Some observations on near-limit flames

Two hundred years of flammable limits

“…This phenomenon is described by Coward and Jones [11, Furno, et al [26], and elsewhere. A contemporary argument for adopting "downward" limits as "true" limits was countered in 1737 by Coward and Jones with reference to near-limit mixtures comprising 6% hydrogen in air and 5.6% methane in air [l]: "Both these mixtures propagate flame upward indefinitely and if ignited near the floor of a closed room would produce pressures of the order 1 and 4 atmospheres respectively, and mean temperatures of about 350" C and 1,200' C. Such conditions would burst windows and burn men.…”

Influence of the ignition source location on the determination of the explosion pressure at elevated initial pressures

“…In each of these three studies the flammable range at atmospheric pressure did not show the sharp cut-off in the explosion pressure-concentration plot. Such behaviour was, however, observed at atmospheric pressure for n-butane-air and hydrogen-air mixtures by Furno, Cook, Kuchta and Burgess (1971) and for methane-air, propane-air and hydrogen-air mixtures by Cashdollar, Zlochower, Green, Thomas and Hertzberg (2000). The main difference between these last two surveys and the others is the size of the explosion vessel: Furno et al (1971) and Cashdollar et al (2000) used vessels of 120 dm 3 and 25.5 m 3 , respectively, while we, Pekalski et al (2004) and Goethals et al (1999) used vessels ranging from 4.2 to 20 dm 3 .…”

“…The concentration range with low explosion pressure ratios is caused by the fact that in this concentration range the flame can only propagate upwards. This is evidenced by visual observation (Cashdollar et al, 2000;Crescitelli, Russo, Tufano, Napolitano & Tranchino, 1977;Furno et al, 1971), thermocouple measurements at different locations inside the explosion vessel (Pekalski et al, 2004) and analysis of combustion products (Crescitelli et al, 1977;Pekalski et al, 2004). The underlying reason for the failure to propagate downwards is the effect of natural convection: if the buoyant rise velocity of the flame kernel is larger than the flame velocity, the flame is seen to travel only upwards (Andrews & Bradley, 1973;Crescitelli et al, 1977;Lovachev, 1971).…”

“…The difference between the measured and the calculated explosion pressures is caused by the large heat losses that occur when the flame reaches the wall. Especially for the top end and central igniter positions, the flame propagation will occur in different stages: upward propagation of a small developing flame kernel to the top of the vessel, followed by downward flame propagation for which a large part of the hot combustion products comes into contact with the cold vessel wall (Cashdollar et al, 2000;Crescitelli et al, 1977;Furno et al, 1971;Pekalski et al, 2004).…”

Flame Propagation in Closed Vessels and Flammability Limits

“…Tn closed vessels the flammability limits are derived by certain arbitrary criteria from maximum pressure increase or conversion degree as a function of mixture composition (Andrews and Bradley, 1973;Dittmar et al, 1965). Furno et al (1971) try to overcome these difficulties by defining two distinct boundary compositions, in connection with the observed flame behavior: the first one is identified by the smaller measurable increase of pressure (upward flammability limit), the second one by the maximum slope of the maximum explosion pressure-composition curve (downward flammability limit). In effect, for mixtures between the two above stated limits, the flame kernels propagate upwards due to buoyancy, and extinguish at the top of the vessel producing very small pressure rises.…”