Design of profile corrections and tolerances in cylindrical gears for noise-sensitive applications in line with production efforts

Abstract In many applications of cylindrical gearboxes, the reduction of noise from the gear mesh is an important criterion. In addition to electrified vehicle drives, gear excitations also play an important role in stationary applications such as wind turbines. When designing gearboxes, elastic surrounding structures or material combinations are increasingly being used in order to reduce mass and manufacturing costs. This has a decisive effect on the misalignment behavior of the cylindrical gears and the need for micro geometric corrections. At the same time, it must be taken into account that deviations can occur during the production of the micro geometries, which must be tolerated and have different effects on the NVH behavior. The aim is to develop a method for the design of profile corrections for elastic environmental conditions and materials. It takes into account that the nominal design is subject to different manufacturing deviations, which have to be tolerated. This is done by means of a tolerance design mechanism that determines the best compromise between tolerance definition and manufacturing costs depending on assumptions for the relationship between manufacturing accuracy and resulting costs. In the first step of the nominal design, the load-dependent misalignments are determined by FE-simulation. This is followed by an optimization of the micro geometric gear profile using a penetration calculation in order to optimize the excitations and improve the NVH behavior. An AI-supported substitute model of the excitation behavior of the entire gearbox is then determined. This is used to determine a suitable compromise between tolerance requirements and expected production costs using a metaheuristic optimization process. This is demonstrated using a cylindrical gearbox for electric cars. The calculation method presented enables the design of profile modifications and production tolerances using modern tools and shows further potential for noise reduction.