Institute of Engineering Mechanics, University of Innsbruck, Technikerstr. 13, A-6020, Innsbruck, Austria
Hochschule für Architektur und Bauwesen - Universität, Marienstr. 15, D-99421, Weimar, Germany
Abstract:
This paper presents a procedure which allows for a stochastic finite element (SFE)-based reliability analysis of large nonlinear structures under dynamic loading involving both structural and loading randomness with relatively little computational effort when compared to traditional Monte Carlo methods. The analysis is based on the identification of important random variables by means of a transformation of the vector of original random variables to the uncorrelated space and subsequent sensitivity analyses. Only few nonlinear computations using the most important identified random variables are performed to determine points on the limit state surface. Subsequently, the response surface method (RSM) is employed to estimate the reliability of the structure.