Transcripción del resumen del autor: This study deals with reactive gas particle flows like the coke combustion during the regeneration of FCC particles. In this kind of reactive flow, the global reaction rate is usually bad predicted. In a first approximation, the chemical scheme can be the reason because of the limitation of its modeling. It is usually based on macroscopic experimental results. The link between these macroscopic measurements and a local kinetics of the heterogeneous reaction occuring at the gas -particle interface is not confirmed. Results of kinetics coming from experimental measurements are used and we try to highlight the problems that appear when the same kinetics are used at different scales. In common industrial computations, coarse meshes are used to solve continuity equations. Averaged or filtered Navier-Stokes and species continuity equations have to be solved in which additional correlation terms appear because of non-linear terms in the original equations, including reaction rate correlation. Therefore a multiscale analysis is performed in order to improve the modeling of this terms. This paper, shows that the eulerian formulation of kinetics has to be improved due to the impact of the particle volume fraction on the reaction rate and the necessity to develop a subgrid model for the reaction rate due to the natural clustering that appears in gas -particle flows and non-linear additional terms appearing in filtered equations.

6