An Improved Non-Probabilistic Monte Carlo Simulation Method for Structural Reliability Analysis Based on Convex Models

This paper proposes an improved nonprobabilistic Monte Carlo simulation method for structural reliability analysis based on convex models. The novelty of the proposed method is to use a unit hyper-sphere sampling technique to ensure the validity of all samples, and to restrict the sampling domain to the evaluated domain which is geometrically the intersection of a hyper-spherical shell and a hyper-spherical sector and generally has a smaller volume than the whole ellipsoid. The nonprobabilistic characteristics and application to the interval model of the proposed method are also discussed. Seven numerical examples are provided to demonstrate the efficacy of the proposed methodology.