The Derivation of Interpolants for Nonlinear Two-Point Boundary Value Problems

There now exist several powerful codes for the numerical integration of rst order systems of nonlinear two-point boundary value problems. All of these codes aim to provide a continuous solution to the problem being solved but the way in which each code achieves this aim is quite dieren t. In this paper we survey some of the approaches that are at present being used to derive interpolants and we provide a new interpolant for use with TWPBVPL.f which is a highly stable deferred correction code based on Lobatto formulae. c 2006 European Society of Computational Methods in Sciences and Engineering