THe Sandia Cooler - Breakthrough in Air Cooling design

[Arkadrel]

https://techportal.eere.energy.gov/technology.do/techID=527

Technology Marketing Summary:

In a conventional CPU cooler, the heat transfer bottleneck is the boundary layer of "dead air" that clings to the cooling fins. This insulating layer is largely unaffected by the impinging airflow generated by the fan. The radically different approach described herein overcomes this thermal bottleneck, generating a several-fold improvement in cooling performance in a device that is smaller, quieter, and immune to clogging by dust.

Description:

In this new device architecture, heat is efficiently transferred from a stationary base plate to a rotating (ccw) structure that combines the functionality of cooling fins with a centrifugal impeller. Dead air enveloping the cooling fins is subjected to a powerful centrifugal pumping effect, providing a 10x reduction in boundary layer thickness at a speed of a few thousand rpm. Additionally, high-speed rotation completely eliminates the problem of heat exchanger fouling. The "direct drive advantage", in which relative motion between the cooling fins and ambient air is created by rotating the heat exchanger, provides a drastic improvement in aerodynamic efficiency. This translates to an extremely quiet operation. The benefits listed below have been quantified on a proof-of-concept prototype.

Benefits:

Dramatic increase in cooling performance without resorting to exotic methods
10x smaller than current state-of-the-art CPU coolers
Exceptionally quiet operation
Immune to dust fouling
Simple, rugged, and cost-competitive design
Provides increased energy efficiency

Whats not said is that they have a prototype, thats abit over 10cm in diameter, that funktions at 0.2C/W. And they believe they might reach as low as 0.1C/W.

The fact that its 10 times smaller, can cool better than most air coolers, and use less power to do so? It almost seems to good to be true.

All this time.... fans blowing air on some metal fins... when its apperntly much more effective to just spin the cooling fins around through the air (instead of the otherway around)



A picture of the prototype:

 

[Ares1214]

How has nobody thought of this yet? Seems genius if the metal is light enough.

 

[Arkadrel]

How has nobody thought of this yet? Seems genius if the metal is light enough.

I have no idea.... its one of those.... "why didnt I think of that" type of things.

Non the less... this is bound to the be the future of air cooling.

Heat pipes? meh... clunky big old 900gram heavy coolers with a fan on them? wont sell at decent prices once these babies are out.

This might put a dent in water cooling too.

 

[Blue Shift]

How has nobody thought of this yet? Seems genius if the metal is light enough.

Transferring heat to the fins while they're spinning is probably an interesting problem. In this case they use an air gap that's as narrow as a thousandth of an inch, between the baseplate and the rotating assembly.

 

[Zink]

I don't see how this is different from a fan + heatsink setup as far as the "dead air". One is air flowing over the heatsink and in the other the heatsink moves through the air. The curved airflow path through the fins in the OP's design though would seem to reduce this issue.

 

[Ares1214]

This could be improved quite a bit, seeing as how this is just prototype.

 

[MarkLuvsCS]

If i was reading correctly, it seemed most of their test loads were around 40W. It could have a problem of dissipating heat loads of 100W+. I think it's awesome they have a design that may allow for even better air cooling then we have now. I wonder if by the time that we have newer cooling tech if processors won't reduce their heat generation a bit.

 

[bridito]

I've commented in the similar Cases & Cooling section thread. IMHO, this is pie in the sky. Justifications over there.

 

[Tsavo]

This design is going to collect lots of knuckles and fingers.

 

[TakeNoPrisoners]

Nice, maybe this will allow CPU makers to go a little more liberal in the heat generating department.

 

[EliteRetard]

LULz I already have one of those, except the fins are "welded" to their own baseplate and theres something like a really thin thermal greese between the fin plate and the baseplate.

 

[greenhawk]

This design is going to collect lots of knuckles and fingers.

Proberly, but then IIRC from the slashdot link about it yesterday(?), an idea is to place a "lid" on the fans as only air in from the middle count.

Personally, As it sucks in air from the middle (so limiting higher speeds / higher fin heights), then the internal space (diameter) would need to increase as the fin height increased, so possibly having a cup shape in the middle. (assuming enough thickness of the blades to allow the heat to rise that high).

