| An experimental and computational study of the effects of diluents on the properties of laminar premixed hydrogen flames in microgravity was carried out during the present investigation. The microgravity experiments made use of a short-drop free-fall laboratory facility that provided at least 450 ms of 10-2 g conditions. Outwardly propagating spherical flames were used to measure laminar burning velocities at fuel-equivalence ratios of 1.0 and 1.8, and at pressures of 0.5, 0.7 and 1.0 atm, containing varying concentrations of He, Ar, N2 and CO2 as fire suppressants. Burning velocities were also computed using the steady, one-dimensional laminar premixed flame code PREMIX with detailed chemical kinetics, transport properties, and radiative heat loss.; Measured and computed results both showed the suppressant effectiveness (based on the reduction of burning velocity for a given concentration) to increase in the order He, Ar, N2 and CO2. This is attributed to two different physical effects: quenching of chain reaction due to the lower temperatures produced by the suppressant effects on the specific heats of the gases, and changes in the transport rates in the flame reaction zone. However the suppressants can also decrease the Markstein number, especially for CO2, causing the flames to become more susceptible to preferential-diffusion instability. The resulting increase in flame surface wrinkling increases the burning velocity, thus counteracting the desired effect of the flame suppressant. Far from the flammability limit, the agreement between measured and computed laminar burning velocities was found to be good. For near-limit flames, however, the computed and measured burning velocities deviated noticeably, with the observed differences being substantially larger than the measured uncertainties. Sensitivity analyses suggest that inaccuracies in three-body termination rates for H+O2+M=HO2+M reactions, particularly in the third-body chaperon efficacy of various species M, and in mass diffusion coefficients, particularly for H2 diffusion, are the most likely explanation for these near-limit differences. |
