ABSTRACT
A short description of the commonly used condition monitoring methods for rotating machinery is given. Some of these methods, such as cepstrum analysis, synchronized time domain signal analysis, frequency analysis and various oil analysis methods are discussed in greater detail. The measurement results from two laboratory case studies performed at the Technical Research Centre of Finland are presented. The first case study was carried out with a one-step gearbox and the measurements consisted of overall vibration level measurements, synchronized time domain signal analysis, spectrum analysis, cepstrum analysis, acoustic emission measurements, automatic wear particle counting, wear particle analysis and spectrometric oil analysis. Cepstrum analysis was a very efficient tool in monitoring the condition of the gearbox. The synchronized time domain signal analysis was also found to be a promising method. This method makes it possible to detect even the faults of an individual tooth of a gear. The second case study carried out with a rolling bearing was aimed at testing a new on-line wear particle sensor, and for comparison also other monitoring techniques were used. The on-line sensor worked very well and failure in the bearing could be predicted well in advance.