Overall, the idea seems a bit "well derrrr", but the catch is getting the heat from the heat sink to the moving blades. That is where I see the interesting features comming from. (ie: axial fans have been around for a while).

 

[EliteRetard]

Oh and just like a blower fan its much easier to direct the outward airflow...just a cage spaced a bit from the blades and a hole for the pressurized air to pass through. The one Ive got also has some different angled blades (almost like regular fan blades, except they connect to the plate and have a higher angle) in the center to help pull air in and reduce hotspot. I think this thing came off a transformer or something, its much to big to use on a CPU.

 

[Nemesis 1]

Ya this is a neat way to cool,with air . I believe shortly also you will be seeing a new water cooling system for PCs that will likely change that industry. Even tho its been around along time.

 

[StrangerGuy]

What if that exotic heatsink is mounted vertically like in our nice PC tower cases? Don't tell me anyone didn't thought about that.

 

[drizek]

OK, now take two of these things and hook them up in a push-pull configuration to a regular tower heatsink.

 

[Dadofamunky]

Looks like a bench grinder.

I hope it realizes its potential in commercial PC applications (e.g. guys like us).

 

[Patrick Wolf]

What if that exotic heatsink is mounted vertically like in our nice PC tower cases? Don't tell me anyone didn't thought about that.

They did. From page 11 of the PDF.
"..can be mounted in an arbitrary orientation (e.g., up-side-down, sideways, etc.)."

http://prod.sandia.gov/techlib/access-control.cgi/2010/100258.pdf

 

[Binky]

Who writes this crap? Something cannot be "10x smaller" than something else! You can say 90% smaller, or 99% smaller, but 10x is meaningless.

/rant

This cooler looks like a circular saw. I'm assuming that the heatsink itself is spinning...yikes.

 

[sm625]

Fail. Note the large the relative mass of the spiinning portion vs the stationary portion. This cpu cooler will work fine for a couple years. Then it will wobble and grind and basically act like your typical 7 year old fan on its last legs. The greater the relative mass, the lesser of an imbalance it takes to destroy it.

 

[CU]

Fail. Note the large the relative mass of the spiinning portion vs the stationary portion. This cpu cooler will work fine for a couple years. Then it will wobble and grind and basically act like your typical 7 year old fan on its last legs. The greater the relative mass, the lesser of an imbalance it takes to destroy it.

That was my thought also. Seems easy to break it.

 

[Ratman6161]

I've commented in the similar Cases & Cooling section thread. IMHO, this is pie in the sky. Justifications over there.

Being its Sandia National Labs and thus a government project, it probably cost about $64 billion in development cost and will have a unit cost of about $43 million each to get in production.

I think I'll probably end up sticking with my $26.00 Hyper 212+

 

[Rifter]

Fail. Note the large the relative mass of the spiinning portion vs the stationary portion. This cpu cooler will work fine for a couple years. Then it will wobble and grind and basically act like your typical 7 year old fan on its last legs. The greater the relative mass, the lesser of an imbalance it takes to destroy it.

I also had this thought. Will be great shrt term and horrible long term. Think ill sick with my NH-D14.

 

[MustangSVT]

Did people take physics? its amazing how many think this will work.
This won't work better than current heatsink + fan system. Both in cost and performance.

The main issue is of course the part where blades meet the surface. Or simply, how the heat is effectively transferred. ANY gap and it will be worthless, too much friction requires other details that just isn't cost effective at all. not even close.

NEXT!.

 

[Arkadrel]

Did people take physics? its amazing how many think this will work.
This won't work better than current heatsink + fan system. Both in cost and performance.

The main issue is of course the part where blades meet the surface. Or simply, how the heat is effectively transferred. ANY gap and it will be worthless, too much friction requires other details that just isn't cost effective at all. not even close.

NEXT!.

There is a 48 page long PDF file, thats full of math and physics calculations that shows why it works.
Physicists with math > your beliefs.

There is a WORKING prototype, that SHOWS that the design works.
Working prototype design > your beliefs.

The prototype is ~10cm in diameter, and has a cooling efficiency of ~0.2 degrees Celcius pr Watt (0.2C/W).

The thing works, and it works better than current designs.