Development and Application of an Eulerian Density Function Methodology coupled to Flamelet Progress Variable Approach for the Simulation of Oxyfuel Combustion

In the prevailing situation of unsustainable fossil fuel resources and the elevated levels of air pollutant emissions, the state-of-the-art of combustion investigations confronts primarily two challenges. These are, on the one hand, the optimization of the fossil fuel combustion efficiency and, on the other hand, the development and the application of robust strategies to reduce the amount of the released pollutant gases with respect to the new emission standards in accordance with the global energy policies. Within this context, the carbon dioxide capture and storage (CCS) technologies play an important role as an accepted strategy towards the mitigation of CO2 emissions. One of the important aspects of the CCS techniques is the oxidation of natural gas under oxy-fuel combustion conditions. However, very few scientific contributions have been devoted to the research of these systems, so that there is a lack of understanding of the oxy-combustion processes. The present work aims at the development and the application of an advanced numerical approach for the simulation of oxy-fuel combustion in which the TCI is adequately accounted for within non-premixed combustion regimes using the OpenFOAM platform. The suggested model, which is designed for both RANS and LES applications, consists of a combination of a transported probability density function approach following the Eulerian Stochastic field methodology and the flamelet progress variable (FPV) chemistry reduction mechanism. In the LES framework, the proposed method accurately represents the effect of the sub-grid fluctuations on the flame structure and on combustion characteristics, along with the interaction between turbulence and chemistry. The implemented developed combustion model is first verified, and then validated and applied to different turbulent non-premixed combustion configurations featuring an in- creasing order of complexity. In particular, Sandia flame D, which consists of a turbulent piloted methane-air jet flame, is first employed for model validation in both RANS and LES contexts. The next flames are more challenging cases, namely the non-premixed Sandia oxy-flame series (A & B), which are operated under different Re numbers and characterized by various CO2 and H2 enrichments in the oxidizer and fuel streams, respectively. All investigated cases are well documented with available experimental measurements. The comparison of the obtained results with experimental data in terms of temperature, scalar distributions, PDFs, and scatter plots agree satisfactorily, essentially in the LES context. This work finally reveals that the hybrid ESF/FPV approach removes the weaknesses of the presumed probability density function-based FPV modeling (β-PDF)