One of the more interesting cases I worked on was when I took ultrasound readings on the
bearings of a slow moving (about 9 RPM), large (100 ton plus) crusher.
The reading showed an interesting graph. (See graph 1) There were distinct increases in harmonics, which did not correlate with the RPM speed.
Back at the office, Ian Ware and I used the UE Systems Spectralyzer software to analyse the recordings and came to the following conclusion that there was an outer race defect bearing. How did we get to that conclusion?:
We put the dual cursors built into the software, on the start of the 3rd and 4th events and that calculated a frequency of 0.141Hz = 8.46RPM which we assumed is the running speed at the time of the recording..
We then zoomed in on the 3rd event and put the dual cursors on the 2nd and 3rd event and got a calculated frequency of 1.96Hz See graph 2
So taking the FAG data if you look at the BPFI (inner race fault frequency) = 13.8773 x 0.141 (running speed) = 1.96Hz
If it were an outer race defect the spacing would have needed to be 1.56Hz and the reading was way off that.
Whilst not an exact science as far as cursor positioning and resolution goes, it would be a relatively educated guess that it is an inner race defect in the bearing. The conclusion was that due to it being an inner race defect, the friction only showed when the defect in the bearing was at the bottom of the cycle and carried the full weight of the crusher.
If you have seen similar examples please let us know .