Estimation and Control Error Based on P-Convergence,

Abstract

The relative error in energy can be estimated on the basis of the asmptotic estimate ot thte rate of p-convergence. Nearly optimal control of error can be achieved when feedback information, generated from p-convergence data, is used for deciding whether the mesh should be refined or the polynomial degree of elements increased. Partial Contents: Convergence in energy norm; Problems with smooth solutions; Problems with stress singularities; and Estimates of error in energy norm.