The fanless heatsink: Silent, dust-immune, and almost ready for prime time

http://www.extremetech.com/computing...

The fanless, almost-silent, dust-immune, 30-times-more-efficient Sandia Cooler heatsink is almost ready for prime time. Sandia National Laboratories has announced that two companies — one computer heatsink maker, and one LED light maker — have licensed the technology.

Badass.

Pretty amazing. One of the things that concerns me though is that I've never seen it mounted vertically; it's always displayed horizontally and parallel to the ground. In most modern cases, it would need to be mounted vertically, which makes me wonder how many folks could actually take advantage of it.

That's pretty nifty. It'll be interesting if it ends up being practical.

AnimeJ wrote:

Pretty amazing. One of the things that concerns me though is that I've never seen it mounted vertically; it's always displayed horizontally and parallel to the ground. In most modern cases, it would need to be mounted vertically, which makes me wonder how many folks could actually take advantage of it.

It'll still be a money maker for them. A lot of cooling applications will still be horizontal. In the case of CPUs, rack mounted servers, for instance.

I'll switch to a horizontal case for something like this.

That does look quite cool, but no, I don't think that will ever work in a vertical application. But one of the larger reasons to make PCs towers in the first place is to get the noise further away from your ears, so going back to the old 'pizza box' style casing, with a nearly silent machine, wouldn't be that big a deal.

One problem, though, is that you couldn't use this tech on both the CPU and the GPU at the same time, unless we switch to a riser+horizontal card layout.

Malor wrote:

One problem, though, is that you couldn't use this tech on both the CPU and the GPU at the same time, unless we switch to a riser+horizontal card layout.

Or put both in the same package, or have the GPU as a socket on the same motherboard. The second is pretty unlikely, and the first currently only viable for low-mid end graphics.

Imagine this in a future console.... No more noise!!

I didn't see what makes this dust immune? It's only dust 'immune' when it's moving so, just like most fans that get clogged with dust in consumer electronics, so will this.

I also don't see why it couldn't be vertical. If it's balanced on the central strut/prop then there's nothing stopping the impeller from working in a vertical manner. The brushless, contactless electric motor we used for one of our projects at uni was mounted vertically and it was held in place by a central bar like this one is.

Not strictly fanless given that it is a very smartly designed type of fan. V. Smart though

Duoae wrote:

I also don't see why it couldn't be vertical. If it's balanced on the central strut/prop then there's nothing stopping the impeller from working in a vertical manner. The brushless, contactless electric motor we used for one of our projects at uni was mounted vertically and it was held in place by a central bar like this one is.

See also jet engines.

Looks cool. One question for the engineers: how is an impeller not a fan?

Maq wrote:

Looks cool. One question for the engineers: how is an impeller not a fan?

I'm pretty sure it is in this case ("fan" defined as a number of rotating surfaces that displace air and produce a current). I think what they're referring to is that there's no fan on the heatsink.

With my rough-shod wiring skills this would end in nothing by tears...

The fan is the heat sink. The heat sink is the fan.

(Moggy, there's something messed up in your quote tags there.)

They say that the motor noise will be completely inaudible once the outer casing is on.

One thing that I've been wondering about, though, is that they say that the impeller spins on an air cushion. Wouldn't that insulate it, and prevent it from dissipating very much heat?

Malor wrote:

(Moggy, there's something messed up in your quote tags there.)

They say that the motor noise will be completely inaudible once the outer casing is on.

One thing that I've been wondering about, though, is that they say that the impeller spins on an air cushion. Wouldn't that insulate it, and prevent it from dissipating very much heat?

Yeah, that's where my skepticism lies... they had like a 2-sentence "fluid dynamics mumbo jumbo" explanation around the middle of the clip, which didn't satisfy me. I'm sure that's probably where the meat of their presumed patents lie, and probably doesn't make for a great video tech demo, but still...

I get it. My fans just need to spin at 2000 rpm backwards.

There’s a nice whitepaper on the Heatsink technology…..
http://prod.sandia.gov/techlib/acces...

Their release documents says it rotates at 5k not 2k
https://ip.sandia.gov/techpdfs/Sandi...

As for the heat transfer between the spinning impeller and the stationary base, that’s the whole point of the impeller spinning only 0.001” from the base. The insulation effect of the boundary layer of air is reduced by a factor of 10 so heat transfers into and across the air gap to the impeller.

I thought that was only on the fins... and I'm having trouble imagining that even a thin layer of air is going to be anywhere near as effective as a layer of heatsink compound.

