich-type structure going from top down:
1. metal foam
2. thin solid metal plate
3. CPU
Or a more cylindrical structure where a solid metal bar that interfaces to the CPU is surronded by a metal foam.
Anybody have any leads for me, or are we going to start a grass-roots effort on our own?
Regards,
Steve
Anyone Experimenting with Metal Foams?
Moderators: NeilBlanchard, Ralf Hutter, sthayashi, Lawrence Lee
I think someone said that they were interested in using a metal foam in a DIY hard drive heatsink, but gave up when they found out how expensive it was. A 48" x 48" sheet of 1" aluminum foam is US$385 at mcmaster.com. Granted, that big a sheet would be enough for about 50 hard drives, but I don't know of any place selling it in smaller sizes, which makes it hard for the tinkerer.
If you know somewhere where a person can get small quantities of it, please do tell!
If you know somewhere where a person can get small quantities of it, please do tell!
Well, after a little more surfing around, I found the following:
http://www.pocothermal.com/
This company has licensed some technology for graphitic carbon open cell foam. I haven't had a chance to read carefully, but it sounds like they do some processing to preferrentially align the carbon molecules in the foam. End result - super lightweight, with thermal conductivity higher than pure copper!
The problem, of course, is cost. A small slab of the stuff (2" x 2" x 1.5") costs about $150. I haven't done the calculations, but it might be enough to make a CPU heatsink out of.
It also looks like somebody did get a small trial sample from them a while back (when prices were higher), and then tried to do some CPU cooling.
Look here:
http://www.spodesabode.com/content/article/poco
They were shooting for a watercooling block material, but did get pretty impressive results using it as an air-cooled block as well.
So, it looks like this stuff is for real. I'm thinking about starting to pester them for an engineering sample, but I'm not hopeful. It would definitely be fun to play with! I'll also look into the metal foams.
http://www.pocothermal.com/
This company has licensed some technology for graphitic carbon open cell foam. I haven't had a chance to read carefully, but it sounds like they do some processing to preferrentially align the carbon molecules in the foam. End result - super lightweight, with thermal conductivity higher than pure copper!
The problem, of course, is cost. A small slab of the stuff (2" x 2" x 1.5") costs about $150. I haven't done the calculations, but it might be enough to make a CPU heatsink out of.
It also looks like somebody did get a small trial sample from them a while back (when prices were higher), and then tried to do some CPU cooling.
Look here:
http://www.spodesabode.com/content/article/poco
They were shooting for a watercooling block material, but did get pretty impressive results using it as an air-cooled block as well.
So, it looks like this stuff is for real. I'm thinking about starting to pester them for an engineering sample, but I'm not hopeful. It would definitely be fun to play with! I'll also look into the metal foams.