FAULT DETECTION OF GEARS WITH DIFFERENT ROOT CRACK SIZE USING WAVELET

摘要

Although tremendous effort has been applied to develop reliable strategies for detecting tooth cracks of gearboxes, these methods have generally fallen short of the required performance. Cracks are usually initiated at the root of a tooth and are very difficult to be identified from time-domain measurement. The vibration signal transformed by wavelet is sensitive to energy change. In this study, the transient vibration variations induced by different sizes of cracks at the tooth root are captured using wavelet. Firstly, as the main parametric excitation, the time-varying gear meshing stiffness caused by the alternating of engaged gear teeth is accurately calculated based on energy method, in which comprehensive deformations including Hertz contact, axial compression, bending, shearing and fillet-foundation deflection are taken into consideration. Moreover, a sophisticated dynamic theoretical model is used to simulate a practical gear system. Unique vibration signatures are captured through the comparison of cracked and perfect gear system.