Model originally uploaded to Thingiverse at https://www.thingiverse.com/thing:382809.
This is a continuation of my fan experimentation that I started with http://www.thingiverse.com/thing:381529
What happened to version 2? Well, strike it up to a lesson learned with a design that didn't work nearly as well as I hoped.
This version, unlike the previous one, has a very definite directional bias at higher RPM (10K+ RPM). The top of the blades (the ones on the tapered side of the spindle) are what you want to use as the leading edge. In my print, I took another user's advice and sanded the blades, then hit them with some acetone. I'm not sure how much of a difference this makes, but this fan does seem to move a good amount of air.
The blades are a bit thicker, they print out a little better because of this, and they don't seem to flex as much (printed with ABS) as my previous version. There is still a bit of flex when you hit higher RPMs, which might actually be beneficial as the blades could be seeking a more optimal angle of attack. In this design it doesn't seem to affect the blade radius as much when they flex as it did in the previous design.
I've gone down to four blades, much wider, with a slightly shallower angle. From what I've been reading, this increases efficiency, though I'm still experimenting to see how good I can get these fan blades to be for my purposes (Expect more versions coming out!).
The pipe fittings I'd promised in my other version are designed, I'll have those up as soon as they are printed and tested. The design is for the motors I've been using (Radio Shack Hi-Speed 9-18VDC motor Part # 2730256).
** Obligatory Warning **
Rotating something at high speed can be dangerous, always be sure to take proper safety precautions such as eye protection, etc to protect against possible failures. Also, take care not to touch the blades during operation. This design has some sharp parts that can cut you (speaking from experience... ouch).