The Fanless Spinning Heatsink: More Efficient And Immune To Dust

Looks very promising, but its tiny heat transfer air boundary between a fixed heat-generating object and the rotating heatsink, would seem to me to be the biggest hurdle for long term reliability in a real world product (for example a laptop subject to orientation change and movement).

Why is that not a fan? And how does it eliminate boundary layers?

This is a fan, but an unusual one as the blades are heated via the small air-gap.

The air in the gap is sheared heavily so heat transfer is much better than on low shear air to cooler interface of ordinary fans.

Centrifugal forces (acting on the air) help in transportation of air so give good heat transfer (low thermal resistance) from blades to environment.

Maybe obstacles: the airbearing gap is said to be 25 to 30 µm, so any dust particle is allowed to be maximum 1/3 of this or 8 to 10 µm. So a good filter is necessary to remove coarse particles (cigarette smoke and bigger).

And: known good airbearings have gaps in-between 15 and 5 µm (and below) so it has to be calculated if the low load in this application tolerates a big airgap.

In the example shown in the Sandia-article the airbearing is externally pressurised - not suitable for mass applications.

But self pumped airbearings (made by Speedring for example) are in use in many applications with quite good results. Good scanners, FLIRs, video-tape-drums use these. But: these applications usually have enclosures for the airbearing so "no dust".

This problem may be solved by the outer part of the airbearing pumping outward and the inner part in inward direction thus having the possibility to have an inlet filter.

Another maybe problem: this is a spinning rotor on an air-gap that has flexibility, so a flexibly supported rotor. This may have - depending on speed - considerable gyroscopic movements if excited by angular movements of the frame.

In total: I would bet this will be successfull and very much worth while!

Below: one of my own experiments on grooved airbearings.

RHABE

Air cooling, even with impingement to strip off the insulating boundary layer, has a very small heat flux (<1 W/cm2) because there are very few molecules working on the job of carrying away the heat. So even the good idea of spinning the heatsink instead of blowing air at it is still going to fall short.

Liquid cooling, where many molecules can contact the heated surface, is going to be the ultimate solution. The highest heat flux cooling method of all is forced convection boiling (100 W/cm2). The incipient boundary layer of vapor is immediately stripped away bearing its latent heat to the ultimate means for heat rejection, thus preventing the vapor from superheating.

Surely in this arrangement the thermal resistance between the hot chip and the spinning heatsink is compromised.

Also with the heatsink itself spinning there is considerable angular momentum involved and therefore balance becomes most critical .....vibration - fatigue of components and pcb....

There would need to be some form of protective cage over the heatsink to keep fingers out and the rotating heatsink in!

Perhaps the thermal gains would be less than the other costs involved ...... but are there overall thermal gains between the die and ambient?