Intervening variables and constraint approximations in safety index and failure probability calculations

The emphasis of this paper is on developing suitable intervening variables and constraint approximations for structural reliability analysis. Traditionally, these procedures are used in structural optimization, whereas this research work adopts these concepts to safety index and failure probability computations. The use of these concepts enables the development of an efficient and stable iteration algorithm for identifying the most probable failure points (MPPs) of the limit state functions. An approximate second-order failure probability is calculated at this MPP with no extra computations of the limit state function and gradients. The efficiency and accuracy of the proposed algorithm are demonstrated by several examples with highly nonlinear, complex, explicit/implicit performance functions.