Basic Analysis of Structural Safety

TLDR

The study reviews recent advances in structural reliability analysis, proposes new interpretations and useful quantities, and suggests dimensionality‑reduction techniques for limit‑state probability calculations. It reviews first‑order second‑moment methods, applies a Lagrange‑multiplier formulation to compute safety indices and design points, employs the Stokes and Gauss divergence theorem to reduce integration dimensionality, and recommends Monte Carlo simulation and maximum‑likelihood points as practical alternatives. The authors demonstrate that the Lagrange‑multiplier approach accurately evaluates safety indices and design points, endorse Monte Carlo techniques for estimating limit‑state probabilities, and propose using the maximum‑likelihood point as an alternative to the design point.

Abstract

The present study reviews some of the more significant recent developments in the area of structural reliability analysis, proposes new interpretations for and emphasis in some of these crucial theoretical developments, and introduces additional useful quantities. To be specific, the first‐order second‐moment methods are reviewed. It is then shown that the Lagrange multiplier formulation (and thus any algorithm associated with it) can be used to evaluate the safety index and the location of the design point. Monte Carlo techniques are recommended for use in estimating limit state probabilities as a practical alternative to other methods. The use of the Stokes and Gauss divergence theorem in two‐ and three‐dimensional integral expressions of a limit state probability is suggested in order to reduce the dimensionality of the integrations by one. Finally, it is recommended that the point of maximum likelihood be used as an alternative to the design point based on the advanced firstorder second‐moment method.