Practical Solutions to Machinery and Maintenance Vibration Problems
Chapter 8, Vibration in Bearings
Section 10, Vibration Symptoms of a Loose Bearing
A bearing can be installed in such a way that "hand feel" indicates a proper fit. However, when the machine is running, the bearing fit might not be tight enough for operating conditions. While running, the fit is loose and the race that is not supposed to rotate, rotates. This, of course, causes excessive rubbing and most often produces symptoms similar to that of a rub or looseness.
If no bearing defects are present and the only problem is bearing looseness, the usual symptoms are more harmonics of 1 x rpm than usual. For example, unbalance alone produces almost no larger than usual harmonics, such as a 2, 3 or 4 x rpm. Shaft-to-shaft misalignment usually increases the amplitudes of the harmonics in this low end of the harmonic range (see section "Evaluating Harmonics for Complete Analysis to Determine Vibration Source"). Harmonic amplitudes caused by misalignment usually diminish considerably after 3 or 4 x rpm (unless magnified by resonance or produced by the segments of coupling jaws). However, higher than usual amplitudes of harmonics due to bearing looseness can occur in the same low end of the harmonic range (but most often do not show until at least 3 or 4 x rpm) and continue on to still higher harmonic frequencies. These harmonic frequencies are usually still in the range that Update calls "lower harmonics" (approximately 10 x rpm and below).
Although more harmonics of higher than usual amplitude are present, they are not absolutely sequential. For example, the harmonics created by looseness at 2 and 3 x rpm may be small and appear higher than usual at 4, 5 and 6 x rpm. For greater looseness or harder rubs, the harmonics can continue, for example, on up to say 8 x rpm; but occasionally intermediate harmonics may be missing, such as the 4th and 7th harmonics. For extreme cases, these harmonics can continue on up to very high frequencies. Synchronous time averaging will not eliminate the true harmonics of 1 x rpm. As this condition is not an actual bearing defect, such as a spall, pit or crack, the harmonics created are not considered to be "bearing defect frequencies." True bearing defect frequencies, such as spalls, cracks and pits, emit frequencies that are non-synchronous with running speed. So when there is question as to whether a peak is due to a bearing defect frequency or due to some other source, it's a good idea to determine whether it is synchronous. Spectra showing rubs or looseness always vary from machine to machine. Only a general idea can be formed by pattern similarity.
1. IBF units are an early warning system only.
2. IBF units should be trended over time where change is more significant than absolute values.
3. IBF units are sensitive to outside influences such as cavitation, transducer mounting, lubrication and so on.
4. When observing velocity spectra, pattern recognition rather than absolute amplitude values are most important.
5. All bearing defects (other than too much clearance or looseness) create peaks at frequencies non-synchronous with running speed.
6. As defects progress, look for increased sideband activity.
7. Difference frequencies between sidebands usually indicate the rpm of the defective bearing.
8. Don't attempt to be too specific in predicting time to failure.
This textbook contains only part of the information in our Practical Solutions seminar.