Study on multi-body dynamics based suspension-rig simulation for early stage road noise prediction and bushing stiffness optimization

In NVH performance refinement, road noise plays important role than before due to matured powertrain NVH and emergence of hybrid, electrical and hydrogen driven vehicles. While vibration and acoustic properties of suspension and vehicle upper body are both crucial for road noise performance, suspension is designed prior to vehicle upper body in usual vehicle development process. Therefore test and numerical (simulation) method for early stage road noise evaluation is required for suspension module without influence of upper body contribution. In this paper, multi-body dynamics based suspension-rig simulation method is newly developed to predict road noise performance in early stage. Conventional multi-body dynamics simulation has shown frequency coverage up to 50 or 100Hz, which limits its usage on vehicle NVH analysis. The present method utilizes precise numerical model for non-linear components, such as bushing/insulators, and realized dynamic properties of suspension vibration to enhance frequency coverage up to 300Hz. The simulation results of the present method show 95% agreement compared to suspension-rig test in frequency range up to 300Hz. With the presence of road noise prediction method, bushing stiffness optimization process has been established by combining multivariable optimization technique. The present optimization process consists of design sensitivity analysis (DSA), design of experiment (DOE) and meta-model construction stages. In DOE stage, optimal Latin-Hypercube Design (OLHD) is used for minimum sample generation by maximizing distance and entropy among sample points. For meta-model construction, radial basis function (RBF) based regression method is used, which can effectively represent non-linearity of objective function. The present optimization process is applied to FF type mid-sized sedan, and 1.5dB overall road noise reduction is achieved without sacrificing R&H performance.