Stochastic finite element analysis of local average random fields

This paper discussed the concepts underlying a new approach to representation and analysis of homogeneous random Fields. An orthogonalized isoparametric local average model of random fields is proposed, and the optimum selection of random variables is determined based on the theory of nonlinear programming. The finite element method is used to develop a methodology for random field analysis. Numerical examples showing the accuracy and efficiency of the method are studied.