Slashdot Mirror
The Fanless Spinning Heatsink
An anonymous reader writes "There's a fundamental flaw with fan-and-heatsink cooling systems: no matter how hard the fan blows, a boundary layer of motionless, highly-insulating air remains on the heatsink. You can increase the size of the heatsink and you can blow more air, but ultimately the boundary layer prevents the system from being efficient. But what if you did away with the fan? What if the heatsink itself rotated? Well, believe it or not, rotating the heat exchanger obliterates the boundary layer, removes the need for a fan, and it's so efficient that it can operate at low and very quiet speeds. That's exactly what the Air Bearing Heat Exchanger, developed by Jeff Koplow of the Sandia National Laboratories, has developed. It's even intrinsically immune to the build up of dust and detritus!"
I think a better description would be a heatsink that is a fan or probably more accurately an impeller but without the tube enclosure.
These comments are my own and do not necessarily reflect the views or opinions of my employer or colleagues...
I think the distinction between a traditional fan + heatsink combo and what is described in the article is that the impeller blades are dissipating the heat instead of merely blowing cooler air over the fins of a stationary heatsink.
These comments are my own and do not necessarily reflect the views or opinions of my employer or colleagues...
But think of all the homeless Dust puppies!!!
have you no shame?
You sound angry. At thermodynamics.
I think he just had an exothermic reaction.
yes. the spinning heat sink is attached to a spinning cpu, which is in turn attached to a spinning motherboard mounted to a spinning case. these are only available in funhouses btw.
The stories and info posted here are artistic works of fiction and falsehood.
Only fools would take it as fact.
This is all explained in the article and PDF.
Don't come here and start mouthing off like you know what you're talking about when you clearly are too lazy to get past the summary and expect everyone else to do the work explaining it for you. You must be an MBA graduate.
Some of what I say is fact, some is conjecture, the rest I'm just blowing out my ass...you guess.
You sir, need a hug. It's a pretty good article.
*Man* you read and analysed those 44 pages of maths quickly.
That's exactly what the Air Bearing Heat Exchanger, developed by Jeff Koplow of the Sandia National Laboratories, has developed.
So I get it was not Jeff Koplow who developed it, but the Air Bearing Heat Exchanger did develop it. The Air Bearing Heat Exchanger in turn was developed by Jeff Koplow.
Oh, and BTW the link was missing a PDF warning.
The Tao of math: The numbers you can count are not the real numbers.
If you read the PDF article, it says it works like an air-hockey puck or hard-drive platter, there's an extremely thin layer of air between the spinning surface which is under high shear which is conducive to heat conductivity. The PDF goes on to explain that this dramatically reduces the size of the boundary layer, not eliminate it as the summary says (since this isn't even logical what I remember of my fluids class). I didn't read the whole thing but I think it's the fact that the heat sink blades themselves are spinning at very high speeds, rather than having dirty air blown on them, that prevents dust build-up.
... made from the hooves of a unicorn.
Godaddy is a scam and a ripoff.
Gamingmuseum.com: Give your 3D accelerator a rest.
+1 sweetness factor, such a rare moment in /. land.
Life is a great ride, the vehicle doesn't matter
I've read the paper and what you said is just silly, not insightful. The heat sink is separated from the base plate by a layer of air on the order of 1E-5m thick. This layer of air experiences large shear stress that keeps its thermal resistance low. It's basically an air bearing for the spinning heat sink. The stackup is thus:
1. CPU
2. Disk-shaped base plate
3. Air gap
4. Heat sink impeller
The major difference is that in normal coolers, fan has no heat dissipating function at al. There's no functional heat flow through the fan. In this design, the fan is the heatsink: heat does flow through it, and that's what makes it work so well.
From what I can tell, it's a truly revolutionary device. It has 5-10x lower thermal resistance than regular coolers, consumes ~5x less power than coolers of same capacity, and generates less acoustic noise to boot (it wasn't quantified, though). Ah, and also it doesn't get fouled by dust: ever notice how in usual CPU coolers the fan is usually clean or just sprinked with dust, when the heatsink is pretty much plugged with dust? In this device, the heatsink spins, so it stays clean, just like a fan would.
Whoever commercializes this for the HVAC market will be financially set, as in "playboy mansion" financially set :)
A successful API design takes a mixture of software design and pedagogy.
Essentially, they claim the really thin (~0.03mm) layer of air between the stationary plate and the rotating heatsink is thermally conductive and agitated by the rotation, so no static boundary layer.
Nope, the "connection" is a thin (1E-5m) air gap experiencing high shearing and thus providing very low thermal resistance. The gap's thermal resistance contributes very little (on the order of 10%) to the overall thermal resistance of the cooler. It is a truly revolutionary design, no shit here.
