Vibration signal model for fault diagnosis of sun gear in epicyclical gearboxes

Abstract

Signal models of machinery components are helpful to understand influence and character of various fault modes without conducting costly and timeconsuming experiments; it also brings convenience to validate new fault diagnosis methods for scientist and engineers. Researchers have built models such as lumped parameter model and finite element model to study problem of interest. However, these models are time-consuming to simulate faults in epicyclical gearbox, some of them could not implant fault on specified gear tooth. In this paper, a simplified fault implantable signal model is established considering meshing regularity between given sun gear tooth and planet gear, as well as amplitude modulation effect caused by rotating of carrier. Theoretical derivations are validated through spectrum analysis of simulation signal, moreover, experimental data collected from fault-seeded and baseline experiments are analysed to validate the model, computing time of new model and lumped parameter model are compared to show computational efficiency of the model. According to signal analysis and computing time comparison, the signal model proposed in this paper is effective and efficient.