A fundamental difference between a partially premixed flame and an equivalent premixed (or nonpremixed) flame pertains to the existence of multiple synergistically coupled reaction zones. A “triple flame” is a type of partially premixed flame that contains a fuel-rich premixed reaction zone, a fuel-lean premixed reaction zone, and a nonpremixed reaction zone. The objective of this investigation is to examine gravity effects on the flame structure and flow instabilities related to partially premixed triple flames. (An earlier investigation by us dealing with gravitational effects on partially premixed double flames essentially considered steady 0- and 1-g flames.) A detailed numerical model is employed to simulate a methane-air triple flame established on a slot burner. A relatively detailed mechanism involving both C1- and C2-containing species and 81 elementary reaction steps is used to represent the CH4-air chemistry. Validation of the computational model is provided through a comparison of predictions with nonintrusive measurements. The results indicate that the overall triple flame structure is determined by interactions between the three reaction zones, and can be controlled by changing the mixture velocity, equivalence ratio, and gravitational acceleration. While the inner rich premixed reaction zone is weakly affected by gravity, the central nonpremixed and outer lean premixed reaction zones exhibit significant differences at 0 and 1 g. For 0 g flames, these two reaction zones move away from the centerline compared to the corresponding 1 g flames, since the entrainment of the lean outer flow is reduced in the absence of buoyant advection. Velocity vectors outside the lean premixed zone are directed away from the centerline due to flow dilatation in a 0 g flame, whereas they are directed towards the centerline due to the buoyancy-induced entrainment that occurs in a corresponding 1 g flame. Consequently, there is an increased physical separation and reduced heat and mass transport between the three reaction zones of the 0 g flame. The nonpremixed reaction zone height decreases due to the increase in residence time at 0 g. The reduced advection and (transport) at 0 g results in a flame that is less compact and has thicker reaction zones and which, therefore, is more sensitive to flow or stoichiometry perturbations. The flame structure and the interactions between the three reaction zones are found to be well-represented in terms of a modified conserved scalar ξ. A fundamental difference between the 0 and 1 g triple flames is due to their transient behavior and markedly different response to changes in the coflow velocity. Our simulations indicate the presence of a shear-induced convective instability and a buoyancy-induced global instability in laminar triple flames that depend upon the magnitude of coflow velocity and gravitational acceleration. The outer premixed reaction zone of 0 g flames exhibits a hitherto unreported weak oscillatory behavior at higher coflow rates that is related to a Kelvin–Helmholtz instability of the momentum shear layer. The instability in the 0 g flame is confined to the outer premixed zone. In contrast, decreasing the coflow velocity at 1 g causes a well-organized flickering of the outer reaction zone that can affect all three reaction zones. The flickering frequency is relatively insensitive to changes in the coflow velocity, as both computed and measured frequencies are found to be in a narrow range of 9–12 Hz. However, the flickering amplitude exhibits strong sensitivity to the coflow velocity as it is reduced to a value smaller than the inner jet velocity, and becomes noticeably large causing oscillations in all three reaction zones. There is a good agreement between the predicted and measured dynamics of the flickering 1 g flames.

No abstract

A triple flame is a partially premixed flame that contains two premixed reaction zones (one fuel-lean and the other rich) that form exterior wings and a nonpremixed reaction zone that is established in between these wings. The three reaction zones merge at a "triple point." Triple flames may play an important role in the stabilization and liftoff of laminar nonpremixed flames. They are also of fundamental importance in the reignition of turbulent mixtures. Despite their importance, many aspects of triple flames have not been adequately investigated and are, consequently, not clearly understood. Herein, laminar triple flames stabilized on a Wolfhard-Parker slot burner are investigated. The flow consists of a rich mixture of methane and air emerging from the inner slot and a lean mixture from two symmetric outer slots. In this configuration the three reaction zones that characterize a triple flame can be clearly distinguished. The loci of the "triple points" form a "triple line" in this planar configuration. The velocity field is characterized using laser Doppler velocimetry, and the temperature distribution using laser interferometric holography. In addition, C* 2 -chemiluminescence images of the three reaction zones are obtained. A detailed numerical model is employed to completely characterize the flame. It is based on a 24-species and 81-reaction mechanism. The numerical results are validated through comparisons with the experimental measurements. Our results focus on the detailed structure, the interaction between the three reaction zones, the dependence of the flame structure on the initial velocities and mixture equivalence ratios, and the dominant chemical pathways. The lean premixed reaction zone (external wing) exhibits different features from the rich premixed reaction zone. In particular, it is characterized by strong HO 2 formation and consumption reactions, and by relatively weak methane consumption reactions. Radical activity is higher in the nonpremixed reaction zone than in the other reaction zones. Overall, radicals from the nonpremixed reaction zone are transported to both the rich and lean premixed reaction zones where they attack the reactants. Simplifying the chemical mechanism by removing the C 2 -containing species produces significant differences in the predicted results only for the inner rich premixed reaction zone.

No abstract

We have investigated lifted triple flames and addressed issues related to flame stabilization. The stabilization of nonpremixed flames has been argued to result due to the existence of a premixing zone of sufficient reactivity, which causes propagating premixed reaction zones to anchor a nonpremixed zone. We first validate our simulations with detailed measurements in more tractable methane-air burner-stabilized flames. Thereafter, we simulate lifted flames without significantly modifying the boundary conditions used for investigating the burner-stabilized flames. The similarities and differences between the structures of lifted and burner-stabilized flames are elucidated, and the role of the laminar flame speed in the stabilization of lifted triple flames is characterized. The reaction zone topography in the flame is as follows. The flame consists of an outer lean premixed reaction zone, an inner rich premixed reaction zone, and a nonpremixed reaction zone where partially oxidized fuel and oxidizer (from the rich and lean premixed reaction zones, respectively) mix in stoichiometric proportion and thereafter burn. The region with the highest temperatures lies between the inner premixed and the central nonpremixed reaction zone. The heat released in the reaction zones is transported both upstream (by diffusion) and downstream to other portions of the flame. Measured and simulated species concentration profiles of reactant (O 2 , CH 4 ) consumption, intermediate (CO, H 2 ) formation followed by intermediate consumption and product (CO 2 , H 2 O) formation are presented. A lifted flame is simulated by conceptualizing a splitter wall of infinitesimal thickness. The flame liftoff increases the height of the inner premixed reaction zone due to the modification of the upstream flow field. However, both the lifted and burner-stabilized flames exhibit remarkable similarity with respect to the shapes and separation distances regarding the three reaction zones. The heat-release distribution and the scalar profiles are also virtually identical for the lifted and burner-stabilized flames in mixture fraction space and attest to the similitude between the burner-stabilized and lifted flames. In the lifted flame, the velocity field diverges upstream of the flame, causing the velocity to reach a minimum value at the triple point. The streamwise velocity at the triple point is Ϸ0.45 m s Ϫ1 (in accord with the propagation speed for stoichiometric methane-air flame), whereas the velocity upstream of the triple point equals 0.7 m s Ϫ1 , which is in excess of the unstretched flame propagation speed. This is in agreement with measurements reported by other investigators. In future work we will address the behavior of this velocity as the equivalence ratio, the inlet velocity profile, and inlet mixture fraction are changed.