Monitoring Natural Pitting Progress on Helical Gear Mesh Using Acoustic Emission and Vibration

Abstract

Over the last few decades, there has been an awareness of the role of predictive maintenance on large industrial gearboxes. To apply effective condition base maintenance on gears, it is important to diagnose the gear fault at its earliest stage of development. The application of the acoustic emission (AE) technology in relation to gear box diagnosis is well documented [1–4]. Asperity contact has been found to be the major source of AE activity within the gear mesh [5]. Researchers have investigated the influence of other operational factors such as oil film thickness [6], load and speed [7, 8] on the level of AE activity. Tan et al. [9] presented the first and only conclusive research in which the applicability of AE in monitoring the natural progressive pitting was investigated. Tan employed a back-to-back gearbox to study the accelerated surface pitting on spur gears while a comparative study between effectiveness of spectrometric oil analysis, AE and vibration techniques was undertaken. Their research showed that the AE r.m.s from the pinion gear was more sensitive in detecting the pitting progress on the gear surface. Raja et al. [7] performed a comparative study on the effect load and speed on the generation of AE for both helical and spur gears, while Eftekharnejad et al. [10] investigated the capability of AE technology in locating and identifying seeded surface defects on helical gears. To date, apart from the work by Eftekharnejd et al. and Raja et al. [7, 10], there has been no other investigation into application of AE in diagnosis of the helical gears. The use of vibration analysis for gear fault diagnosis and monitoring has been widely investigated and its application in industry is well established [11]. This is mainly reflected in the aviation industry for diagnosis of the helicopter engine, drive trains and rotor systems [12]. The spectral kurtosis (SK) is gaining ground in application to vibration diagnosis. To determine the SK, the signal is first decomposed into the time-frequency domain after which the Kurtosis values are determined for each frequency band [13]. The concept of SK analysis was first developed by Dwyer [14] as a tool which was able to trace non-Gaussian features in different frequency bands using the fourth order moment of the real part of the short-Time Fourier transform (STFT). Dwyer only investigated the application of SK on stationary processes but did not account for non-stationary vibration signatures typical of rotating machines. To date, the most comprehensive calculations of the SK has been developed by Antoni [15] as the fourth order cumulant of the spectral moment (K):