Master of Science in Mechanical Engineering (MSME)
Preheating a combustible mixture enhances the laminar burning flux characteristic mad with high reaction firing rates. As a result, the flammable zone as defined by inlet conditions of equivalence ratio and temperature is expanded beyond that available at standard ambient conditions; however, fundamental questions in these combustion regimes have not been addressed. In this thesis, preheated lean and rich combustion of methane/air mixtures is studied numerically and experimentally to catalog and confirm expected trends in these regimes. Numerical simulations were completed using both GRI-Mech 3.0 and San Diego mechanisms in the combustion code Cantera. An adiabatic simulation data set is obtained over a vast range of equivalence ratios (φ = 0.15-3.5) and inlet temperatures Tin = 200-1000 K), while further study is completed at lean (φ < 0.89) and rich conditions (φ > 1.3). Detailed analyses of flame structure and reaction pathway analysis, sensitivity, and heat release are completed at a total of ten reference cases, five lean and five rich, selected along contours of constant equivalence ratio φ = 0.7, 1.6 and mass flux mad = 0.2190 kg/m2-s. A regression analysis of each regime links adiabatic flame propagation to a characteristic temperature T★, shown to be primarily a function of m, while φ and Tin are shown to play a subordinate role. Analyses together reveal causal kinetic phenomena contributing to differences in lean and rich combustion. Experiments connect the adiabatic findings to the simplest non-adiabatic application, where stand-off distances of a flat flame burner are used as a metric for flame behavior. Viable flames are established at ultra-lean and rich conditions, but results show mechanism uncertainty at preheated conditions in addition to unmodeled heat transfer phenomena. Further study of flat flame behavior is performed in the computational fluid dynamics code Fluent 12.0, where a two dimensional axisymmetric flame is stabilized for three mass fluxes at a reference case of φ = 0.7, Tin = 300 K. The model does not attempt to replicate the exact conditions seen experimentally, rather it seeks to evaluate boundary effects and other two dimensional flame structures resulting from exceeding the laminar burning flux.
Document Availability at the Time of Submission
Release the entire work immediately for access worldwide.
Gibson, Joseph Edward, "Numerical and experimental evaluation of preheated premixed flames at lean and rich conditions" (2012). LSU Master's Theses. 2168.