Comprehensive scaling analyses of fire whirls based on experimental data

The phenomenon of fire whirls has been studied for many decades to understand the hazards they pose during wildland fires. Experimental and numerical simulations have been used to predict their occurrence, although challenges remain in using laboratory-scale data to predict occurrence in the field. In this work, we present comprehensive scaling analyses of experimental work on fire whirls in the literature. Several combinations of characteristic length scales (pool diameter, flame width, flame height, etc.) were used to define nondimensional quantities for burning rate, circulation and flame height. The relationships between these quantities were explored to compare the different normalization methods prevalent in the literature. The most important parameters to predict flame height are the representations of momenta in the axial and azimuthal directions, defined by nondimensional buoyant and circulation momenta, which are in turn estimated using heat-release rate and circulation, respectively. The data for experimental data obtained using gaseous-fuel burner and liquid-fuel pools show distinct differences in scaling behavior, which stems from differences in the flame width. Finally, for experiments with fuel pool diameter, d = 10.5 cm, in the classic four-wall, natural air entrainment configuration, the critical enclosure-wall height (H) required for stable fire whirl formation was found to be 3.5d.