Assessment of uncertainties of laminar flame speed of premixed flames as determined using a Bunsen burner at varying pressures

Laminar methane/air premixed flames at different pressures in a newly developed high-pressure Bunsen flame rig are studied using detailed numerical simulations and laser diagnostics. In the numerical simulations, one-dimensional and two-dimensional axisymmetric configurations were considered employing detailed transport properties and chemical kinetic mechanisms. In the measurements, OH PLIF was employed. The aims are to improved the understanding of the structures of the flames at varying pressures, to measure the laminar flame speed at different pressures, and to quantify the accuracy of the Bunsen flame method for measurement of laminar flame speed at different pressures. The stoichiometric and fuel-rich flames were found to exhibit a two-reaction-zone structure: an inner premixed flame in which the fuel was converted to CO and H2, and an outer diffusion flame in which CO and H2 were oxidized further to form combustion products. With increasing pressure, the inner premixed flame becomes thinner and the flame as a whole has the tendency to become unstable. Using the numerical and the experimental data, the methods of flame-cone-angle and flame-area were used to extract the laminar flame speed for different equivalence ratios and pressures. The flame-cone-angle method showed slightly better accuracy than the flame-area method did. The accuracy of both methods became lower under high pressure conditions. The inlet velocity of the burner was shown to affect the accuracy of the extracted laminar flame speed. For a stoichiometric atmospheric flame it was found that the most suitable inlet velocity for the fuel/air mixture was about 6 times the laminar flame speed, yielding a flame length about 7 times the radius of the burner. With appropriate flame length, the mid-height of the flame showed a rather low flame stretch rate, the laminar flame speed being in close agreement with the unstretched laminar flame speed, the error being less than 6% for the flames that were studied.