Heatsink design- the fins

[fri2219]

What I don't get...

Why don't they mimic biological systems and put in as many bumps and squiggles on the fins as possible, in order to increase the surface area? That would reduce the need for 1kg chunks of case filling metal, cooled by Pratt and Whitney Turbofans.

Some possible answers:
1) Lack of imagination
2) Cost of manufacture
3) The resulting airflow would be so turbulent it would actually reduce cooling

Some possible rebuttals:
1) That couldn't possibly be it- there's almost no lazy copying that goes on in computer component manufacturing and marketing, right?
2) There has to be some sort of stamping process that would allow them to be produced without having to smooth/machine them off afterwards.
3) The right bumps and nooks in the right places actually enhance laminar flow, at least with vertebrate lungs, insect wings, and whale flippers.


Anyone else see opportunities here?

[jaganath]

From a previous thread on a related subject:

improving heat transfer

Jumper@I think that the increased pressure loss due to the rough surface and the corresponding decrease in mass flow rate on a large scale will cancel out any improvement due to increasing the area of the fins on a microscopic scale.

The right bumps and nooks in the right places actually enhance laminar flow, at least with vertebrate lungs, insect wings, and whale flippers.

right, but for better cooling you actually want to disrupt laminar flow and make it turbulent. the right bumps and nooks can do that too, but according to the above thread the airflow through an ordinary heatsink is probably turbulent anyway, or transitions to turbulent almost immediately.

[Felger Carbon]

A perfectly flat .03" or .04" aluminum sheet bends very easily, as any Ninja owner can confirm. Take such a sheet and put a 90 degree bend in it near the edge. Now try to bend it!

If you will notice, the bottom and mobo support plate in your computer case have impressed depressions. This stiffens the plate. Side panels have the bends at the edges or they would be insufferably floppy.

Have any of you noticed how "stiff" the ZeroTherm HSF fins are? There's an embossed depression around the edges, just like the ones on the bottom of your case. Even though this depression is very shallow, it's enough to stiffen the cooling fin tremendously. Whether it affects the cooling one way or the other I dunno.

[fri2219]

Sorry about my inexact language, I was trying to keep it layperson friendly. I wasn't thinking of macro-modifications, but something on the scale of 10^-4 - 10^-6 m.

By squiggles, I mean cut holes in the the fins, in the middle and the leading edge to increase the surface area available for exchange, as in Brush border or Microvillus structures. It might be fun to vary things with testing circular perforations with variable pore sizes in the 10^-4 m range.

A s for the bumps, one could probably tune the ridges and bumps to induce laminar/turbulent flow as one saw fit- plenty of examples of either being exploited in nature. If I had the patience, I suppose I could tear apart a heatsink and spackle it with some sort of pliant, conductive, compound, like thermal paste.

It might be fun try it for a project between Christmas and New Year's- I have enough spare parts lying around to give it a try, and report back anything I find, we'll see about the time. The holes would definitely be easier to do in a reproducible manner, and it would be pretty straightforward to use null hypothesis testing to make the conclusions rigorous.

[Cerb]

I have seen heatsinks with ruffled fins, jagged edges, and some with holes (TR's newer ones use holes). So, some ideas:

0. Holes in the fins work. It makes sense, and TR has a few coolers with them. These coolers perform well in quiet conditions.

1. It has to be done just right, or it could cause it to be less effective. Biological systems have gotten where they are by trial-and-error over more samples than any of us can conceive. There's a huge R&D factor involved, and possibly a more expensive product from it.

2. Fancy designs may not be as effective as better materials and simple tried-and-true solutions. Look at how soldered coolers still do great, despite the hyped advantages of skyved fins, chemically bonded parts, and milled blocks. A better heatpipe and small bends for how the fan pushes air in do the trick. Spacing fins out but making them huge does well.

3. I wouldn't be surprised at all if a high surface area design is a crazy dust magnet. Smooth fins are certainly going to be easier to keep clean. If your surface is covered with dust most of the time, you've negated that surface area advantage, and possibly even created a worse cooler.

[tehfire]

I remember a news post a while back about ionic wind and how heatsinks suffer from a lot of inefficiency due to the fact that air molecules tend to stick to a heatsink's surface, and if these molecules moved out of the way HS effectiveness would go way up. This could be the case against having rough surfaces on a HS - it would cause more drag and make the air more stagnant.

[nightmorph]

--

[jaganath]

If your surface is covered with dust most of the time, you've negated that surface area advantage, and possibly even created a worse cooler.

good point. possibly a simple solution that would have the greatest effect would be some kind of anti-dust coating; it's inevitable in a non-filtered aircooled system that dust will accumulate, and you can't always count on the user to clean it out on a regular basis. ISTR some researchers have developed some nano-textured self-cleaning surface based on the lotus leaf?

http://www.physorg.com/news88088727.html