A FLAMELET APPROACH FOR LARGE EDDY SIMULATIONS OF COUPLED COMBUSTION AND RADIATION IN TURBULENT BUOYANT DIFFUSION FLAMES

In traditional computational fluid dynamics (CFD) descriptions of fires, the combustion and radiation models generally rely on a global combustion equation and the assumption of a linear relationship between radiative power and heat release rate. These models may lead to a crude treatment of important phenomena such as flame extinction, formation of soot and toxic species, and the change of radiant emissions in response to evolving fire conditions. The general objective of this Ph.D. study is to evaluate the potential of advanced combustion and radiation models for large eddy simulations (LES) of fires. A flamelet-based modeling framework is proposed that considers established or modified steady and unsteady flamelet formulations. This study is part of an international collaborative project between the University of Maryland and the University of Poitiers (France) aimed at providing a fundamental understanding of coupled combustion-radiation phenomena in fires. It consists of two parts. The objective of the first part is to bring fundamental information on the coupling between combustion and thermal radiation occurring in laminar flames. The study considers a simplified configuration corresponding to one-dimensional counterflow planar laminar diffusion flames subjected to time-evolving moderate-to-slow mixing conditions that are representative of fires. The analysis demonstrates that for conditions far from the extinction limits, the flame belongs to the semi-unsteady regime in which mixing processes occurring in the outer diffusive layers of the flame are unsteady whereas heat release processes occurring in the inner reactive layer remain quasi-steady. The objective of the second part is to develop and validate a fully coupled flow-flame-radiation fire modeling framework. A novel unsteady flamelet model is developed that includes: detailed information on combustion chemistry through a tabulated chemistry approach; a careful description of the combustion-radiation coupling; a description of subgrid-scale turbulence-radiation interactions; and a description of non-grey radiation effects through a Weighted-Sum-of-Grey-Gases (WSGG) model. This new combustion/radiation model is then incorporated into the LES solver FireFOAM (developed by FM Global) and is evaluated by comparisons with experimental data obtained in a turbulent line burner experiment previously studied at the University of Maryland. Results on the global radiant fraction (GRF) obtained in cases with nitrogen dilution suggest that provided that the WSGG radiation model is used, the new modeling framework is capable of simulating changes in the flame radiative emissions with the predicted GRF within 20% of the measured values. Comparisons between the grey and WSGG options in the flamelet model show that, with the WSGG model, the simulated flame is no longer optically thin (the ratio of global absorption divided by global emission is close to 40%). Note that while the flamelet combustion model presented in this study has provided unique insights into the micro physics of fires, it is not a modeling approach that is recommended for engineering-level simulations of fires. First, the flamelet combustion modeling approach assumes the availability of a detailed chemical kinetic mechanism to describe fuel oxidation and this type of mechanism is typically not available for practical fuels in fire problems. Second, the flamelet combustion modeling approach treats the heat release rate implicitly and numerical tests show that the implicit heat release rate is described with limited accuracy (the error in the simulated global heat release rate ranges takes values between a few percent up and 20% in the present work). This limited accuracy on the description of the fire power is viewed as a strong limitation of current tabulated chemistry approaches for engineering-level simulations of fires