Probabilistic Response Analysis of Geometrically Nonlinear Structures Based on Systematic Response Surface Method

CHENG Jin

DOI: https://doi.org/10.3321/j.issn:0253-374x.2006.09.003

2006-01-01

Abstract:A stochastic finite-element-based algorithm for the probabilistic response analysis of geometrically nonlinear structures is proposed in this paper by combining the advantages of the response surface method,finite element method and Monte Carlo simulation.The accuracy and efficiency of the method are demonstrated through a numerical example of geometrically nonlinear truss structure with system parametric uncertainties.The choice of response surface models is a key element for the proposed method.Particular attention is given to investigate effects of response surface models on statistics of the response of both linear and geometrically nonlinear structures under certainties.To this aim five response surface models are considered,namely,star experiment design based on a quadratic polynomial without cross terms,hyper-cube experiment design by using a quadratic polynomial without cross terms,complete factorial experiment design based on a quadratic polynomial with cross terms,minimum experiment design based on a quadratic polynomial with cross terms and composite experiment design by using a quadratic polynomial with cross terms.Results obtained by the proposed method based on the five response surface models are compared and appropriate conclusions are drawn.