Towards a reliable non-linear dynamic model of damaged gear transmission

Most diagnostic algorithms of mechanical systems are based on empirical observations derived from actual measurements of vibration patterns. In this study, a different approach was examined. This approach is based on the assumption that reliable dynamic models can be utilised to predict the effects of faults on vibration patterns. Among the benefits of the proposed approach are an increased understanding of various phenomena as expressed in vibration patterns, enhanced quality of the diagnostic algorithms and possibly reduced effort and cost related to extensive seeded tests. In the study, a simple spur gear transmission system is modelled and analysed. In the first stage, a non-linear dynamic model of a spur gear transmission was developed. In the second stage, an analytical method was proposed to estimate the gear mesh stiffness as a function of mesh angle in the case of both healthy and faulty gears. The dynamic response of the system was computed by applying the gear mesh stiffness derived from the analytical model. The model was applied to simulate the theoretical vibration response in healthy and spalled tooth gearwheels. The dynamic responses derived from the model were compared to and verified with experimental vibration data. The experimental phase included varying the load and speed settings, allowing the examination of the fault effect on the vibration signature under different test conditions.