Kinetic Combustion Neural Modelling Integrated into Computational Fluid Dynamics

A methodology has been developed to replace traditional chemical kinetics model calculation with Neural Models and applied to air-methane combustion. The reacting flow field has been described taking a detailed chemistry reaction mechanism into account. Convective and turbulent diffusive transport of species have been taken into consideration by means of a well known finite volume CFD code. Two version of such a mechanism have been developed. The first one is based on traditional differential equations representing the specie production rates. Such equations are integrated over the time intervals related to the cell volumes and local volumetric flows.A reduced combustion mechanism involving twenty species and sixty-eight reactions has been developed both for traditional calculation as well as for the Neural Model ones. At the real end combustion flow fields are calculated with kinetics Neural Models with a CPU time forty-two times smaller than that of traditional procedures with a comparable solution accuracy of the combustion flow fields.