The E ect of Flow and Mixture Inhomogeneity on theDynamics of Strained FlamesbyYoussef

Changes in ow strain and mixture composition on the order of a ame time scale are characteristic of many practical combustion processes. Accurately predicting the unsteady reponse of burning to these changes requires detailed modeling of species transport and chemical kinetics. This thesis formulates a detailed one-dimensional computational model for arbitrary unsteady conditions of strain and mixture, demonstrating its applicability to subgrid modeling of turbulent combustion. A novel numerical formulation, based on a globalized Newton iterative method and a precondi-tioned Krylov subspace linear solver, ensures eecient and robust convergence despite the stiiness of detailed chemistry. The model is validated via comparison with OP-PDIF, a well-benchmarked steady-state strained ame code. The model is then used to characterize the fundamental interactions of ow and mixture inhomogeneity| examining the dynamic response of ame structure and burning to linear variations in mixture equivalence ratio, and capturing the eeect of unsteady strain on a ame surface interacting with a vortex in two dimensions. The latter example leads to a redeenition of the appropriate subgrid strain for ame embedding simulation of premixed turbulent combustion. Acknowledgments Thanks are rst due to my thesis advisor, Professor Ahmed Ghoniem. His guidance and insight have brought coherence and rigorous thinking to this work, as well as a strong awareness of its context and implications. I would also like to thank Dr. Habib Najm, who hosted me at Sandia National Laboratories in August of 1998, during the early stages of this eeort. His attention to detail and his practical know-how, shared in many discussions over the past year, helped speed the progress of my research. Also, Habib generously provided the data from simulations of ame-vortex interaction that are used in this thesis. I owe a debt of gratitude to my family, who have been constant and supportive throughout my education; my appreciation of their support continues to grow. Thanks are also due to my fellow students (and postdocs) at the Reacting Gas Dynamics Lab|Jean-Pierre Hathout, Issam Lakkis, Shankar Subramaniam, and Mah-moud Fleiil. I have relied on them for friendship, technical advice, and a healthy share of lunchtime diversion. I also thank Constantin Petrov, a laboratory alumnus, for introducing me to the elemental ame code. Finally, I would like to acknowledge the support of the Fannie and John Hertz Foundation, which has sponsored me as a graduate fellow for the past two years. Numerical simulations in this thesis were performed at Sandia …