Reliability sensitivity analysis based on stress-strength model of bearing with random parameters.

Abstract

On the basis of the stress-strength model of the ball bearing with random parameters, the reliability sensitivity of the raceway strength is examined. The basic parameters are regarded as random variables subject to normal distribution. The Latin hypercube sampling method is adopted to obtain the samples, which are brought into the bearing model to obtain the corresponding maximum orthogonal shear stress. The explicit expression of shear stress was obtained by genetic algorithm-back propagation neural network fitting, and the limit-state equation is established in combination with the yield limit of bearing materials. First, this study analyzes the sensitivity of the maximum shear stress with respect to various parameters and obtains the effect of parameters on shear stress at various rotational speeds. Then, based on the stress-strength state equation, the strength reliability is obtained by using the improved first-order second-moment method, which is verified by Monte Carlo simulation. Finally, the reliability sensitivity with respect to the mean and standard variance of random variables is analyzed. This research can provide theoretical guidance for the design, production, and use of bearings.