Duoae wrote:

I didn't see what makes this dust immune? It's only dust 'immune' when it's moving so, just like most fans that get clogged with dust in consumer electronics, so will this.

Unlike regular heatsinks, where the dust will settle between the vanes and reduce efficiency, the dust will be shaken off when it starts to run.

It really didn't seem that much quieter to me. Sure, the impeller was quiet, but the motor they were using was noisy. It'll take more of a motor to spin a heavy heatsink shaped like a fan than to spin a lightweight fan.

Edit: Thanks Malor

So without reading the whitepaper because I'm lazy, does this device only work one-way-up? How does it cope with a tower format case? Or when being moved?

DanB wrote:

Not strictly fanless given that it is a very smartly designed type of fan. V. Smart though

Duoae wrote:

I also don't see why it couldn't be vertical. If it's balanced on the central strut/prop then there's nothing stopping the impeller from working in a vertical manner. The brushless, contactless electric motor we used for one of our projects at uni was mounted vertically and it was held in place by a central bar like this one is.

See also jet engines.

The issue isn't the motor, it's the impeller. The impeller isn't actually connected. As for transfer from the chip, I wager you still need heatsink compound for that while fluid dynamics veteran heat to the impeller.

The key thing to think about regarding the "air gap" between the spinning and stationary portions of the heat sink:

The levitation effect is caused by molecules bouncing around a lot smashing into both sides of the tiny gap between the two.

Heat is transferred by molecules smashing into things.

It's a very small gap, and the force with which the smashing is going on is strong enough to exactly counter-balance the force of gravity. That's not an insignificant amount of force. I am not at all surprised that heat transfers efficiently across the interface.

Hmm, good point. It's not just a passive interface, it's an active one, with weight on it.

I took a heat transfer and thermodynamics class back in the day. It's really fascinating stuff. I can't remember much from it, unfortunately.

I wonder how well this tech could be implemented in a laptop, as they are mostly horizontal.

Probably not at all, I don't imagine it will appreciate being moved.

Malor wrote:

Probably not at all, I don't imagine it will appreciate being moved.

Why not?

Well, a heavy device spinning that fast is going to be basically a gyroscope, and moving it around will cause all kinds of weird forces to be exerted on the laptop's internals, and will cause the laptop to move strangely.

You could probably have the equivalent of a hard drive head crash pretty easily, where the spinning bit drags on the surface underneath -- not likely to be as damaging as it would be on a hard drive, but I can't imagine the scratches would be good for cooling. (otherwise, they would pre-scratch it, no?) And if it leaves a horizontal position, like if you prop the laptop on your chest in bed, I don't think it will work at all.

Oh, an assumption I didn't really think about is that the cooler needs to be fairly large... if it could be much smaller than the one they're showing in the video, then you probably wouldn't notice the gyroscopic forces, and it would be harder to damage by moving around. But I still don't think it will work properly if it's vertical, and that's a fairly common use case for laptops.

Malor wrote:

but I can't imagine the scratches would be good for cooling. (otherwise, they would pre-scratch it, no?)

Might be that the math to get the pattern of scratching right is way too hard/time consuming

This is not a gryoscope.

wikipedia wrote:

A gyroscope is a device for measuring or maintaining orientation, based on the principles of angular momentum.[1] Mechanically, a gyroscope is a spinning wheel or disk in which the axle is free to assume any orientation. Although this orientation does not remain fixed, it changes in response to an external torque much less and in a different direction than it would without the large angular momentum associated with the disk's high rate of spin and moment of inertia. Since external torque is minimized by mounting the device in gimbals, its orientation remains nearly fixed, regardless of any motion of the platform on which it is mounted.

There will be a "correcting" centripetal effect but that should not materially affect the distance of the air pocket (or bearing as the guy in the video above calls it) below the spinning heat sink portion because it is, as is mentioned at around 2:21 in the video, the spinning itself that causes the separation of the two surfaces.

Malor wrote:

Oh, an assumption I didn't really think about is that the cooler needs to be fairly large... if it could be much smaller than the one they're showing in the video, then you probably wouldn't notice the gyroscopic forces, and it would be harder to damage by moving around. But I still don't think it will work properly if it's vertical, and that's a fairly common use case for laptops.

As pointed out before - if they're planning to use this tech in solid state lighting fixtures then there's no way it'll be even as big as it currently is.

I also don't think it's that heavy, tbh. A lump of aluminium of that size weighs almost nothing.