Reliability evaluation for rolling bearings with small sampling based on bayes method

As to these bearings, whose outside diameter is large (d ges 120), price is high, and life is long, the number of samples chosen is small in order to save sample quantity and test cost, not according to the standard. There are a lot of users to choose the bearings less than 8 pcs to test the life for comparing the different bearings quality. According to the trait of rolling bearing submitting to Weibull distribution and test data, this paper adopts no information prior distribution based on Bayes method, and presents point estimation of shape parameter and reliability, interval estimation of shape parameter, reliability, reliability life and mean life. This paper provides an available method for quantitative assessment of bearing reliability in terms of small sampling.