AN ALGORITHM FOR THE NUMERICAL EVALUATION OF THE RANDLES-SEVCIK FUNCTION

An accurate numerical method for the evaluation of the Randles-Sevcik function Chi(x) is given. The method is an improvement over the trapez oidal quadrature approach of Lether and Wenston. We obtain significant computational economy by adding correction terms to the quadrature su m, which compensates for the effect of the singularities of the integr and. For intermediate values of x, this method is shown to be more sui ted than other series methods. Practical error bounds are given for th e discretization and truncation errors. Copyright (C) 1997 Elsevier Sc ience Ltd.