Asymmetric Gear Blank Optimization to Reduce Gear Noise for Electric Drive Units

Gear whine has emerged as a significant challenge for electric vehicles (EVs) in the absence of engine masking noise. The demand from customers for premium EVs with high speed and high torque density introduces additional NVH risks. Conventional gear design strategies to reduce the pitch-line velocity and increase contact ratio may impact EV torque capacitor or its efficiency. Furthermore, microgeometry optimization has limited design space to reduce gear noise over a wide range of torque loads. This paper presents a comprehensive investigation into the optimization of transfer gear blanks in a single-speed two-stage FDW electric drive unit (EDU) with the objective of reducing both mass and noise. A detailed multi-body dynamics (MBD) model is constructed for the entire EDU system using a finite-element-based time-domain solver. This investigation focuses on the analysis and optimization of asymmetric gear blank design features with three-slot patterns. A DOE methodology is employed to identify pivotal gear blank design parameters, including the blank thickness and slot angle. The radiated sound power and mount vibration responses from the EDU are predicted and correlated with test data. The time-varying stiffness at the meshing point gives rise to sidebands around the transfer gear orders, which are accurately captured using the MBD time-domain solver. The asymmetric gear blank stiffness changes the torsional vibration transfer path, necessitating microgeometry re-optimization to fully capture the NVH benefits. A case study is conducted based on the Ultium EDU, focusing on transfer gear blank design. It is demonstrated that a selected three-slotted gear blank design, in conjunction with optimized microgeometry, results in a 0.46 kg reduction in mass and reduced gear noise at both high and low torque conditions.