Moment-independent sensitivity analysis of a fold mechanism based on an active learning kriging method

Abstract

The present article shows the reliability and sensitivity of a missile folding rudder calculated by MSC.ADAMS and MATLAB. For folding rudder, mass of the counterweight, maximum of the motive force, preloading force, and stiffness coefficients of the spring 1 and spring 2 are considered when calculating the reliability. Two methods, active learning kriging (ALK) and Monte Carlo simulation (MCS), are introduced to calculate the reliability. The results obtained by ALK and MCS are compared to ensure the accuracy of the reliability. Three failure modes are considered respectively in reliability calculation, including the unfolding time exceeding the specified value, the max load of the lock structure 1 or lock structure 2 exceeding the specified value. Moment-independent sensitivity index is introduced to assess the importance rank of each factor. Both cases with and without gap are discussed to explore the influence of gaps between sleeve and shaft. Comparison of two cases show that gaps could reduce reliability of the folding rudder. The influence of gaps cannot be neglected during the reliability calculated, especially for mechanism with several motion joints like the rudder folding. The works aim to provide valuable guidance for design and optimization of unfolding rudder. In addition, it is a beneficial trial for introducing reliability theory in weapon equipment field.