Gearbox Condition Estimation Using Cyclo-Stationary Properties of Vibration Signal


Article Preview

The paper explores the cyclo-stationary properties of vibration signals for estimation of gearbox condition. The advantage of such approach may be clearly seen especially for so called multi-faults problem, i.e. for more than one faults that occurred in the system. In complex mechanical systems like multistage gearboxes, such situation may be often seen. Although this approach becomes more and more popular, it has been noticed that there is difficult to find examples highlighting its potential, especially for real industrial situations. In order to fill partially the gap, the paper deals with the multi fault detection in complex mechanical systems like multi-stage gearboxes: fixed axis and planetary. It has been discussed that during the operation in such machines many faults may appear simultaneously and the classical method like envelope analysis is difficult to use. The paper presents the use of cyclo-stationary properties of signals to identify and characterize sources of modulation. From Spectral Correlation Density Map or more precisely Spectral Coherence Map have been observed the number of sources with different properties of modulation. It is shown that the number of harmonics is important for a kind of fault extraction and interpretation. This approach has been applied to two, three and five stage gearboxes used in mining industry. Vibration signals received in industrial environment during normal operation of objects are considered. It has been also proposed the simple diagnostic feature to estimate the changes of condition with application to a planetary stage in a 5-stage gearbox.



Key Engineering Materials (Volumes 413-414)

Edited by:

F. Chu, H. Ouyang, V. Silberschmidt, L. Garibaldi, C.Surace, W.M. Ostachowicz and D. Jiang






R. Zimroz and W. Bartelmus, "Gearbox Condition Estimation Using Cyclo-Stationary Properties of Vibration Signal", Key Engineering Materials, Vols. 413-414, pp. 471-478, 2009

Online since:

June 2009




In order to see related information, you need to Login.

In order to see related information, you need to Login.