ALBERT

Description: We employ the opposed flow flame-spread configuration in order to examine flame-front instability of diffusion flames near cold, solid boundaries. The thermo-diffusive and hydrodynamic instabilities can transform an initially planar flame front into an irregularly curved, corrugated, possibly fragmented front. Under ordinary 1-g conditions, the buoyancy-induced flow masks the thermo-diffusive and hydrodynamic instabilities and produces planar flames. Such stable spreading flames have been observed for decades in laboratory experiments. Experiments in zero gravity are necessary to produce unstable flame fronts. The thermo-diffusive/hydrodynamic microgravity instability appears in diffusion flames such as, for example: the candle flame oscillations observed by Dietrich et al.; smolder instabilities on a recent Space Shuttle flight. Drs. T. Kashiwagi and S. Olson have attributed the latter to a lowered oxygen transport rate to the hot, reactive surface. Consider a burning surface near the flame extinction limit. The flow, or stretch, induced by the diffusion flame is weak, hence buoyancy plays a small role, thereby enabling previously secondary mechanisms, such as differential thermo-diffusion, to become the most important mechanisms. The flame leading edge becomes unstable; and diffusion flame breakup, oscillation, and rejoining all occur at a measurable frequency of approximately O(1 Hz). This project has only begun in January of this year, 1999. To date, there have been no flight experiments on flame spread instabilities. However, we have made numerous experiments in the NASA 2.2 and 5 second drop towers on flame spread over very thin cellulosic fuels. We have been very fortunate through a combination of factors, to be explained, to obtain some interesting, perhaps even compelling, results on diffusion flame instability in the presence of heat losses to cold surfaces.

Description: The effects of imposed flow velocity on flame spread along open edges of a thermally thin cellulosic sample in microgravity were studied experimentally and theoretically. In this study, the sample was ignited locally at the middle of the 4 cm wide sample, and subsequent flame spread reached both open edges of the sample along the direction of the flow. The following flame behaviors were observed in the experiments and predicted by the numerical calculation, in order of increased imposed flow velocity: (1) ignition but subsequent flame spread was not attained, (2) flame spread upstream (opposed mode) without any downstream flame, and (3) the upstream flame and two separate downstream flames traveled along the two open edges (concurrent mode). Generally, the upstream and downstream edge flame spread rates were faster than the central flame spread rate for an imposed flow velocity of up to 5 cm/s. This was due to greater oxygen supply from the outer free stream to the edge flames and more efficient heat transfer from the edge flames to the sample surface than the central flames. For the upstream edge flame, flame spread rate was nearly independent of, or decreased gradually with, the imposed flow velocity. The spread rate of the downstream edge, however, increased significantly with the imposed flow velocity.

Description: Low-pressure blowoff experiments were conducted with a stagnation flame stabilized on the forward tip of cast PMMA rods in a vertical wind tunnel. Pressure, forced flow velocity, gravity, and ambient oxygen concentration were varied. Stagnation flame blowoff is determined from a time-stamped video recording of the test. The blowoff pressure is determined from test section pressure transducer data that is synchronized with the time stamp. The forced flow velocity is also determined from the choked flow orifice pressure. Most of the tests were performed in normal gravity, but a handful of microgravity tests were also conducted to determine the influence of buoyant flow velocity on the blowoff limits. The blowoff limits are found to have a linear dependence between the partial pressure of oxygen and the total pressure, regardless of forced flow velocity and gravity level. The flow velocity (forced and/or buoyant) affects the blowoff pressure through the critical Damkohler number residence time, which dictates the partial pressure of oxygen at blowoff. This is because the critical stretch rate increases linearly with increasing pressure at low pressure (sub-atmospheric pressures) since a second-order overall reaction rate with two-body reactions dominates in this pressure range.