Application of Model Order Reduction to Nonlinear Finite Element Tire Models for NVH Design 2019-01-1507

In current practice, tire development and testing are typically experimentally driven. However, as the need to simultaneously optimize multiple noise vibration and harshness (NVH) performance criteria increases and development cycles become shorter, predictive numerical simulation techniques are becoming necessary. In addition, many tire performance areas are coupled and therefore the experimental approach often lacks detailed insights which numerical simulations can provide. Currently, no industrially applicable fully predictive high-fidelity numerical approach that incorporates the use of nonlinear Finite Element (FE) tire models for NVH design is available in literature. Therefore, a fully predictive numerical simulation approach that predicts the rolling of a tire over a coarse road surface is described in this work. The proposed approach allows to predict the dynamic contact- and hub forces that arise during rolling without the need for experimental data. Based on these results the NVH performance of a specific tire design can be assessed and optimized. One of the main drawbacks of using nonlinear FE tire models for NVH design, is the large computational cost associated with running the numerical simulations. Therefore, application of a nonlinear Model Order Reduction (MOR) and hyper-reduction technique to the nonlinear FE tire models is described in this work as well. It is shown that application of the MOR and hyper-reduction techniques greatly reduces the total computational time and costs, leading to acceptable computational times for engineering practice. The simulation results show good correspondence with experimental data. This confirms the potential of this efficient predictive numerical simulation approach as a viable alternative to the experimental based approach in tire NVH design.