Research on interior noise optimization of minibus based on vibration transfer path analysis using super element method

Abstract

In this paper, aiming at the problem of excessive low-frequency noise in a minibus, the SEM (Super Element Method) was applied to the NVH (Noise, Vibration and Harshness) dynamics simulation analysis of the vehicle, and the vibration and noise control schemes were proposed from the aspect of vibration transfer path to improve NVH performance. First, the relationship and difference between SEM and LFEM (Legacy Finite Element Method) were elaborated theoretically. Afterward, the legacy finite element model of the full vehicle, which comprehensively considered all related components such as BIW (Body In White), chassis, door covers, body accessories and interior and exterior trims, was established based on LFEM. Furthermore, the subsystems of legacy finite element model of the vehicle were divided into super elements based on SEM, and the differences between SEM and LFEM in model processing were discussed in detail. In addition, the modal analysis and frequency response analysis of the full vehicle were carried out by LFEM and SEM respectively, and the computational results were compared with the testing results to verify the simulation accuracy and feasibility of the SEM in solving the vehicle body dynamics problems. Meanwhile, through the comparison of simulation analysis process of SEM and LFEM in practical case, the advantages of SEM compared with LFEM were discussed in terms of computational efficiency and storage space occupied. Finally, the stiffness optimization schemes of rubber bushing in the middle of driveshaft were proposed to improve vibration and noise from the aspect of vibration transfer path, and the testing results verified the effectiveness of the optimization schemes in reducing the low frequency noise inside the vehicle. The study concluded that SEM had obvious advantages over LFEM in terms of computational efficiency and storage space occupied, while ensuring the computational accuracy basically consistent with that of LFEM.