Takuji Suzuki

Gas Movements in Front of Flames Propagating Across Methanol

Abstract The gas velocity profiles in front of the leading edges of flames propagating across methanol at initial temperatures Ti from -5°C to 35°C were measured by using high-speed schlieren photography combined with a hot gas tracer technique. For Tl much lower than the flash point Tlf, any appreciable gas movement could not be observed in front of the leading flame edge. Therefore, as considered in previous studies, preheating in this case was supposed to be mainly caused by convection in the liquid phase. For Tl slightly lower than Tlf, the maximum value of the extrapolated gas velocity across the flame front was found to be a fairly high value. This result could be consistently intepreted by considering the increase of the methanol vapor concentration in front of the leading flame edge due to preheating. For Tl above Tlf, the experimentally predicted aspects of the gas movements near the leading flame edges were found to coincide with the theoretically predicted aspects for the flame propagation thro...

Ion-current fluctuations recorded with a cylindrical electrostatic probe passing premixed flames

Abstract The ion-current fluctuation recorded with a cylindrical, electrostatic probe traversing a premixed flame has been studied. The experiments were performed by moving the probe across a rectangular or circular-nozzle burner flame. Effects of the ion-denesity distribution on the ion-current fluctuation were estimated by analyzing the phenomena near the probe on the basis of the continuum probe theory, and the deviation of the ion current was shown to increase with the ion-density gradient. By comparing the maximum ion currents to the probe traversing the flame in different directions, the deviation was found to decrease with increasing probe velocity. It was estimated to be less than 15% if the probe was used in the range of probe velocities higher than 3 m/sec. An analysis was made for inferring the ion-current fluctuations to be recorded with the probe being passed through flames of known configuration. The results showed that the ion-current fluctuation of an inclined or curved flame could be inferred fairly well on the basis of the fluctuation derived from a plane flame parallel to the probe axis.

Theoretical Simulation of Gas Movements in Front of Propagating Flames Through Layered Flammable Mixtures

Abstract Detailed gas movements ahead of propagating flames through layered flammable mixtures have been theoretically simulated on the basis of convenient models. In the analysis, two-dimensional inviscid flow fields and constant flame velocities during propagation are assumed. Thermal expansion of gas due to combustion at the flame front is replaced by an imaginary source behind the flame front. Two cases are analyzed. In one case, a flat, flammable mixture layer, through which a flame propagates, is assumed to be established in an open space, and in the other case, it is assumed to be established in contact with a liquid or solid surface parallel to the layer. Equations representing the flow fields, fuel concentration distributions, gas particle paths, and flame front shapes are derived. The theoretical results obtained in the present study are shown to agree fairly well with the experimental results presented in previous reports.

Flame Spread Over Fuel Soaked Sand In An Opposed Air Stream

Flame spread over kerosene soaked sand in a quiescent, atmosphere and an opposed air stream has been examined, and the effects of the air stream on the fl arne spread mechanisms are discussed. The aspects of spreading flames, the flame spread rate, and t.he temperature d i.s t.r i buti.ons near leading flame edges were examined for various initial temperatures T. of kerosene soaked sand and free stream velocities U of the opposed i. air stream. As U increases, the flame near Lt.s leading edge approaches the sand surface. The flame spread r a t.e Vf at U=O increases considerably with T.. As U increases, V f decreases rapidly and then gradually decreas1?s until U becomes 275 cmls in the Li.mi ts of present e xpe r i.men ts . When U exceeds 275 cm y s , the flame spread becomes unstable and V f decreases rapidly. In the case of U=O, the heat transfer from the flame zone t.o the unburned region ahead of the leading flame edge seems t.o depend largely on the flame radiation. On the other hand in the case of U 0, the heat transfer t.o the unburned region seems to depend largely on t.he heat conduction through the sand layer. The stabil izat ion of the leading flame edge associated with the aerodynamic atru c tur e of the reverse flow region is found t.o be necessary for the stable flame spread.

Flame propagation across a liquid fuel in an air stream

The behavior of flames propagating across methanol in opposed and concurrent air streams was examined using high-speed schlieren photography, and the propagation mechanisms of these flames are discussed. For the free stream velocity U f gradually decreases as the absolute value |U| of U increases. The absolute value |U c | of the critical value U c at which V f becomes 0 is found to be much larger than the flame propagation velocity V f 0 at U=0. While for U>0, i.e., in the case of the concurrent air stream, V f remains almost constant until U becomes equal to V f 0 , and becomes nearly equal to U in the range of U>V f 0 . In the case of U f 0 , the effects of the free stream on the flow in front of the leading edge of the flame seem to be reduced to a great extent, so that the flame can propagate without a remarkable decrease in V f even in the opposed air stream of a much higher velocity than V f 0 . On the other hand in the case of U>V f 0 , the flame propagation seems to be closely related to the behavior of the hot gas overhanging the methanol surface in fron of the leading edge of the flame. The heat from the overhanging hot gas is expected to increase the methanol vapor concentration in front of the leading flame edge and ignite it. Thus, an inclined flame can follow the hot gas even when the initial methanol temperature T 1 is lower than that of the flash point. The relation (ϱu/ϱy) sc 2 =B·S L 2 ·δ 0 is found to be valid for T 1 >15°C, in which ( ϱu/ϱy) sc is the critical velocity gradient at the methanol surface in the approach flow, B a constant, S L the burning velocity at the leading flame edge, and δ 0 the flammable layer thickness at U=0.

