An Accurate Analysis Method to Calculate Planetary Gear Set Load Sharing under Non-Torque Load 2022-01-0653

Given their high-power density, large range of speed change, and reputation of being quieter than counter-shaft gear sets, planetary gear sets (PGS) have advantages to be applied in electric vehicle (EV) applications. Since electric drive unit (EDU) designs are often subject to accelerated development timelines with more versatile gear set layouts than conventional automotive transmissions, accurate prediction of PGS load sharing is needed. In the past, PGS load sharing imbalance used to be considered as a gear set problem focusing only on the effect to gear performance. Finding a closed-form formula has been a focus in gear design. However, early bearing failure in wind turbine gearboxes exposed the limitation of this strategy. With extensive field and laboratory testing, engineers started to notice that load sharing imbalance is essentially a system issue. Non-torque loads on PGS should be considered in the estimation by a gearbox system model. In this study, a virtual design, development, and validation (VDDV) method is proposed to calculate PGS load sharing accurately by using a validated deformable finite element contact analysis tool. The study concentrates on the effect and interaction of the non-torque load and planet carrier compliance that hasn’t been investigated before. To understand the PGS load sharing problem thoroughly, we define bearing load sharing for the first time, in conjunction with conventional planet gear load sharing. The study also represents PGS load sharing as a periodic variable instead of a constant factor, which supplies a mathematical tool for data interpretation. Through a parametric study of four EDU models, some design tips are clarified, such as increasing load or reducing carrier stiffness to improve load sharing. The work here confirms the complexity of load sharing problem which needs a system level numerical analysis. The proposed VDDV method can help hardware reduction in future EDU development. It also provides a validation foundation for more efficient analysis tools.