"Tornado" Shaped Cyclone

[Anthony Smaldone]

Why not invert the tornado... Have the motor on top and the wide end of the Tornado on the bottom, emptying into a can.(sort of a lid if you will) lol Dont think it will work, but if your brainstormimg might as well go all the way.

[John Messinger]

I still like to see a sketch of exactly the shape you are thinking of. ....

I have looked at several photos of tornadoes over the past few days, and found that they are much like fingerprints. However, this seemed to be an attractive shape to me.

[Eric Gustafson]

I couldn't resist any longer, so I am throwing in my two cents.
I do not think that the cyclone needs to look like a tornado to work efficiently. I am not a scientist and I do get lost in all those mumbo-jumbo calculations, but I have the clearvue 1800 and the the mini clearvue and can watch them work. That observation plus some applied physics tells me a couple of things: One, friction must be slow down sawdust as it slides against the inside wall of the cyclone. However, slowing down the sawdust will not help separate the dust from the moving air.
Think about it, if you slowed down the dust completely, it would get picked up by the moving air in the cyclone and not even make it to the bin below. If I load my pickup with sawdust and drive down the road, moving air will empty the truck bed quickly! The inside of a cyclone, however, is like a centrifuge. The heavier sawdust is cast against the side of the cyclone and air gets sucked up the center. The taper at the bottom of the cyclone compensates for the energy loss the sawdust forfeits to friction, and energy spent in order to change direction. The taper at the bottom of the cyclone keeps the relative velocity of the sawdust high enough compared to the exhaust air to keep dust pinned against the cyclone inner surface until it reaches the bottom of the cyclone where there is virtually no air movement, the farther the dust gets from the air rushing into the exhaust the better, but there is a balance. The sawdust must reach the bottom before it slows completely, or it will get sucked back into the air vortex and expelled into the filters.
To see this in action, watch the first video of the Clearvue cyclone CV06. The sawdust is clearly not 'falling into the lower container. It is being propelled into the bin opening:
http://clearvuecyclones.com/Videos.htm

The goal of the cyclone, therefore, is to force the sawdust to the exit above the dustbin more quickly and forcibly than it has a chance to slow down and get sucked up by the air exhaust.
To answer the OP, would the dust/air separation be further optimized if the diameter transistion of the cyclone is made non-linear, as a tornado appears, rather than the linear sloped side of a common cyclone? Perhaps. Regardless, there is little to be gained. Anyone who gets the chance to experiment with a cyclone knows that they are already extreemely efficient. I still have my original bag in my festool CT-22, but I have emptied the 1 1/2 cu ft container below the clearvue cyclone twice already. There just isn't much added performance to extract.