The rotor is a stack of twelve 3mm thick discs to hold the magnets onto the soon to be double slotted 8mm shaft and transmit power through a pair of 3.175mm round rods fitting into round cutouts in all discs. The sharp angle cuts on the outer perimeter will be coated with lots of epoxy and a layer of dyneema thread wound through the slots/grooves prior to placing the magnets in the slots/grooves of the stack. The combination of adhesive, dyneema, and magnets on the outer perimeter should provide enough mechanical reinforcement to greatly delay the likely inevitable self-destruction of the acrylic, and the use of two full-length round rod 'keys', instead of square key stock, in the inner perimeter should also reduce stress concentrations in the acrylic. Using lots of thinner pieces bonded together should also help slow the propagation of cracks since the thinner material is more flexible and a crack in one does not mean as large a loss in material still connecting the magnets to the shaft compared to using fewer pieces of thicker material to achieve the same overall thickness.
Assembly might be a bit tricky since I did not include a jig in the order to help hold the magnets in place before the epoxy is sufficiently set to keep them from snapping onto each other and still do not have a working 3D printer yet. Basically have to get the discs reasonably well assembled onto the shaft and pins/rods then coat the exterior with epoxy and wrap with dyneema along the bottom surface of the deeper grooves and crossing around the shaft to grooves on opposite side, then insert the magnets using either a non-magnetic band or ring to help keep them from popping out of their grooves. After the magnets are placed, and if there is still room, going to wrap the exterior with another layer of dyneema thread (around the outer diameter this time to really hold the magnets well) since I have two giant spools of it and when combined with epoxy should make a nice composite structure to keep the magnets flying off into the stator (assuming the epoxy actually bonds to the very slippery UHMW-PE fiber that is dyneema thread).
As for wheel hubs, I've been working on a very much overkill design using cheap generic 0.75" timkin (tapered roller) bearings intended as replacement kits for highway use trailers. Unfortunately, my thought to use 2.25"-OD 1.5"-ID tube and 0.375" plate steel as endcaps to attach to the wheel rims and brake rotors would be really expensive plus involve quite a bit of machining, and thinner, cheaper tube would require inserts of some sort to provide the steps needed to maintain the preload of the tapered roller bearings. Could also just use thinner walled tube and add big welds to the interior to create the bearing retaining steps instead of using smaller tube as a liner, but would probably be harder to get properly balanced.
Thinking about how to make the hubs actually drove me to think about buying/building a furnace to get lengths of 1" diameter steel rod (preferably AISI 1144) up to ~1200[C] and forging them into the rough tube shape. After that, press/upset forge the tubes in closed dies to create the lips on the ends and reheat as necessary to 790[C] for a decent soak then furnace cool to a full anneal state. Finally, chuck them in a lathe to get them to their final dimensions and maybe another heat treatment although 1144 is already quite strong in its annealed state. Unlikely I would buy a furnace as they are quite expensive when new from McMaster and also quite small. If the interview on Tuesday goes well, thinking I might try making a cheap refractory lined box in the backyard to create a basic furnace/kiln that could have many different uses (brazing, heat treating, metal casting, ceramic firing, etc.). Of course, I would need to get the 20 ton press built well before the furnace construction should really be attempted.