Behavior Of The Reverse Flow In Front Of The Leading Flame Edge Spreading Over Fuel-soaked Sand In An Air Stream

The behavior of the reverse flow in f r o nt. of t.h e leading flame edge spreading over kerosene-soaked sand in an a i r s t.r e a:n has been examined using a few flow visualization techniques, and r.h e role of t.h e reverse flow in the flame spread is discussed. In a wide range of the free stream velocities U from 30 t.o 210 cm/s, a stable reverse flow region in f r o nt. of the leading flame edge was observed clearly, and its horizontal dimension was found to be almost independent of U. As U increases, the velocity of the reverse flow increases. The reverse flow takes an important role in the stable flame spread in an opposed air stream, although it has no apprec iable e f f e ct. on the f lame spread rate. The reverse f low provides a slow gas stream region, through which gasified fuel as well as heat from the reacteion zone would be transferred in the upstream direction.

Flame front movements of a turbulent premixed flame

Flame front movements of a propane-air turbulent premixed flame were examined by using a micro-electrostatic probe with a pair of identical sensors. Based on the fact that the variation of the ion current to one sensor with time was very similar to that to the other, it was shown that local and instantaneous flame front movements could be determined by analyzing the ion currents. The most probable direction and velocity V fm of local flame front movements were estimated by examining the correlation between the fluctuating parts of ion currents to the sensors. Near the unburned mixture side of the turbulent flame zone, V fm was found to be smaller than the mean velocity of the unburned mixture stream at the burner port and the flame front moved most probably downstream gradually approaching the burner axis. Across the turbulent flame zone from the unburned mixture side to the burned gas side, V fm increased gradually and the most probable direction of flame front movements was away from the burner axis. A large number of data for the periods during which flame fronts moved from one sensor to the other were analyzed for a typical case and were divided into those obtained when flame fronts passing the sensors were moving toward the unburned mixture and the others. Consequently, the periods for these two groups were found to be differently distributed. Based on this result, it was confirmed that the local and instantaneous flame front movements of a turbulent premixed flame could be consistently interpreted by taking into account the effect of the laminar burning velocity on the flame front movements, in addition to the effect of the turbulence of the unburned mixture stream.

Simultaneous measurements of gas flow and flame front movement in a turbulent premixed flame zone

Instantaneous local flame front movement and the flow velicity of unburned mixture or burned gas near the flame front in a turbulent flame zone have been measured simultaneously by using a micro-electrostatic probe with three sensors and a dual beam LDV system. The mean gas flow was also measured to explore the flow field of the turbulent premixed burner flame. It was found that there are evident differences in the velicity and direction between the mean gas flow and the most probable flame front movement at the side of the turbulent flame zone. The flow velocity of the burned gas near the flame front was found to be generally larger than that of the unburned mixture when the flame front moves toward the burned gas side, although no appreciable difference in their flow velocities is seen when it moves toward the unburned mixture side. Based on the relation between the flame front velocity and the flow velocity of the unburned mixture near the flame front, it was estimated that the local flame velocity is equal to the vector sum of the laminar burning velocity and the flow velocity of unburned mixture near the flame front. The relation between the overall flame front movement and the instantaneous local flame front movements was discussed using a simple model of a flame front movement.

Flame behavior near steps bounding layered flammable mixtures

A study was made of flame behavior near steps bounding layered flammable mixtures established over liquid or solid surfaces. Experiments were conducted by using a long and narrow tray filled with methanol at temperatures above the flash point and steps of various heights placed at one end of the tray. The flame movements near the steps were observed in detail by using high-speed schlieren photography. When the step height h is less than a few times the thickness δ u of the flammable mixture layer before ignition, a flame is shown to climb on the step and to propagate along its top surface. The distance l of flame propagation along the top surface is found to increase as h decreases or as δ u increases. The flame movements near steps are simulated by an inviscid theory in which convective motion ahead of the leading flame front is evaluated by replacing thermal expansion of gas due to combustion with an imaginary source behind the leading flame front. The results show that the theoretically simulated flame movements near steps agree fairly well with the experimentally observed ones. Both the experimental and theoretical results indicate that l/h depends on δ ue / h , the ratio of the effective initial thickness δ ue of the flammable mixture layer to h.

Dynamic characteristics of flame fronts in a turbulent premixed flame zone

Instantaneous local flame front movements in the flame zone of a turbulent premixed burner flame have been examined using a micro-electrostatic probe with three sensors. Based on the measured values of the instantaneous inclined angle and velocity of the local flame front, frequency distributions at representative locations in the flame zone were derived. The results are shown to be consistent with those of the most probable direction and velocity measurements. Each frequency distribution is found to flatten as the measuring point moves from the unburned side of the flame zone to the burned side. A large number of velocity vectors of instantaneous, local flame front movements at representative locations are shown in vector distribution maps. Each vector was determined by using the values of the direction and velocity of the local flame front evaluated by analyzing three simultaneously recorded ion-current peaks. Local flame front movement is shown to change considerably from one instant to another. The scatter of the vector directions and magnitudes increases as the measuring point moves from the unburned side of the flame zone to the burned side. At the outermost point of the flame zone, the flame front moves downward at a certain instant, while at another instant the flame front velocity occasionally attains a value 3 times as large as the average value in the direction close to the average direction. The flame front movements at the top and side of the flame zone are shown to be almost the same, although the mean ion current and the flame zone width are different at these two locations.