Fault detection of a two-stage gearbox with artificially induced gear crack using vibration monitoring

Abstract

The fault diagnosis of a gear crack in a two-stage gearbox using the vibration signal processing techniques is the aim of the present work. The online vibration data were recorded for a set of normal gear and cracked gear, at a particular speed and various loading conditions on Drivetrain Diagnostics Simulator. Spectral analysis is carried out for a healthy gear and a cracked gear to demonstrate modulation sidebands from cracked and normal spur gear vibration measurements. The amplitude of sidebands can be used as an input parameter in various artificial intelligence techniques for fault detection. In this paper experimental study has been done to find relation between incipient gear crack at different loading conditions and amplitude of sidebands.