« Ecriture et validation d’une méthode de réduction de la cinétique chimique de type Computational Singular Perturbation : Application à la combustion numérique du biogaz ».

Abstract

The main objective of this thesis is to write a code to reduce the kinetics che mical mechanism by Computational Singular Perturbation method and to validate the reduced mechanism by the numerical combustion of biogas. This reduction approach is based on the analysis of the importance indices of the reaction groups on the evo lution of species and the degree of participation of elementary reactions. To validate our code, the Yang and Pope’s mechanism has been reduced to 22 reversible reactions.This mechanism has been validated in three steps :Firstly, we used the chemkin II package, with DVODE solver for solving systems of differential equations deriving from reduced and detailed mechanisms. good agreement of the numerical temperature and average mass fractions of the majority species were observed between the two mechanisms. In a second step, two numerical simulations of a turbulent diffusion DLR-Flame-A are made in CFD OpenFoam code. Partially Stirred Reactor (PaSR) and the ‘’Modèle In termittente Lagrangien” (MIL) are used as combustion models to take into account the interaction between chemistry and turbulence. The choice of MIL model is motivated by the aim of validating the ignition delays table of reduced mechanism. Good agree ment is observed between the numerical and experimental data following the radial and axial profiles. The experimental data are obtained from the Sandia laboratory website. Finally, a simulation of the numerical combustion of biogas is made. These biogases it is from anaerobic digestion of household garbage, plant biomass and pig slurry. In order to optimize the heat during the combustion of biogas, we have proposed four numerical tests of the combustion of dihydrogen-doped biogas. The results obtained are satisfactory, the heat released during the combustion of biogas is proportional to the content of methane and added hydrogen. In view of the results of the applications of the reduced mechanism by the Computatio nal Singular Perturbation method, we claim that this technique is suitable for reducing mechanisms that may help to study turbulent diffusion flames, the rate of ignition, the ignition time on engines using natural gas or biogas.