Automatic derivation of the optimum dynamic model for analysis and decision making in gearboxes

摘要

Eliminating problems in dynamic behaviour through simulation requires suitable application of model creation, mathematical solution and result interpretation. The creation of lumped-parameter models from a continuous system is necessary to carry out such dynamic simulations, yet it is a time-consuming and iterative process, often prone to errors. Different models with different approaches to discretisation give different simulation results even when derived from the same geometry. A discretisation algorithm has been implemented to automate this process and to create optimum dynamic models for a given frequency range. This gives simulation results that are accurate and at the same time computationally efficient. In this document, we validate this discretisation algorithm using various examples and show the errors that can arise from inappropriate discretisation. It goes on to show how an informed selection of model and solver facilitated decision making and problem elimination in a real-life case study for gear rattle.