Health monitoring for gearbox systems based on damage dynamics

Health monitoring of gearboxes can significantly improve the ability of Condition-Based Maintenance (CBM) of the systems. One of the motivations of traditional health monitoring is to support fault diagnosis, but the monitoring results can´t provide available information for CBM. Hence, this paper proposes an approach to select optimal monitoring parameters and the corresponding condition indicators for a gearbox system based on damage dynamics. With the analytical formulations of tooth crack, the dependencies among crack levels and the reduction of mesh stiffness are able to obtained, then the damage dynamics of the gearbox system can be built. Afterwards, the dynamic responses of generalized parameters in the models of various damage levels can be simulated, and their condition indicators can be obtained by using effective signal processing methods. Then, the sensitivities of each generalized parameter to tooth crack can be calculated, and the optimal monitoring parameters can be determined according to the size of sensitivities.