COMPUTATIONAL TECHNIQUES FOR ADAPTIVE HP FINITE-ELEMENT METHODS

In this paper we present some new computational techniques designed fo r adaptive hp methods. We also list some established and successful te chniques. In particular we introduce a new robust solver for such meth ods using a novel ''space-filling curve'' based ordering of the unknow ns and a multilevel substructuring type factorization. We also outline an a posteriori error estimator that can be easily used in conjunctio n with this solver. The ordering introduced is of general applicabilit y and can also be used with non-adaptive methods to great advantage.