احسان بهمیاری

A novel computational scheme called the Chaotic Radial Basis Function is presented in this paper for quantification of uncertainties in bending analysis of moderately thick plates with elastically restrained edges under lateral loading. The plate modulus of elasticity, the stiffness of the elastically restrained edges, and the lateral loading on the plate are considered as random processes and are represented by using the Karhunen–Loève expansion which is based on a linear combination of the eigenfunctions of the covariance function with uncorrelated random variables. These random variables which could have any types of probability measures constitute the random space of the stochastic problem. The proposed method relies on discretizing the random space by a set of properly distributed nodes and then employing a Galerkin-Chaotic Radial Basis Function scheme to derive the coupled deterministic set of equations governing the stochastic plate bending. The accuracy of the proposed method is investigated by comparing the results obtained by the method with those of the Monte Carlo method. It is seen that the presented method provides accurate results with a considerably less computational cost in comparison to the Monte Carlo simulation. Further, the effect of coefficients of variations of the plate module of elasticity, applied loading and the stiffness of the restrained edges as well as the influence of the correlation lengths, aspect ratio and the plate thickness are investigated on the statistics of the plate deflection.