I extracted the bearings to measure the internal dimensions of the bearing shoulders.
The axle shoulders are 38.15mm (+/-0.05) wide, while the hub bore hole bearing surface is around 37.82mm (+/-0.1). I have to measure the hole indirectly because the hub body tapers in the center, so this measurement isn't precise. But at least the difference between the body and and the axle is around 0.33mm (+/-0.15).
(I later measured this more precisely to around 0.40 to 0.55mm)
Is it better to have this difference in width as close to zero as possible? How is this normally designed by other manufacturers?
Or is it better to have the hub slightly narrower, or the axle slightly narrower?
I assume axle slightly narrower by less than 0.05mm is better, so that outer bearing trace will contact the hub sides, and the inner trace will add a little bit of preload when clamped into the dropout.
( I later measured the axial play of the traces at about 0.02mm, so I could potentially add a preload of about 0.04mm)
With the way it is now with this novatec hub, the inner trace will be clamped tightly against the axle shoulders, but there is a gap between the bearings and the inside surfaces of the hub. Wouldn't the bearings slip around?
I'm thinking of adding shim washers to fill the gap. I can find some din 988 19x26mm in 0.1, 0.2, 0.3mm thicknesses.
The lathe that i have access to has a lot of backlash so I'm not confident that I'll be able to cut the shoulder narrower.
I created another post comparing the exterior dimensions of other 74mm hubs for folding bikes here:
Novatec and Kinetix 74mm front hub dimensions for Dahon and others
Also, one side of the axle has an interference fit with one bearing, while the other side of the axle has an "easy" fit that slights in without much interference. The bearings are TTN chinese branded, and strangely the widths are 8.05mm instead of 8.0mm.
During extraction, a lot of metal shavings came out, and I'm not sure whether it came from the hub surface, axle, or residual waste from machining.
(shavings came from threads on the bolt I used for extraction) The inner surface was dry without grease, and very dirty.