A successful API design takes a mixture of software design and pedagogy.
it's a truly revolutionary device
haha! Excellent pun.
SJW n. One who posts facts.
Yes, a layer of air does form between the heat spreader base, and the base of the rotating heatsink. This is called an air bearing. It's extremely thin, and for that reason an excellent thermal conductor even though it's conducting heat poorly. You see, it has a surface area of 100 cm squared, but it is less than 0.03 mm thick. So, heat transfer is inefficient, but its so thin as to be negligible.
And no boundary layer forms (well, it does but it is reduced by a factor of 10) on the fins because they are rotating. The equations for fluid dynamics are quite different between an inertial reference frame and a rotating one. Basically, the fluid cannot settle into little pockets because the (fictional) centripetal force is pushing it outwards along the fin channels.
ASCII stupid question, get a stupid ANSI
You're wrong.
Basically, the layer of air between the thermal spreader (base plate) and the impeller if very thin and very turbulent, because it is 'grinded' between the the impeller and the base plate. That actually makes it a very good heat conductor.
It's explained very well in the Sandia Labs paper. Seems like a very plausible and good design.
assignment != equality != identity
I like the "protective wire mesh container". Tell me it's not an office trash can.
I thought a long standing goal of PC manufacturers was to do away with moving parts. I dont think fans will go away anytime soon as long as they are cheap to replace. From the comments hear I'd assume this heatsink spins on a platter essentially taking the place of the fan. What do you do when it fails? Can you replace it for less than $10?
I had the same question but it is very well addressed in the PDF:
During operation, these two flat surfaces are a separated by a thin (~0.03 mm) air gap, much like the bottom surface of an air hockey puck and the top surface of an air hockey table. This air gap is a hydrodynamic gas bearing, analogous to those used to support the read/write head of computer disk drive (but with many orders of magnitude looser mechanical tolerances).
Heat flows from the stationary aluminum base plate to the rotating heat-sink-impeller through this 0.03-mm-thick circular disk of air. As shown later in Figure 18, this air-filled thermal interface has very low thermal resistance and is in no way a limiting factor to device performance; its cross sectional area is large relative to its thickness, and because the air that occupies the gap region is violently sheared between the lower surface (stationary) and the upper surface (rotating at several thousand rpm). The convective mixing provided by this shearing effect provides a several-fold increase in thermal conductivity of the air in the gap region.
The PDF also goes into how this tech could have serious applications in things like home AC and refrigerator heat exchangers as well.
I know this is /., but the lack of TFA reading comprehension for this article is crazy even for slashdot standards.
The PDF never claims that the spinning heat sink eliminates the boundary layer. They only claim that spinning the heat sink reduces the boundary layer thickness by several orders of magnitude. And it makes sense, as the speed of air over the impeller blades moving at several thousand RPM is quite a bit faster than the speed of air that can be pushed over a static heat sink by a traditional fan. The faster the airflow, the smaller the boundary layer. The only way to get air moving that fast over a static heat sink is with a jet, and establishing such a jet in a computer enclosure environment exceeds reasonable power requirements.
Good Lord. Have your psychiatrist adjust your dosage.
TL;DR version: Stationary heat spreader surface on top of the IC. Teensy tiny air gap, small enough to permit heat transfer while functioning as an air bearing between heat spreader and... the next part, a heat-absorbing rotary impeller which pulls heat through the air gap into its fins, which are in turn cooled by air flow caused by centrifugal acceleration of the air through the rotating impeller assembly (squirrel-cage-fan style).
I'm not gonna pretend that there's no boundary-layer effect over the impeller blade surfaces, but I expect it'll be less than the effect caused by the common "push air down into the cooler and have it decelerate and turn 90 degrees to exit" cooler. Flow-through coolers would be more efficient than that, but air still has to decelerate through the cooler, whereas this impeller cooler makes the air accelerate during the cooling action. That might make a difference.
How well do bearings conduct heat?
The generic answer is "depends on thickness of air bearing surface (i.e., how big of an air gap), coverage area of bearing surface (i.e., is the heat spreader the size of the entire impeller, or just the small central portion of it), and the rotational speed of the rotating part on the other side of the gap -- moderate rotation speeds, in the 2k to 10krpm range, make the air in the gap turbulent and sheared rather than laminar, forcing mixing and heat transfer.
WTF happened to /.
Well, in this case, an actual scientific research article of relatively high coolness and technical merit leaked past the editors. I understand how this could be upsetting to most slashbots, given the novelty and rarity of this type of thing. Certainly, t
Welcome to the Panopticon. Used to be a prison, now it's your home.
