Global reliability sensitivity analysis of motion mechanisms

Abstract

The kinematic failure of a mechanism is commonly caused by the random input errors such as the errors of component dimensions and motion inputs. For identifying the main source of the failure probability, the global reliability sensitivity analysis is introduced. The method is based on decomposing the variance of the failure domain indicator function into partial variances of increasing orders, which enable attributing the failure probability to each of the random input errors and their interactions. Then the analytical solutions of the global reliability sensitivity indices are derived based on the first-order Taylor series expansion of the motion error function. Compared with the traditional local reliability sensitivity indices, the global reliability sensitivity analysis technique has two main merits. First, it is more suitable for ranking the importance of random input errors and identifying the source of failure probability. Second, it provides not only the individual effect of each random input error on the failure probability but also their interaction and total effects. The engineering significance of the global reliability sensitivity indices as well as the effectiveness of the analytical method for computing the global reliability sensitivity indices are demonstrated with a four-bar sine function generator mechanism and a rack-and-pinion steering linkage.