A number of factors impact the service life of bearings like manufacturing tolerances, alignment errors, overloading, or overheating. Bearing damage can be caused by for example fatigue cracks, lack of lubricants, breaking, spalling and cracking of the rolling elements. Whenever a component comes into contact with the damaged surface, the result is periodic impulses, also called shock impulses. These can excite resonance frequencies that can be measured and interpreted.
There are two measurement principles that are applicable and complimentary to bearing fault detection and monitoring:
- Scalar measurements - These describe the overall bearing condition. When trended, they facilitate that maintenance can be planned ahead of time. Typical condition monitoring scalar measurement techniques include the BCU (Bearing Condition Unit) or the BCUp (Bearing Condition Unit Peak).
- Spectral analyses - These involve narrow-band measurements that identify the type and location of damaged parts. Typical spectral analysis techniques include BCS (Bearing Condition Signature) and SED (Selective Envelope Detection). Both techniques are based on the fault amplitude modulation of a carrier frequency, i.e. envelope analysis.