This is meant to be my entry for the "Fan Show Down" hosted by the YouTube channel Major Hardware. It is a fairly simple looking fan design with an A12x25 hub mount. The twist in this design is that I used a genetic algorithm to optimize the blade cross section, chord, and pitch along the length of the blade. I am not entirely sure if the assumptions I made for the algorithm's fitness function were all correct, but the end result looks like a fan should :P. I have no doubt that there are more creative designs that could perform better as my search was somewhat limited. Regardless, I found this to be an interesting and enjoyable project.

The blade design for this fan was optimized using a genetic algorithm along with MIT's XFoil software.

For each section, the rotation speed, radius, and guessed pitch angle was used to estimate the Reynold's number at that blade section. With this Reynold's number and a guessed 4 digit NACA airfoil, and alpha sweep was run in XFoil. I assumed that the air around the blade section was in steady state when the blade no longer produced any lift (or no circulation around the blade section). Admittedly, this is my weakest assumption and is probably not entirely correct.

From the alpha sweep results, the algorithm computed the zero lift angle of attack. Using this, the guessed pitch angle, and the rotation velocity, the estimated air intake velocity could be calculated. This was used to perform a second XFoil run at the zero lift AOA to find the pressure distribution along the blade section which would allow for the trailing edge velocity to be calculated. The training edge velocity was then transformed out of the blade section reference frame into a non-rotating reference frame to find the outflow velocity at that blade section.

The fitness of each blade section design was based on the magnitude of the outflow velocity parallel to the intake direction, and the drag force experience by the blade section. The aim was to maximize outflow velocity while minimizing drag. This would allow the design to pull more air while also allowing the fan motor to spin faster for a given amount of power.

The genetic algorithm optimization was performed at 3 sections along the blade. The results of my optimization are included with the cad files.