A multiphysics optimization approach to design low noise and light weight electric powertrain noise, vibration and harshness (NVH) prediction of electric powertrain using finite element analysis (FEA) and optimization

This paper presents a multiphysics modeling approach to design an electric powertrain assembly in order to predict its NVH performance and perform optimization to minimize noise while keeping weight under control. The powertrain assembly contains two components - synchronous motor and a geartrain. Numerical simulation approach involves three steps; 1. A two-dimensional (2D) finite element electromagnetic analysis (FEA) of motor to predict harmonic excitations in frequency domain on stator and rotor; 2. A three-dimensional (3D) structural harmonic analysis of the powertrain assembly to predict the structure's dynamic response3. Harmonic acoustic analysis to model noise radiation due to vibration of assembly by transferring surface excitation velocities to 3D domain of air around the assembly. The excitation spectrum for the gearbox are taken from [1]. An average A-weighted sound pressure level (SPLA) sound is predicted at 1.5-meter distance from assembly. Comparison of results between motor alone and full powertrain assembly is carried out to show the effect of coupling of motor with geartrain. Finally, optimization is done for identified geometry parameters to minimize SPLA while keeping the weight in acceptable range.