Abstract
The parametric effect on performance of Moving Least Squares (MLS) interpolation based recovery technique is evaluated in this paper. The Moving Least Squares (MLS) fitting recovers the field variable derivatives over nodes patch using background element meshes. The recovered errors at element and global levels in the finite element solution are presented in energy norm. The study considers three basic recovery parameters (i.e. parameters affecting the post-processed results) namely shape of influence (support) domain, dilation parameter and order of poly nominal basis function. Numerical experiments on elastic plate problems are carried out for parametric effect of interpolation based post processing technique on effectivity of error estimation and rate of convergence of the recovered solution with fineness of the meshing scheme. The linear and quadratic triangular elements have been used for the discretization of the problem domain. The circular and rectangular shape domain of influence is used to form the node patch. Six different dilation parameters and three different number of basis function terms are selected in the moving least squared interpolation formulation. The study shows that recovery parameters of MLS interpolation method have pronounced effect on the post-processing recovery of finite element solution and optimal alternatives are to be adopted for better performance of the recovery procedures.
Published Version
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have