Fault Diagnosis of Gearbox Based on Order Domain Averaging

Abstract

Against the non stationary random process of gearbox's velocity variation, an analysis method for the vibration signals at shift process in the rotate mechanism is explored. The vibration signals and the speed signals at start-up in the gearbox are sampled in time-domain using multi-analyzer and then resampled the speed signals in angle-domain using spline interpolation arithmetic, so the equal angle distributed data of the speed signals were obtained and then the vibration signals were resampled too. The data of the vibration signals were analyzed using order tracking analysis in the order domain, the order domain averaging was applied into the fault diagnosis of gearbox. The outcome shows that there is more superiority in dealing with the instantaneous signal with order tracking analysis than traditional frequency spectrum analysis. It can avoid effectively the "frequency smear" phenomenon, which cannot be solved with the traditional frequency spectrum method. It can be reduce the demodulation of the vibration signals with the order domain averaging, and the method has the identify ability to the forepart fault in the gearbox and the result proves that the order domain averaging analysis is trenchancy supplement for traditional spectrum analysis.