A new correlation between diluent fraction and laminar burning velocities: Example of CH4, NH3, and CH4 + NH3 flames diluted by N2

Modern combustion processes widely use exhaust gas recirculation, oxyfuel combustion, and other techniques that alter the concentration of diluent gases from that of the air. The dilution's impact on the laminar burning velocity, S-L, is therefore a crucial effect that has been studied experimentally and numerically in the literature. However, an accurate fitting correlation with physical meanings is lacking, making it difficult to extrapolate S(L )data to real-application conditions. To address this gap, in the present work we have derived a novel correlation between diluent fraction and laminar burning velocity, S-L, through new analysis of the maximum temperature gradient and heat release rate as ln(S-L/S-0 ) = a( 1/Yu - 1/Y-0 ), where Yu is the reactant mass fraction in the total unburnt mixture, and a being a constant when only the diluent concentration is varied. To provide data for the analysis and validation, SL of CH4 + O-2 + N-2, 40 %CH4 + 60 %NH3 + O-2 + N-2, and NH3 + O-2 + N-2 flames were measured using the heat flux method at varied oxygen ratio x(O2) and equivalence ratio phi. Simulations were carried out using three kinetic mechanisms from Han, Konnov, and Okafor, which have been validated using both CH4 and NH(3 )burning velocities. Our experimental and simulation results for various fuel types and equivalence ratio conditions demonstrate good linearity with R-2 > 0.985 over all ranges of x(O2) spans, from the upper limit of x(O2) = 1.0 to the lower limit where SL is below 5 cm/s, confirming the accuracy of the correlation equation. This correlation is also found valid under complex conditions with various unburned temperatures, pressures, diluent types, and fuel types, indicating its wide applicability. Sensitivity analyses revealed the kinetic origin of the linear ln(S-L) vs. 1/Y-u relationships. Specifically, the absolute values of a sensitivities are much smaller than those of S-L, and they remain nearly the same for different oxygen ratios. Therefore, even by tuning the rate constants of the highly sensitive reactions, the a at different 1/Yu conditions will change uniformly, resulting in a linear ln(S-L) vs. 1/Y-u variation though with a different slope value.