Combustion Chemistry and Parameter Estimation

Abstract

Combustion processes in practical systems are marked by an huge complexity stemming from their multi-dimensional character, from the interaction of physical processes (including diffusion, flow dynamics, thermodynamics and heat transfer), and from the extremely complex chemistry potentially involving up to hundreds of species and thousands of elementary reactions. The determination of accurate parameters accounting for the chemical kinetics is an essential step for predicting the behavior of practical combustion systems. This is usually done by carrying out specific experiments in simplified systems whereby many of the physical phenomenons mentioned above can be neglected. The present paper aims at giving an overview over the approaches currently followed to estimate kinetic parameters based on experimental data originating from these simplified systems. The nature and mathematical description of such problems are presented for homogeneous systems where all variables depend on time only. The techniques for the identification of the most significant reactions (and hence parameters) are shown along with methods for mechanism reduction considerably alleviating the computational burden. As an application example, Mechacut, a C++ procedure written by the authors is employed for the reduction of a detailed reaction mechanism aiming at describing the combustion of methane CH₄. In the following section, the estimation of kinetic parameters is formulated as an optimization problem and different approaches found in the current literature are examined.