Citation

White D, Beyler CL, Fulper C, Leonard J. Fire Safety J. 1997; 28(1): 1-31.

Copyright

(Copyright © 1997, Elsevier Publishing)

DOI

unavailable

PMID

unavailable

Abstract

Flame spread rates and flame spread induction times have been measured for JP-5, JP-8 and mixtures of these fuels over the temperature range 10-90[deg]C. The flame spread data were correlated on the basis of the initial liquid temperature relative to the closed cup flashpoint. Liquid-phase-controlled flame spread is observed for liquid temperatures 15[deg]C above the closed cup flashpoint, flame spread is via gas-phase-controlled flame spread and flame spread rates range from 12 to 160 cm/s. The transition at V = 12 cm/s and the maximum flame spread rate are consistent with present knowledge of gas-phase flame spread and burning velocities.Differences between the present data and other available data from the literature are the result of differences in flashpoint interpretation for hydrocarbons versus alcohols and the use of very narrow flame spread pans by early investigators. The use of pans of only 1-6 cm causes large changes in flame spread rate with temperature which are not observed in wider pans. Therefore the use of narrow tray flame spread experiments gives an incorrect indication of the flame spread rate to be expected in large-scale (realistic) conditions. Pan widths of at least 20 cm are required to avoid these small-pan-width effects.Variations in the flashpoint of mixtures of JP fuels were successfully predicted using the method of Affens and McLaren by treating each JP fuel as a single fuel characterized by a single vapour pressure-temperature relation and assuming that mole fractions in the liquid phase may be approximated by the volume fraction of the JP fuel. Flame spread induction times were shown to be a function of the liquid temperature relative to the closed cup flashpoint. The induction time was also shown to be dependent on the strength and nature of the ignition source.