[elhigh]

The two gases aren't that dissimilar in their molecular weight, nitrogen is about 88% of oxygen by weight.

I'm not very good at this advanced stuff so let me just play with the math on my back-of-the-envelope level.

80,000 rpm
15 cm diameter = 176 cm3 per cm of centrifuge length

Let's assume for the sake of charity that the centrifuge is 100cm long. That gives it a volume of 17,600 cm3.

Now let's assume a perfectly average 2-liter four-cylinder car being fed by this contraption.

Every revolution of the crank consumes (ideally) 1,000cm3 of air. So in 18 revs it uses the entire volume of the centrifuge.

Let's assume that engine runs at 3000rpm when cruising at 60mph. It's a little on the buzzy side for such a modest engine, but I'm trying to keep the math simple. In one second, it turns 50 revs. It consumes the entire volume of centrifuge three times.

Granted the centrifuge is generating over 53,000 G - and I'd be curious to know what they plan to make this out of, and whether they can make it lightly enough, plus its containment, that any benefit accorded by it isn't overshadowed by its own weight and support systems - but with the two gases so close in molecular mass, is it possible to achieve useful separation with that kind of throughput? The device has only one-third of a second to separate the two in a really short stack height, not a lot of room for stratification there.

And finally, any kind of harvesting system that pulls the separated gases must generate some turbulence that might well completely undo all the hard work of the separation in the first place.

With 53,000 G's to play with, I think this would be a LOT more useful as a plain ol' supercharger.

I'm kind of a hack at this stuff, if anyone can (gently) point out where I have gone horribly wrong I would be very interested to hear it.