Riding Naked All The Time
Join Date: Oct 2014
Rep Power: 1
Now lets look at tolerances that were likely applied to this design. The bearings are ABEC3 (claimed from all-balls). That means that the spec should be 12mm wide +0 -0,0127. This is interesting because my calipers are only precise to that same number (0.0005"). The fact I measured 11,96 means they may be under spec if indeed made to ABEC3 spec's (should have been 11,9873 min). HOWEVER, I am sure some of you are out there screeming about splitting hairs (one RCH?)
I am however pretty confident that the upper bound of the stack will NEVER exceed the 114mm reverse engineered width.
Typical tolerance that would be expected from a manufacturing company on something like the spacer should be no more than 0.005". Given it's intention to not let the bearings come too close together you don't want the thing too short. So I suspect mine may actually be under spec, but appears to not allow the bearing on the LH side bind (you always fully seat the R.H. Bearing and only push L.H. in to center spacer)
So, Let's call that stack 114mm -0, -0.006" (0,1524)
Given I have actually measured my stack at 113.84, this is pretty darn close to being at the bottom end of the expected tolerance zone (0.003") and likely a result of the "squishing" of the spacer when the bolt is torqued the numerous times it has been in it's life.
Therefore as it stands right now I intend to make my R.H. spacer 8,75mm +0,1 -0 to cover the range of maximum material conditions to my emperically measured stackup. Alternatively a one-piece R.H. spacer would be 25,25mm +0,1 -0
The whole point of the above is to get the wheel centered by finding the virtual centerline of the wheel, then getting the R.H. stack correct. I believe I have no found some pretty good numbers to go by.
Should get wheel centered within 0,1mm to the theoretical center. Which is about 10x what is probably necessary.
2004 919 Uranium