Fast Algorithm for Calculating Chemical Kinetics in Turbulent Reacting Flow

Abstract

The design of a fast, automatic solver for homogeneous, gas-phase chemical kinetic equations requires understanding the physical and numerical sources of computational inefficiency. Several inefficiency sources are reviewed and specific techniques for improvement are recommended (and some advised against). It is argued that exponentially-fitted integration algorithms are inherently more accurate for chemical kinetics modeling than classical, polynomial-interpolant methods for the same computational work.

new algorithm incorporating an exponentially-fitted trapezoidal method is presented. The two- part predictor-corrector algorithm includes filtering of ill-posed initial conditions, automatic stepsize selection, and automatic selection of Jacobi-Newton or Newton-Raphson iteration to achieve maximum computational efficiency while satisfying a prescribed error tolerance. The new algorithm compares favorably with the code LSODE on two representative test problems from combustion kinetics.