Hypothetical Cooling.

I was wondering when you'd spot this :)

Any idea on practicality of making copper foam? It's a nightmare to machine, and I'd expect it to form a very viscous melt, but I don't know the details of making these materials.

I think heat transfer from cpu to foam is achievable by having a single block of copper where one side was against the mould and therefore isn't full of holes.

And it's possible that sufficient pump pressure would achieve a reasonable flow rate through the foam without destroying the copper walls.

It's not in any sense a cost effective solution, but I'm not convinced that it's inherently inferior to the current designs for moving heat away from a processor
 
Makes me wonder why the finished prototypes were shown in 06 and theres no sign of them or mention of temperatures anywhere in your link.

Also makes me wonder if you understand any of the limitations posted above by AcidHell2, myself and mrthingyx.
 
Superewza said:
That article i posted is actually quite a good read. Makes you wonder why we're still using traditional heatsinks...
Click to expand...

I expect it's because airflow through sponge will be terrible, and you'd need a delta fan to see any benefit. Having the fins lengthways in the direction of airflow is an efficient way to get high surface area with minimal restriction to airflow, whereas in a sponge you have pockets of dead air where the fan won't aerate and which just insulate it. In fact, iirc, you don't necessarily see benefits from roughening heatfins, because the added surface area is negated by the air forming an insulating boundary layer in the roughened surface that is more difficult to disrupt and replace with fresh air, so the air stays there, insulating the metal. In standard heatfins louvres can be added (slits where the fin is then angled, as seen on radiator fins) These are angled so as to provide a fresh edge and disrupt the boundary layer of warmed air that has accumulated at the metal-area boundary as the air flows over the fin. Done right, multiple louvres can increase the heat transfer from a fin by around 30% (compared to a flat sheet of metal), but they need to be at the right angle so as not to impede airflow too much, because then total airflow drops and less heat is transfered. In a sponge the same sort of thing is happening, but the angles are not right (lots of sponge cell sides perpendicular to direction of airflow), so although there'll be lots of surface area, the airflow will be crap and total heat transfer will likely be poor.

Additionally, I expect it costs an awful lot more to make metal foam than it does to cut fins, press and solder them on.
 
Am I correct in thinking that water does something similar as far as a near stationary layer forming goes?

Further, would this be why the flow rate through a loop has to be above a certain number, in order to keep the static layer from becoming too thick?
 
Monkey Puzzle said:
I expect it's because airflow through sponge will be terrible, and you'd need a delta fan to see any benefit. Having the fins lengthways in the direction of airflow is an efficient way to get high surface area with minimal restriction to airflow, whereas in a sponge you have pockets of dead air where the fan won't aerate and which just insulate it. In fact, iirc, you don't necessarily see benefits from roughening heatfins, because the added surface area is negated by the air forming an insulating boundary layer in the roughened surface that is more difficult to disrupt and replace with fresh air, so the air stays there, insulating the metal. In standard heatfins louvres can be added (slits where the fin is then angled, as seen on radiator fins) These are angled so as to provide a fresh edge and disrupt the boundary layer of warmed air that has accumulated at the metal-area boundary as the air flows over the fin. Done right, multiple louvres can increase the heat transfer from a fin by around 30% (compared to a flat sheet of metal), but they need to be at the right angle so as not to impede airflow too much, because then total airflow drops and less heat is transfered. In a sponge the same sort of thing is happening, but the angles are not right (lots of sponge cell sides perpendicular to direction of airflow), so although there'll be lots of surface area, the airflow will be crap and total heat transfer will likely be poor.

Additionally, I expect it costs an awful lot more to make metal foam than it does to cut fins, press and solder them on.
Click to expand...

Well you obviously didn't read the article, because while it does cover metal foams there are several other unique designs for heat sinks and completely new cooling systems.

http://www.frostytech.com/articleview.cfm?articleID=2424

It's quite long though, so i just skimmed through it.
 
JonJ678 said:
Am I correct in thinking that water does something similar as far as a near stationary layer forming goes?
Click to expand...

Yeah, I think so - for laminar flow the water at the very edge (by the pipe wall) is assumed to be stationary (effectively trapped in the rough surface of the metal).

Further, would this be why the flow rate through a loop has to be above a certain number, in order to keep the static layer from becoming too thick?
Click to expand...

Not sure - could well be why older style blocks needed the water to flow fast enough to become turbulent over the copper. Impingement style blocks don't need the water to be going that fast to get turbulence as they shoot a jet directly at the copper, so I expect that's why flowrate isn't as important.

I think the flow-rate of 1gpm or whatever is more to do with diminishing returns from the radiators above that. Could be wrong though.
 
Superewza said:
Well you obviously didn't read the article, because while it does cover metal foams there are several other unique designs for heat sinks and completely new cooling systems.

http://www.frostytech.com/articleview.cfm?articleID=2424

It's quite long though, so i just skimmed through it.
Click to expand...

No need to get irate - I read the page you linked to on the metal foam/sponge for heatsinks.:) That's what you were talking about...

Foamed Metal Heatsinks and Heat Exchangers:

While this may look like sponge, what these panels and prototype heatsinks are made of is foamed metal, either copper or aluminum. The benefit of foamed metal is incredibly large surface areas, the downside is restricted airflow. High pressure fans are a must.
Click to expand...
 
Cheers. I'm a bit confused by impingement blocks at present, though a lot less than I was a couple of days ago. I think the issue with them is that their presence reduces water flow considerably relative to less restrictive blocks, so while they work relatively well at low flow rates, they also cause low flow rates. Watercooling is not as simple as I first thought
 
JonJ678 said:
Cheers. I'm a bit confused by impingement blocks at present, though a lot less than I was a couple of days ago. I think the issue with them is that their presence reduces water flow considerably relative to less restrictive blocks, so while they work relatively well at low flow rates, they also cause low flow rates. Watercooling is not as simple as I first thought
Click to expand...

Yeah, you are trying to put about half a dozen blocks and 3 radiators in one loop though. :D
 
I liked the look of the carbon fibre heat exchangers... and this:

hex2.jpg
 
Monkey Puzzle said:
Yeah, you are trying to put about half a dozen blocks and 3 radiators in one loop though. :D
Click to expand...

I'm a bit surprised this is so abnormal, it makes so much sense to me

one block good, two blocks better, all the blocks...
plus as many radiators as will fit, can't have spaces :)
 
You must log in or register to reply here.
Back
Top Bottom