Axial flux motors are difficult and expensive to make.
Motors need to be made of laminated steel sheets to reduce parasitic eddy currents. The laminations need to be thin in the direction of the direction of the flux. For radial flux motors you just punch out a shape and stack a bunch of sheets up. For axial flux you have to wind a strip: https://15658757.s21i.faiusr.com/2/ABUIABACGAAgmviFqAYozvPw-...
Each layer of that strip has a different cut in it, so its much more complicated to make. The shape and manufacturing method typically impacts efficiency; YASA avoids that by spending more money. Efficiency is an unavoidable requirement of high power density- heat is the limiting factor, and going from 98% to 96% efficient means double the heat.
The mechanical demands on the motor are also much higher- radial flux is balanced since the magnetic force pulls the rotor from opposite sides. Axial flux motors are usually one-sided, so the magnets are trying to pull the rotor and stator together with incredible force. That also makes vibrations worse. Extremely strong, expensive bearings are required to handle it. With permanent magnet rotors you need a jig to lower the rotor into place; they can't be assembled by hand. That also makes maintenance more difficult and expensive.
>> Each layer of that strip has a different cut in it, so its much more complicated to make.
You can roll a spool of that material and then machine the shape out of it. I've seen this done for axial flux motors. There are other approaches as well, and the cost differences get even smaller if you throw automation at the production process. I used to believe axial flux motors were one of those oddities that won't win in the end, but now that I work with them I'm not so sure. They are at least competitive with radial flux machines.
Can you (or anyone) explain where the power density improvement comes from with axial flux? When I work through the first-order math it seems like it should come out the same as radial flux. Is it just that the geometry allows better cooling from the coils? Or the use of grain oriented steel?
I was wondering where the seam is supposed to be in that picture and then it turns there is no seam. It's just one continuous roll.
https://www.youtube.com/watch?v=6ZVQWgk6tRc
Isn't the whole premise of this breakthrough that they don't need the lamination, and thus can stamp it out?