Algorithm-Based Optimization of Gear Mesh Efficiency in Stepped Planetary Gear Stages for Electric Vehicles
The electrification of the automotive powertrain confronts the gearbox development with new challenges. High-speed concepts require higher gear ratios, which cannot be optimally achieved with simple cylindrical gear stages. For this reason, stepped planetary gear stages are increasingly used, as they offer high power density at high gear ratios. To increase range and energy efficiency, the gear mesh losses are of great importance and must be considered in the gear design.
The design of the macro geometry of gears is usually focused on ensuring the load-carrying capacity. In the design of stepped planetary gear stages, there are constraints due to assembly restrictions as well as additional degrees of freedom, such as the division of the total gear ratio. Due to many adjustable geometry parameters and design combinations, manual optimization of the gear geometry would not be effective.
In this paper, a method for an automated optimization of the macro geometry of stepped planetary gear stages to improve the gear mesh efficiency is presented, which considers the assembly restrictions. An FE based tooth contact analysis is used to evaluate the design objectives: NVH (Noise, Vibration, Harshness), load-carrying capacity, and efficiency. Since these objectives require different design strategies, a weighting of the objectives is necessary. A particle-swarm algorithm is used to optimize the gear geometry and the tool data. Tooth flank pressure, peak-to-peak transmission error, tooth root stress, and efficiency are evaluated. The influence of the weighting of the design objectives on the gear design is shown. The results of various optimizations are compared, and an efficiency-optimized variant is selected for a specific application.
With the method presented in this paper, it is possible to design the macro geometry of stepped planetary gear stages using FE-based tooth contact analysis and to optimize the operational behavior for a given application.
Authors: Christian Westphal, Jens Brimmers, Christian Brecher
Discounted member price:
You could save: