Numerical Prediction Modeling for Fatigue Experiments on Straight Face Gears Produced via Hot Rolling with Insufficient Information

Due to their characteristics of a high power-to-weight ratio, stringent lightweight requirements, and harsh working environments, straight face gears are prone to issues such as tooth fracture and inadequate fatigue strength. Meanwhile, because of the lack of fatigue information and weak fatigue life prediction method, the fatigue life of face gears cannot be effectively evaluated. In this study, the key technologies involved in the hot rolling forming process, fatigue experiments, and numerical modeling of straight face gears are studied. A technical foundation for straight face gears formed by hot rolling processing is established, and a fatigue experiment of the hot rolling forming of straight face gears is carried out. Due to the lack of information on fatigue experiments, a numerical prediction model is constructed. Sample expansion is carried out using a BP neural network–Bootstrap model to calculate the reliable lifespan of hot-rolled straight face gears, and fatigue life prediction for hot-rolled straight face gears is completed via the improved GM(1,1,λ) model based on the artificial bee colony algorithm, and thus the accurate evaluation of the fatigue life of rolling forming face gears is realized. The feasibility and superiority of the improved fatigue life prediction model are demonstrated by comparing it with the traditional prediction model and experimental results. The theoretical basis and technical support for the research of the fatigue resistance and installation application of face gears are provided.