Originally Posted by
Barry Ott
Both my bearing issues had a small amount of play after everything was assembled. David Clubine of Britain West told me of his experience with bad bearings out of the box and, after discussing the situation, we thought it was a likely scenario. My front assembly is a simple stub axle pressed through the inner races with a big nut on the other end. Looking at the upright bore, it looked fine, the bearing was reasonably difficult to press in to the upright, so that didn't seem to indicate the upright. The real tip off was that adding torque on the big nut didn't have any effect on how easy the upright spun around the stub axle, which told me the bearing wasn't being preloaded. After I removed a little material from one of the inner races (where it touches the other inner race), adding torque to the big nut made the upright harder to turn. CAREFUL not to get carried away!
With this test in mind, I started working on the rear. This assembly has two "hubs" that sandwich the inner races, with the outer hub being pressed almost all the way through both inner races. Note that the two hubs should NOT touch in the middle, if they do, then they can't maintain the clamping force on the inner races and the bearing will loose preload. When I went out for my shake down runs, I had a tiny bit of play in the assembly, and ~12 laps turned that into a bit more. This was because the inner hub mating surface to the bearing inner race wasn't quite right so it deformed the inner hub surface a little, which allowed the two hubs to touch, blah, blah, balh. The good part was that the inner hub to bearing interface was deformed enough that it wasn't going to move any more. So, I removed some material from one of the hub surfaces to allow the two hubs to clamp around bearing without touching. Back to the preload check, I assemble the upright, the two hubs and the bearing in the press and confirmed that the upright spun around the "axle" with more resistance as I increased the clamping force (CAREFUL!), confirming the bearing was preloaded. So, long story, hope it helps!
So, yes, if the bearing has too much interference between the two inner races, the assembly will have play. The interference I'm talking about is where the two inner races touch each other, which keeps the the surfaces where the inner and outer balls roll too far apart, so the balls can't be squeezed on to the outer race ball surfaces.
For me, testing the bearing meant assembling the whole thing, that's the only way I could feel the play at the wheel (+/- .0025 at the wheel, which is ~.00029 at the bearing inner race)