Time-Dependent Global Reliability Sensitivity Analysis of Mechanisms with Correlated Random Input Variables  
Author Pengfei Wei
Co-Author(s) Wenxuan Wang
Abstract Time-dependent global reliability sensitivity (GRS) analysis is an effective technique for quantifying the different types of effects of the uncertainty of each independent random input variable on the time-dependent reliability of motion mechanism. However, when the random variables are correlated, the current time-dependent GRS indices are not affordable for distinguishing their different types of effects. To overcome this difficulty, we firstly study the connotations of the main and total effect indices of correlated variables, and then propose three other time-dependent GRS indices, i.e., the individual uncorrelated effect index, the total correlated effect index and the individual interaction effect index, for deeply studying the different types of effects of the correlated random variables on the time-dependent reliability. At last, the envelope function method combined with Monte Carlo simulation (MCS) is extended for efficiently estimating all the five timedependent GRS indices. The significance of the proposed method in identifying the failure sources of motion mechanism as well as the effectiveness of the numerical method for estimating the proposed GRS indices are demonstrated by a four-bar mechanism and a rack-and-pinion steering linkage.
Keywords Motion mechanism, Global reliability sensitivity, Correlated random variables, Envelope function
    Article #:  23-039
Proceedings of the 23rd ISSAT International Conference on Reliability and Quality in Design
August 3-5, 2017 - Chicago, Illinois, U.S.A.