Combustion and Flame, Vol.112, No.4, 575-592, 1998
Velocity and thermal structure, and strain-induced extinction of 14 to 100% hydrogen-air counterflow diffusion flames
Extensive results from axisymmetric convergent-nozzle and straight-tube opposed jet burners (OJBs) characterized strain-induced extinction of unanchored (free-floating), laminar H-2/N-2-air flames. Parameters included (a) plug-few and parabolic input velocity profiles, (b) jet exit diameters ranging 2.7 to 7.2 mm for nozzles and 1.8 to 10 mm for tubes, (c) various relative jet gaps, and (d) 14 to 100% H-2 in the fuel jet. Extinction, a sudden rupture (blowoff) of the mostly-airside disk flame, occurred as fuel and air flows were slowly increased and a critical radial strain rate was exceeded. The disk flame was restored at much lower flows, unique to H-2 systems. Focusing schlieren, thermocouple, and airside LDV (and PIV) data confirmed the one-dimensional (1-D) character of nozzle-OJB flow fields; axial widths of velocity-and thermal-layers varied as (input strain rate)(-1/2) for both nozzles and tubes. The global approximation of a 1-D applied stress rate (ASR), using average air jet velocity divided by exit diameter, enabled high quality correlations of extinction data with varied H, concentrations, for both nozzles and tubes. Pre-extinction ASRs for nozzles agreed closely with LDV-measured centerline input strain rates; for tubes, however, an empirical factor of 3 produced close agreement. For methane-air extinction, nozzle-OJB ASRs agreed within 4% of independent nozzle and Tsuji burner results. For extinction of 100% H-2-air, an ASR of 5670 1/s compared with 7350, 8140, and 8060 from independent 1-D numerical evaluations using potential-flow inputs; for 50 to 14% H-2 inputs, agreement was much closer. The nozzle-ASR/tube-ASR ratio for extinction was greater than or equal to 3 for <50% H-2 inputs, 2.74 +/- 0.03 for 50 to 100% H-2 inputs, and 2.83 for methane-air. Because these ratios exceeded 2.0, which "accounted" for centerline velocity inputs from parabolic profiles, an additional 3/2 radial strain component was apparent and was supported by the axial velocity gradient measurements.