Laminar flame speed and stretch sensitivity of hydrocarbon fuels at high preheat, pressure and vitiation

This thesis investigates the laminar flame speed of C₁-C₃ alkanes and their binary mixtures at conditions of interest in natural gas based gas turbines viz. high temperature, pressure and dilution. Laminar flame speed has been found useful not only for validating chemical kinetics mechanisms but also for developing empirical scaling laws for practical combustion systems. The thesis addresses the lack of laminar flame speed data of C₁-C₃ alkanes at preheat (300-650 K), pressure (1-10 atm) and significant oxidizer dilution (15-21 vol% O₂). Over 400 measurements are reported over a wide range of conditions along with comparison to predictions from leading chemical mechanisms. Unstretched flame speed measurements were performed using a modified Bunsen flame technique based on reaction zone area from chemiluminescence imaging, whereas the strain sensitivity measurements were performed using a bluff-body stabilized stagnation flame with high resolution PIV. These measurements are used to: (i) discern the uncertainties associated with the measurements, (ii) understand the effect of fuel mixture and vitiation on flame speed, and (iii) validate the performance of the leading chemical kinetics mechanisms. Extensive testing shows the unstretched flame speed measurements from the modified Bunsen technique are reasonably accurate. Vitiation studies for methane and propane flames at high preheat show the reduction in flame speed results primarily from the thermal effect of the diluent and that the relative change in flame speed from the undiluted mixture is well correlated to the fractional change in the adiabatic flame temperature over a range of conditions. Significant difference in the measured and predicted flame speeds were observed for rich, atmospheric pressure, propane and lean, high pressure, methane/ethane mixtures with dilution. This highlights possible avenues for improvements in the chemical kinetics mechanisms. Systematic errors were also identified in the Bunsen flame measurements at certain conditions, such as for rich flames with dilution, indicating a need for better understanding of the Bunsen flame technique at these conditions. The difference in the measured and predicted flame speed does not show any clear correlation with the flame height or the strain sensitivity of the mixture. Finally previously proposed mixing rules for estimating flame speed of fuel mixtures from pure fuel components are shown to be reasonably accurate over a range of pressure, reactant temperature and dilution conditions.Ph.D