Prediction of NVH Performance for Electric Vehicle Drivetrain Based on Multiphysics Simulation

摘要

Prediction of noise and vibration performance at the early design stage is highly important for the design of electric vehicle powertrain. The major noise sources of E-powertrain are an electric motor and gears. Therefore, the high-fidelity, multiphysics model combining the electric motor and gear train is required to accurately predict NVH performance of the E-povvertrain system. Ansys Motion, the third-generation multi-body dynamics (MBD) solver, can accurately model powertrain system by considering both macro- and micro-geometry of gears and the complete assembly of powertrain system including the motor, gears, shafts and housing. The electromagnetic forces on both a stator and rotor calculated from Maxwell are transferred to the MBD drivetrain model. Considering electromagnetic forces that drive E-povvertrain system and the characteristics of gears, noise/vibration generated from housing can be predicted in both time and frequency domains. The developed multiphysics workflow can accurately track the order of noise/vibration related to gear mesh frequency, which is a source of gear whine noise, and motor frequencies related to torque ripple and stator forces. Such a complete workflow enables both designers and analysts, especially in automotive industry, to expedite design iterations and to evaluate electric powertrain NVH performance in the early design stage.