My first simulation of the motor showed a 1 ton force attracting the rotor disc to the stator disc, whether the motor was on or off. This is caused by the permanent magnets “wanting” to connect to the nice flux-conducting electrical steel cores, and has several undesirable effects, one of them being the probable self-destruction of the motor.
There’s been a few suggestions on how to improve the axial flux design in order to compensate the 1 ton static axial force.
NJay came up with a more solid mechanical design to withstand the forces without bending the materials. He also made this nice fly-by 3D movie (SketchUp, what else) to show us. 🙂 It starts with the original simple design, and then goes on to the reinforced version.
The thing is, I myself had also been thinking about structure reinforcements for some time. Here’s some of the sketches I made (sorry for their being in paper, but my wife has been insisting I spend too much time on the computer… They’re also in Portuguese, because I didn’t think I’d be showing them to anyone… I won’t make that mistake again!):
Here I decided to reinforce both discs in 2 different ways:
– with an outer cylinder that embraces the whole set – it stops the discs from bending and can also serve as protection against penetration of foreign objects (or exit of broken pieces, if it comes to that…);
– with triangular and rectangular plates that stop the disc plates from bending and also possibly reduce the effects of stress.
In the next one you can see the triangle plates’ positioning, behind each coil.
About this time, and just like NJay did, I concluded it was a losing battle to insist in the axial flux design as it was, because even if the structure was solid enough, the axial force would make it impossible to assemble and disassemble the motor by hand, and, most importantly, to fine-tune the air gap distance. Rui suggested we used a specialized tool (like the common bearing extractors), but I think “two wrongs don’t make a right”. It also doesn’t take care of the problem of overloading the wheel bearings with that extra axial force (both static and stress components).
So I sketched the radial flux version, which has the great advantage of removing the enormous stator-rotor 1 ton force from the axis of the motor and wheel.
Here you can see my preoccupation with the size of the magnetic set. I don’t know if I can fit all the thickness of materials (magnets, coils, and electrical steel cores for flux conduction) in the vertical direction. Mainly because I still haven’t consolidated the choice of dimensions for the coils. Another worry in this radial design is the air gap itself. It no longer will be adjustable upon assembly to the car. Depending on how well the motor is built, this may actually be an advantage. Build it once, use it anywhere! 😉
But another thing is true: as you can see in the previous post (radial flux alternative design), the air gap will have to be a lot bigger (because there must be clearance space for the edges of the magnets and coils to pass by each other), and the efficiency of the motor will suffer with that. We could design a “normal” motor, with nice round rotor and stator surfaces and a tight air gap, but that is almost impossible to build at home. And we would need exactly the right kind of curved magnets for the job, which would also be very hard (or impossible) to find.
I definitely need to run some magnetic simulations on this new radial model!…