Research on Gear Bending Fatigue Test and Probabilistic Life Prediction Method

The manufacture of precision long life gears is characterized by high cost and long cycle. Moreover, the fatigue tests of gears are time consuming and labor intensive. In order to obtain the life information of gears accurately and conveniently, a model for predicting the probabilistic life of gears is established by using the concept of the minimum order statistic. Furthermore, the probabilistic life relationship among gears with different numbers of teeth is constructed. The life data of specific gears are taken as the input variables of the model, and through the transformation of the model, the probabilistic life of other gears can be obtained. Meanwhile, a gear bending fatigue test is carried out under a constant stress level, and during the test, the methods for monitoring and evaluating gear failures are expounded. Next, two analysis methods of gear failure mechanism and fracture are combined to present the failure mode, and test data are used to illustrate the guiding significance of the model. Finally, the model is verified by the statistical method of random censored data, which proves that the model has good prediction ability and the advantage of handling a small amount of data.