Flammability limits: an invited review

An approximate modeling technique for enclosure fires is proposed and partially tested.The technique is based on the hypothesis that the combustion of a fuel pile in an enclosure, relative to the combustion of the same pile in open space (unconfined), depends on associated differences in the fluid-mechanical properties and the chemical composition of the gas supply to the fire. Previous work on unconfined fires of wood cribs is reviewed and consolidated in a form which relates the burning rate per unit exposed surface area to stick thickness and a function of a porosity factor. This factor differs in a small but significant way from a porosity factor proposed previously by Gross (Ref. 10). Application of the modeling hypothesis, under the assumption of large Grashof Numbers for the convective flow, then leads to the following sufficiency conditions for predicting fire properties on one enclosure scale from experiments on another scale: (1) the enclosures must be geometrically similar; (2) unconfined burning rates of the fuel piles in the two enclosures must be in the ratio of the linear-scale ratio to the 5/2-power; (3) porosities of the fuel piles [defined under Eq. (10)] must be conserved from one scale to the other; and (4) the thermal properties of the walls must be modeled for proper response [Eqs. (17)]. Experiments available for testing the modeling technique did not satisfy the third and fourth conditions. Nevertheless, data on burning rates, selected gas species (oxygen, carbon monoxide, carbon dioxide), and over-all fire behavior conformed well with the modeling hypothesis. Temperatures did not conform, evidently because wall properties had not been modeled. No discernable effects of fuel porosity were observed, but the scope of the experiments was too limited to allow a firm conclusion on porosity effects.

Modeling of enclosure fires

An experimental study of flammability limits of LPG/air mixtures

Flame propagation in methane-air mixtures with transverse concentration gradients in horizontal duct

The concentration changes with time and space of methane following its release into air, under the combined effects of molecular diffusion and natural convection were examined. A smooth, cylindrical, vertical flame tube of 116.0 inches (2.946 m) long and 2.5 inches (63.5 mm) inside diameter was used. The tube was separated into two sections by a plane shutter: a lower small fuel section and an upper large air section. Fuel stratification, caused mostly by convective diffusion, was initiated following the opening of the shutter. The concentration profile as a function of diffusion time at specified regions of the tube was determined by measuring the velocity of sound in the medium using a specially developed “ultra-sonic gas analyzer”. To predict the experimental concentration profile in such a system an eddy diffusivity that is a function of both the diffusion time and the local concentration gradient is needed. The range of diffusion time within which a flammable zone was developed within the stratified mixtures was determined at various locations employing a periodic spark source, for downward flame propagation. It was observed that the flammability limits of downward flame propagation within stratified methane-air mixtures were wider than those obtained for the corresponding homogeneous quiescent mixtures. Flame speeds of both accelerating and decelerating flames within stratified methane-air mixtures caused by the natural convection process were found to be higher than those encountered under homogeneous conditions at the corresponding methane concentration. Some possible reasons for the enhanced flame speeds are discussed.

Ignition and flame propagation within stratified methane-air mixtures formed by convective diffusion

Characteristics of flames propagating in vertical tubes in the direction of steep methane-air concentration gradients of about 1.3 volume-percent methane per cm were investigated using interferometric techniques. The maximum flame velocity for upward and downward propagation in both 2.5- and 5.0-cm diameter tubes occurred in the fuel-rich region; lower flame speeds were observed with the smaller tube diameter at corresponding methane concentrations. Following flame extinguishment in upward propagation, the hot product gas region behind the flame front continued to rise in a predictable fashion as a coherent column into the nonflammable fuel-rich region. This hot column rose at a constant velocity, approximately equal to the flame speed at the time of flame extinguishment. Flame extinguishment limits in the layered mixtures were found not to be strongly dependent on tube diameter. Interferometric techniques developed in this study appear to be useful for the rapid measurements of flammability limi...

Israel Llebman

Papers

Dynamics of Flame Propagation through Layered Methane-Air Mixtures