Detection of Fault in a Bevel Gearbox Under Varying Speed Conditions

Abstract

A bevel gearbox is complicated in geometry and founds application where the shafts are perpendicular such as helicopter and automobiles. Unlike spur gears, three forces exert on bevel gears to transmit the motion. Due to complicated geometry of gears, nonlinear and nonstationary vibrations excited due to meshing exhibiting severe modulations in the gearbox vibration signals. Delay in detection of faults could be fatal; therefore, the condition monitoring and early fault detection techniques are of vital significance. In the present manuscript, a novel signal processing-based approach has been presented to detect the gear faults under varying speed conditions. The vibration signals acquired from the test bench were decomposed using variational mode decomposition (VMD) and the sensitive sub-band is selected upon estimating the instantaneous frequency (IF). The selected mode function (MF) upon decomposition is analyzed using the statistical condition indicators and its FFT is developed. The FFT of the selected mode exhibits the symptoms of the gear faults.