Dulbecco A. , Lafossas F. -A. , Mauviot G. , Poinsot T.J.
Oil & Gas Science and Technology – Rev. IFP, 2009. Vol. 64, No. 3, pp. 259-284Abstract
This paper presents a new 0D phenomenological approach to the numerical modelling of Diesel HCCI combustion. The model is obtained through the reduction of TKI-PDF (Tabulated Kinetics for Ignition, coupled with presumed Probability Density Function) 3D CFD model developed at the IFP. Its formulation is based on physical considerations, to take into account the main phenomena and their mutual interactions that take place in the cylinder during the combustion process. Aspects relating to spray penetration, fuel evaporation, turbulence, mixture formation and chemical kinetics have been studied in detail. The original contribution of this work concerns the modelling of the formation and evolution of the equivalence ratio stratification around the spray, and of its connection to combustion kinetics. In order to achieve this, different tools commonly adopted in 3D modelling have been adapted to 0D modelling. Presumed PDF theory has been extended to a 0D formalism in order to characterize the mixture-fraction distribution. This approach has then been coupled with droplet-evaporation theory in order to have access to the thermodynamic conditions characterizing the mixture. The temporal evolution of the spray is computed in terms of volume and the entrained mass of gases, starting from conservation laws for mass, momentum and energy. An adapted − model is used to take into account the turbulence in the cylinder, which is very important, in an ICE (Internal Combustion Engine), especially during the mixing process. Further, combustion heat-release is computed using an adapted detailed tabulated chemistry method inspired by the FPI (Flame Prolongation of ILDM (Intrinsic Low Dimensional Manifold)) theory. This look-up table allows the simulation of a large range of combustion regimes, since it takes into account the presence of EGR (Exhaust Gas Recirculation) in the mixture. The results of the 0D model are compared in an initial step to the 3D CFD results. Finally, the OD model is validated against a wide experimental database.
