Abstract
This paper is concerned with developing an inner-outer iterative method for steady state problems by setting up an inner iterative procedure (specifically the AGE-1D algorithm, chosen due to its suitability for parallel computers) within the outer iterative procedures that are usually applied (e.g. the ADI method). The experiments are carried out on the Laplace equation in two dimensions. The study involves examining what termination criterion should be applied for the inner procedures. It examines the influence of grid refinement on this criterion, and on the choice of the acceleration parameters of the inner/outer iteration procedures and their effect on the rate of convergence.