*Man* you read and analysed those 44 pages of maths quickly.
I skimmed them.
It seemed to very carefully avoid the issue of the bearing's heat conduction ability while explaining how spinning a heatsink does reduce its thermal resistance vs merely blowing air upon it. So you decrease the resistance at one end while ignoring the increase at the other. Hmm.
The other mystery is the straw dog of cheap and easy to machine heatsink designs (you've all seen them) have moderately bad boundary layer problems, so rather than a more elaborately modeled and machined heatsink design, or even more simply, a larger heatsink, the solution is a very complicated, hard to model, and hard to machine rotating heatsink. So, why not just put the hydrodynamic engineering hours and CNC machining hours into a GOOD passive sink that might work just as well? Or invest in a couple more dollars of aluminum, or skip it all and go for broke with waterblocks. Who knows?
Is there a middle ground for this design to live in between cheap and easy and inefficient non-moving sinks and much higher performance (and cost) waterblocks? I'm guessing, no. Not in any electronic system I've worked on (not just computers, but high power RF amps, high power audio, high power VFDs, etc)
The other problem is it makes for a more brittle design. Now you can usually shut down a system automatically when the cooling system stops, due to thermal mass, limited natural convection cooling, etc. With this, it'll be smaller and lighter, can you shut down in time to avoid frying the CPU (physically) or crashing the filesystem? Its going to make OTHER parts of the system design more complex, not just the cooling system.
Cool engineering (pun intended) but I'm unimpressed from an economic standpoint. It will probably cost more than the alternatives. Unless you're just trying to avoid a patent or whatever.
"Science flies us to the moon. Religion flies us into buildings." - Victor Stenger
A 1 micron air gap... Now not only can the heads on our harddrives crash, our heatsinks can crash, too.
The living have better things to do than to continue hating the dead.
No, the heat is conducted from the CPU to the baseplate, across the 0.03mm airgap, into the baseplate of the "fan", and up into the spiral fan "blades". The heat is actually transferred to the air from the sides of the "blades", ant the warmed air is flung out into the environment.
Consciousness is an illusion caused by an excess of self consciousness.
A head crash is significant because the magnetized substrate is damaged, frequently destroying the data which was stored on it. Here a minor scratch might occur resulting in a who gives a shit. Its two completely different events.
(Emphasis added.)
How well do bearings conduct heat?
Again, the technical document makes it clear that the rotating heat sink is not coupled via a bearing to the surface it's cooling. Rather there is a very thin layer of air separating them. Naively one might think that this layer of air (generally a poor heat conductor) would become limiting, and there would be poor heat transfer from the hot plate to the rotating heat sink. However they address this:
So, basically by keeping the air gap very thin (30 microns), and by substantially shearing/mixing this thin air disk, its thermal conductivity can be sufficient to transfer heat up into the rotating fins. Overall a rather clever design.
WTF happened to /.
I agree a lot of junk gets posted to Slashdot. But in this case, a link was actually provided to a good technical document that answers many questions, provides schematics, and shows graphs of various performance measures.
They do not avoid the bearings heat conduction ability:
As shown later in Figure 18, this air-filled
thermal interface has very low thermal resistance and is in no way a limiting factor to device
performance; its cross sectional area is large relative to its thickness, and because the air that
occupies the gap region is violently sheared between the lower surface (stationary) and the
upper surface (rotating at several thousand rpm). The convective mixing provided by this 11
shearing effect provides a several-fold increase in thermal conductivity of the air in the gap
region.
I see no reason why this should not be subject to cost engineering like any other component. Yes, they machined their prototype our of solid aluminium on a CNC machine. But they are not production engineers. It would seem to me that, as one-moving-part system, this should be subject to manufacturing optimisation over two or three years to be very competitive with equivalent air cooling systems.
Your point about the design being more brittle is relevant: failure of the drive motor will lead to serious overheat in a very short time.
Consciousness is an illusion caused by an excess of self consciousness.
It sounds like it's still in the research phase, which means it's not a viable commercial product yet. All the things you describe would need to be solved in order for it to be commercially viable. But the concept is novel, and deserves credit for what it is.
Whether this will ultimately end up being a replacement or a competitor in the current cooling systems market will be a matter of whether these problems can or need to ultimately be solved. Since this phase of the research deals only with the viability of the new design, I suspect it will be.
"If a nation expects to be ignorant and free in a state of civilization, it expects what never was and never will be."
From TFA, highlighted for your convenience...
The cooler consists of a static metal baseplate, which is connected to the CPU, GPU, or other hot object , and a finned, rotating heat exchanger that are cushioned by a thin (0.001-inch) layer of air. As the metal blades spin, centrifugal force kicks up the air and throws it up and outwards, much like an impeller, creating a cooling effect.
