NASA/EPSCoR KY: Discrete Dynamical Systems Subgrid-Scale Models for Turbulent Combustion

Modeling interactions of turbulent flow fields with the chemical kinetics of combustion is one of the most difficult problems of physical/engineering importance. We are only beginning to develop a reasonable level of understanding of either turbulence or combustion separately, and studies of their interactions are intrinsically beyond the capabilities of models based on Reynolds-averaged Navier-Stokes formulations. Only direct numerical simulation and large-eddy simulation (LES) with high-fidelity subgrid-scale (SGS) models have any prospect of succeeding in the presence of such complex physics, and neither is suitable for practical engineering calculations on any foreseeablE. co;nk~ti~'6 h3-idwarE,. The priEl3.ry goal of the research being proposed here is to continue development of a new class of SGS models for turbulent reacting flows based on methods utilizing discrete dynamical systems that have already been successfully implemented for nonreacting and non-premixed reacting flows by the PI. This modeling approach, which employs the usual decomposition of dependent variables found in LES, possesses three desirable features that make it a candidate for successfully treating turbulent reacting flows. First, governing equations are not filtered-rather, solutions are filtered, thus obviating the need to model SGS stresses and fluxes; second, the SGS quantities that are modeled are actual physical variables instead of statistical correlations, thus opening the way to detailed interactions of turbulence with other phenomena on the small scales. Finally, the SGS results are directly used to enhance accuracy of the under-resolved large-scale calculations. Because of the fidelity of such models, computations can be performed on grids much coarser than those typically employed for LES. Researchers of the Combustion Branch at NASA Glenn Research Center are actively engaged in both experimental and numerical studies of turbulent reacting flows that will be applicable to the design and analysis of propulsion systems to be employed in NASA scientific missions. The proposed study is intended to complement these efforts, and it will be coordinated through regular contact with the NASA Glenn Combustion Branch. In addition, the turbulence models produced will be made available to combustion engineers at General Electric Aircraft Engines, Evendale, OR who have previously supported turbulence research by the PIon a project involving non-reacting flows. At least indirectly, this will benefit the Commonwealth of Kentucky since the gas turbines of many GE engines are assembled in Madisonville, KY.