Why Can't I get my DTI to Repeat

[petertha]

May or may not be your problem but when bearings are going, the readings will repeat every other rotation. This is because the bearings rotate around the spindle half as fast as the spindle rotates.

Some real good detective work in this post. I would never have known to even think of this issue.

I'm trying to wrap my brain around the 2nd rotation reading thing. Are you saying for example a cluster of rolling elements are worn, therefore a bit more annular gap between rings. So if that cluster was phased at the 3-o'clock position you could measure runout in that position. But now at 1 full spindle rotation the bearings would only have migrated to 9-o'clock. Wouldn't this 2:1 relationship be a function of the rolling element diameter kind of like idler gears? ie smaller diameter rolling element take longer to migrate than larger ones?

 

[RJSakowski]

Some real good detective work in this post. I would never have known to even think of this issue.

I'm trying to wrap my brain around the 2nd rotation reading thing. Are you saying for example a cluster of rolling elements are worn, therefore a bit more annular gap between rings. So if that cluster was phased at the 3-o'clock position you could measure runout in that position. But now at 1 full spindle rotation the bearings would only have migrated to 9-o'clock. Wouldn't this 2:1 relationship be a function of the rolling element diameter kind of like idler gears? ie smaller diameter rolling element take longer to migrate than larger ones?

Good catch, Peter!

The balls do rotate around the race at a slower rate than the spindle rotates but I was in error stating that it was half the spindle rpm. I was offered that explanation when I first noticed the every other repeating phenomenon myself and I passed the explanation on without properly analyzing. My bad.

Actually, the rate at which the balls move in the race is dependent upon the ratio of the contact diameters of the inner and outer race. In order to have the 1:2 ratio, the bearing balls would have to have a diameter of one half of the contact diameter of the inner race. This is actually fairly close to the configuration used for bearings with a spacer to keep the balls separated.

I checked a bearing with seven balls separated by a spacer and actually came up very close to a 1:3 ratio which is in line with what the OP observed.

 

[hman]

Last night I was going to respond to the question of how fast the balls circulate, but it was WAY past time for me to hit the sack. Anyway, my concept of how to understand the circulation rate is to treat the races and balls as similar to an epicyclic (planetary) gear train. Instead of counting gear teeth, you can get a good approximation of the behavior by substituting the diameters of the balls, inner race OD and outer race ID. There might be a bit of inaccuracy because of slippage. But for a correctly pre-loaded bearing, this slippage should be very small or zero.

Anyway, it looks like RJ got there first.

Epicyclic gearing - Wikipedia

en.wikipedia.org

Screen shot of a key section of the article: