1. It's already been done (30% of my ideas fall into this category
)2. It's a harder, more expensive, less efficient way of doing it (50%
)3. You just give me a funny look and then carry on doing stuff (20%
)4. It's brilliant and someone wants to send me a big fat cheque (0% so far)
I was generally thinking about fan blades and how they curve to follow the acceleration of the air they are pushing. I was then thinking of the number of leading and trailing edges in a normal fan design. I've read about vortices and turbulence making the noises we hear (aside from clicking and bearing noise) so I was trying to think of a way of minimising these edges and turbulences.
What I came up with was essensially a frustrum of a cone but with the wide bit at the top, spinning through a vertical axis. There would essentially be one fan blade that forms a channel down the narrowing frustrum a bit like a helter-skelter...the floor of the first piece of channel forming the roof of the channel on the next 'lap' down the frustrum. This channel could accelerate the air over a number of turns with a slower smoother acceleration than normal fan blades, possibly eliminating some of the vortices and turbulence. I was also thinking that rolling the outlet like the bottom of an oboe or clarinet might help also.
The channel would start wide and flat, and would end up squarer but would keep the same volume ( i think that's how it should work anyway). Maybe it would need two channels like a double helix to give any pressure near the bottom, and maybe the shape would be more like a log curve rotated around the Y-axis.
The downsides I see are:
1. It might have to be huge to give any airflow
2. I haven't decided where the motor should go..maybe above the inlet, or rim-driven
3. manufacture of the shape might be tricky, or just plain expensive
I hope this bad description put the right images in your heads, I look forward to the cussing