Combustion modeling using Principal Component Analysis: A posteriori validation on Sandia flames D, E and F

The present work shows the first application of the PC-transport approach in the context of Large Eddy Simulation (LES) of turbulent combustion. Detailed kinetic mechanisms, together with advanced computational tools, are needed to advance our knowledge of turbulent reacting systems. However, the cost related to high-fidelity simulations of turbulent reacting flows is still prohibitive for realistic configurations. Therefore, there is a need to reduce the complexity of the problem by identifying low-dimensional manifolds. To this end, the potential offered by Principal Component Analysis (PCA) in parameterizing the thermo-chemical state-space is very appealing. The present paper extends the PC-transport framework to three-dimensional Large Eddy Simulation (LES), coupling PCA with Gaussian Process Regression (GPR). To demonstrate the potential of the method, LES simulations of Sandia flames D, E and F are shown. Results show the great potential of the PC-GPR model, as indicated by the accuracy of the simulation results when compared with experimental data, using only 2 principal components. The sensitivity to the kinetic mechanism and subgrid closure model is also investigated.