The strength analysis and probabilistic design of a bogie frame with incomplete probabilistic information

Abstract

A bogie frame is a key component of high-speed rail; higher requirements of the safety and stability of the bogie frame are put forward with the continuous improvement of the train speed. In this study, we first performed a strength analysis of a bogie frame according to the JIS E 4207 standard by using finite element method. Then, we did a reliability analysis of the bogie frame. The reliability problems are defined as strength reliability and resonant frequency reliability, which indicate the structural safety and stability of the bogie frame. The reliability estimation is realized by experimental design, artificial neural network and stochastic perturbation theory. Additionally, reliability- based sensitivity indices were derived to measure the parameter importance of random input variables. An illustrative example of the bogie frame with incomplete probabilistic information was used to demonstrate the applications of the proposed method for reliability and reliability-based sensitivity estimation in terms of strength and resonant frequency reliability. The results indicate that vertical loads and the material density have more of an impact on reliability than other variables of the bogie frame.