tested: OpenSCAD 2021.1
shifted the vertical axis of the bracket on that occasion as proposed in one of the makes.

Background and first design

Studying some logarithmic spirals I stumbeld over the Archimedes windmill patented and produced by a Rotterdam company called the Archimedes. Because I liked the idea behind it, I started to implement a first version to see it running, then a second version much closer to the original, because I also liked the design itself.

Obviously the company's windmill has blades made from metal sheets rolled around an axis and having no specifiic aerodynamic design. Although they claim an amazingly high efficiency for such a small wind turbine as well as a silent run, I suspected that a proper airfoil design might not be wrong - and also enhance the harmony and statics of the design.

So I started to implement a libary for NACA airfoil data and a sweeper to be able to seemlessly juggle around airfoils with changing parameters in 3D space on arbitrary trajectories.

Putting all this together the outcome turned into a quite interesting design. However, I didn't do any calculations or specific modelling to optimize the shape for a specific wind speed.

One of the design goals was to get a single STL for printing without any support. For this reason I added a foot to stabilize the print. It is meant to be cut off after printing to enhance aerodynamics. As shaft I used a 2mm bike spoke. I drilled up the hole in the turbine (1.9-2mm drill bit) and heated the spoke before poking it in. The holes in the bracket I drilled up with 2.5mm. The bracket can easily be expanded by 10mm to insert the turbine.