Detection of Faults in Gearboxes Through Vibration Signals Time-Frequency Analysis 2000-01-3176
Faults in gear teeth produce vibration components that can be recognized using adequate signals analysis techniques. The techniques based in the spectral analysis work better, when exists a repetition frequency of the occurrence of defect which is well known. However, in the specific case of identification of faults in gearboxes in assembly lines, the sampling of signals is usually carried out with the system operating in acceleration or start up of the drive shaft. Therefore, it is necessary to study the components of these signals, using tools that are able to extract the transient features caused by faults and by variation of the rotation frequency. In this paper, the measured vibration signals in automotive gearboxes have been studied through time-frequency tools, particularly, the Continuous Wavelet Transform (CWT), Pseudo Wigner Distribution (PWD) and Choi-Williams Distribution (CWD). The results prove that such techniques, can be used like convenient tools in the quality control of these vehicular components.