The bearing between the fan and the plate is a very small air gap. Because it's small, and because it's constantly being churned around, it's thermal resistance is low.
Because the movement fan part destroys the normal zone of still air around radiator fins it dissipates heat more quickly and efficiently.
I know its said a lot and should be common knowledge but I think it pays to stress it more strongly on occasion. This seems like an ideal time, READ THE FUCKING ARTICLE.
Several posts now, numerous mod points and dozens of follow ups all frankly making complete asses of themselves ironically complaining about how the IQ of /. has dropped while they make angry complaints and rants about the story that are fully addressed in the documentation.
and if you think that the fact that the summary screwed up is still a good sign of /. intelligence drop then you really need to look right back in the archives because bad summaries have been around on /. and virtually everywhere else pretty much from the beginning. Unsurprisingly the people posting the stories dont have total knowledge of the often fairly complex material posted and they screw it up good and proper on occasion. Which is probably why you should be judging the posts on the documents they link to and not the quickly thrown together summary by an admitted layman. Anything else is ironically a really stupid thing to do.
READ THE FUCKING ARTICLE.
(and no this doesnt mean the documentation is flawless but make commentary on that, not the summary, it will raise that intelligence level a lot of you are so eager to whine about.)
Basically, the fluid cannot settle into little pockets because the (fictional) centripetal force is pushing it outwards along the fin channels.
Oblig XKCD
The exception to this rule is lap top computers, where available electrical power is extremely limited. In this special case, CPU clock speeds and fan rotor speeds are reduced to conserve power, albeit at the expense of CPU performance. At these low fan speeds the residence time of air in the heat exchanger is greatly extended, resulting in much higher exhaust air temperatures.
I will create a sig when innovation restarts in the U.S.
also in the pdf they say that the air bearing effect on this is self regulating and allows for precision without the tight mechanical tolerances a hdd requires.
'...if only "Jumping to a Conclusion" was an event in the Olympics.'
The obvious solution would be to cause the spin to force air into the gap, so that the device is self lifting when operating. Use magnetic induction to do the spin and it pretty much can't fail. (Or at least, would be less prone to failure than existing designs, which should be all anyone cares about).
"Who is the Journal of Quantum Physics going to believe?" --Stephen Hawking
My first thought looking at the linked paper was, "January 2010? Why hasn't this been all over the place if the results are so promising?"
But there could be lots of reasons for that. Just sort of popped out.
Another question that comes up is overall system efficiency. One advantage of the current fan + heat sink paradigm is that in addition to moving air across the heat sink, which is not terribly efficient, the fan also serves to mix the heated air around the heat sink with the larger reservoir of surrounding air. They don't really directly discuss it, but my impression is that their design would result in very little large-scale mixing; one of the efficiency advantages is that they aren't moving large amounts of air around. It seems that in a setting like a CPU cooler this might be a non-issue, as you would still presumably be using case fans to move air through the case (exchanging the reservoir); but in something like an air conditioning unit, it seems like it would become limited by convection for larger scale heat transfer - or require an external fan for air exchange.
Basically, my concern is that while this method might be very efficient in moving heat from the base plate to the air in the immediate vicinity, you would then have the problem of heat building up around it. Perhaps not an issue where you have a small-scale device open to a large, room-temperature environment (or where you already have something in place to move air around, like a computer case), but it seems like it could be an issue moving to something like a residential air conditioner.
Still, it appears to eliminate the boundary layer problem, so you could use a pretty efficient, large, slow-moving fan for air exchange - so probably more efficient than blowing high-speed air across a heat sink, but something that would need to be considered in a full implementation.
You probably couldn't use spinning hard drives but SSD should work fine.
This problem is solved by keeping the platters stationary as the case spins about them.
I wonder how jostling of the PC case might affect this 'bearing'? Is it possible that the spinning disc would then impact the static metal base plate?
Seems like this concept would already have been dealt with via spinning hard disk platters, but those don't weigh nearly as much as a heat sink and are segmented out specially with shock absorbers in the HD enclosure.
People in cars cause accidents....accidents in cars cause people
That "boundary layer" theory is well justified by models and experiments. The lack of a fan does not make it disapear, but rotating the heat sink does make it smaller.
About dust, it only accumulates over a hight speed rotating surface (where the air speed is increasing) up to a certain amount. That amount is certainly way lower than the dust that accumulates at a static place where the air speed drops. Immune do dust is a justfied simplification, altough a bit over confident.
Also, I completely agree with your point about the dust accumulating at the gap. That device won't last for long at the real world.
Rethinking email