Directional Importance Sampling Methods for Reliability Evalaution of Structures with Multimode Failure

摘要

A large variety of sampling / simulation methods for the reliability evaluation of structures have been proposed (in the last two decades). The need for efficient sampling & simulation mehtods emerges from the necesity to reduce the computational effort, which rapidly grows with the size and complexity of the system under analysis and/or with the analytical complexity of the model(s) employed involved by high accuracy requirements. All of the efficient sampling techniques aim to reduce the computational effort in problems of reliability estimation, withouth inducing significant errors of the failure probability / reliability estimations. In this paper we present an attempt to a higher integration of the directional and improtance-based sampling methods. We propose a model for directional-importance simulation in the polar coordinate space.