Flame Propagation Direction

The flame propagation direction affects the type of flame arrester selected. An end-of-line or in-line deflagration flame arrester used for the protection of an individual tank may be of a unidirectional design because the flame will only propagate from the atmosphere towards the tank interior. A bidirectional flame arrester design, however, is needed for an in-line application in a vapor recovery (vent manifold) system because the vapors must be able to flow from the tank interior into the manifold, or from the manifold into the tank interior. Consequently, flame may propagate in either direction. 

---

In small-scale laboratory tests (characteristic sample dimensions of less than 0.2-0.3 m), conduction or convection are the dominant heat transfer routes. It has been shown horizontally or vertically, that 90% of the heat generated in a laminar flame spreading along a thermally thick PMMA sample is transferred by conduction across the condensed phase, and only about 10% by convection through the gas phase. In the case of thermally thin samples, on the other hand, the heat is predominantly transferred through the gas phase, independent of the flame propagation direction... 

This arrester is nsnally designed to be effective in one direction only. However, hydranlic arresters exist that are reported to be effective in preventing flame propagation in both directions. Tests to establish this on a particnlar hydranlic arrester design are described by Flessner and Bjorklnnd (1981). 

Flashback Undesired flame propagation opposite to the direction of flow. It is also used to describe failure of a flame arrester element. 

Unidirectional Flame Arrester An in-line flame arrester that is designed to stop flame propagation approaching from only one direction. 

History of upward flame propagation and extinction in lean limit methane/air mixture. Square 5 cm x 5 cm vertical tube. Green color frames indicate PIV flow images. Red color represents direct photography of propagating flame. Extinction starts just after frame c. Framing rate... 

Quenching process of a flat hmit flame, propagating downward from the open end of the tube in mixture with 2.20% CjHg observed by schlieren system with superimposed direct photography. Square tube 125 mm x 125 mm x 500mm. Time interval between frames 0.3 s. 

On the basis of the observations and results so far presented, it is obvious that the properties of limit flames are very different depending on the width of the quenching channel, the equivalence ratio, and the direction of flame propagation. The reasons for this are detailed in the following four sections. 

Numerical simulation of a spirming detonation in Hj/air mixture in a circular tube at various times. Gray and green space isosurfaces in pressure are the detonation front and the pressure of 6 MPa. White arrow propagating direction of the detonation front, pink arrow rotating direction of the transverse detonation. TD—transverse detonation, and LT—long pressure trail. (Reprinted from Tsuboi, N., Eto, K., and Hayashi, A.K., Combust. Flame, 149,144,2007. With permission.)... 

It is important to specify the direction of flame propagation. Since it may be assumed as an approximation that a flame cannot propagate downward in a mixture contained within a vertical tube if the convection current it produces is faster than the speed of the flame, the limits for upward propagation are usually slightly wider than those for downward propagation or those for which the containing tube is in a horizontal position. 

In configuration (4) a turbulent premixed flame is stabilized at the exit of a tube by a pressure gradient along the direction of mean flame propagation, induced by transverse swirl, but again the turbulence is inhomogeneous and there is appreciable mean flow. 

We note also research by Ya.B. performed together with G. I. Baren-blatt in 1957 [39], which has proved to be a key to understanding not only the problems of stability and flame propagation to which it was directed, but a far wider range of phenomena as well. This is the problem of the stability of invariant solutions to problems in mathematical physics. The question was posed thus what is flame stability Let us perturb the temperature and concentration distribution in a flame. Which flame shall we call stable We recall that the solution of the problem of a flame as a... 

Moreover, the phenomena of combustion themselves prove to be more complicated. For a long period the study of combustion broke away from chemical kinetics and set itself its own specific tasks. These included especially studies of the influence of instrumental parameters on ignition, flame propagation and limits, i.e., the influence of the diameter and length of tubes, the form of vessels, the direction of propagation, etc. 

Results pertinent to the theory of critical diameter are contained for the most part in earlier works by English authors. Despite his erroneous assumptions, Holm obtained the correct relation between the critical diameter and the flame velocity (1.4.6). The remarkable work by Daniell on the theory of flame propagation contains an analysis of the influence of heat losses. The losses enter directly into the equation describing the temperature distribution in the flame zone. A solution exists only for heat losses which do not exceed a certain limit, and under critical conditions (at the limit of propagation), the flame velocity drops to a certain fraction (40-50%) of the theoretical flame velocity. Daniell was also the first to indicate definitely that the flame velocity cannot be constructed from thermal quantities alone and by dimensional considerations must be proportional to the square root of the reaction rate. 

Table 4 Dependence of the Limiting Concentration of Fuel on the Direction of Flame Propagation in Mixtures with a Deficiency of Fuel...

When /k < 1, conditions in the flame pellet, which receives fuel from the surrounding medium by diffusion, are even less favorable than for normal flame propagation with respect to the gas. Therefore the flame moves normally with respect to the gas in both directions, upward and downward, and identical mechanisms yield practically identical limits z is close to unity. 

In the one-dimensional theory of NM we can imagine only a flat flame front the temperature varies only as a function of the coordinate along which the flame propagates, and the direction of the temperature gradient coincides with the direction of propagation. The gradient is small, as is the surface through which heat is transferred (it is equal to the tube cross-section). 

Direct experiments by Payman [8] show that at atmospheric pressure the velocity of uniform flame propagation in tubes of diameter 2.5 -5 cm does not fall below 12-20 cm/sec. On the basis of work by Coward and Hartwell [9], who found the relation between the normal velocity (i.e., the velocity of the flame with respect to the gas which we used above) and the velocity of uniform flame propagation with respect to the walls of the tube in which the gas is enclosed, we know that the data observed by Payman correspond to a normal flame velocity of 4-8 cm/sec. The values of the normal flame velocity at the limit obtained by Coward and Hartwell themselves are of the same order. 

---