Toward an understanding of the stabilization mechanisms of lifted turbulent jet flames: Experiments

This review discusses recent progress in understanding turbulent, lifted hydrocarbon jet flames and the conditions under which they stabilize. The viewpoint is from that of the empiricist, focusing on experimental results and the physically based theories that have emerged from their interpretations, as well as from the theoretically founded notions that have been supported. Pertinent concepts from laminar lifted flame stabilization studies are introduced at the onset. Classification in broad categories of the types of turbulent lifted flame theories is then presented. Experiments are discussed which support the importance of a variety of effects, including partial premixing, edge-flames, local extinction, streamline divergence and large-scale structures. This discussion details which of the categories of theories are supported by particular experiments, comments on the experimental results themselves and their salient contributions. Overall conclusions on the state of the field are drawn and future directions for research are also discussed.     

Stability characteristics of lifted turbulent partially premixed jet flames

The stability characteristics of partially premixed turbulent lifted methane flames have been investigated and discussed in the present work. Mixture fraction and reaction zone behavior have been measured using a combined 2-D technique of simultaneous Rayleigh scattering, Laser Induced Predissociation Fluorescence (LIPF) of OH and Laser Induced Fluorescence (LIF) of C₂H_x. The stability characteristics and simultaneous mixture fraction-LIPF-LIF measurements in three lifted flames with originally partially premixed jets at different mean equivalence ratio and Reynolds number are presented and discussed in this paper. Higher stability of partially premixed flames as compared to non-premixed flames has been observed. Lifted, attached, blow-out and blow-off regimes have been addressed and discussed in this work. The data show that the mixture fraction field on approaching the stabilization region is uniquely characterized by a certain level of mean and rms fluctuations. This suggests that the stabilization mechanism is likely to be controlled by premixed flame propagation at the stabilization region. Triple flame structure has been detected in the present flames, which is likely to be the appropriate model at the stabilization point.     

Acoustic enhancement of combustion in lifted nonpremixed jet flames

Lifted nonpremixed jet flames are often used in industrial processes and present inherent difficulties such as their reattachment to the burner, blowout, and poor combustion. One solution is to control the jet by acoustic forcing. For flames lifted in the hysteresis zone where anchoring may occur, forcing at high amplitudes and middle frequencies (around 200 Hz) changes the combustion regime and prevents reattachement. The common long yellow plume, due to soot radiation, vanishes. The flame becomes shorter, totally blue and stabilizes at a higher position above the burner. The phenomenon is explained using the results obtained by analyzing the flow dynamics with high-speed laser tomography, laser Doppler anemometry, particle image velocimetry, and Mie scattering techniques. Measurements show that the excitation periodically generates axial velocities higher than the maximum velocity of the hysteresis zone, leading to flame liftoff. Some primary and streamwise eddy vortices similar to natural instabilities develop during the jet deceleration. Contrary to the unexcited case, these structures, disorganized by the superimposition of the forcing wave, lead to quasi-homogeneous turbulence which provides efficient mixing and improves the combustion regime. Finally, the frequency is sufficiently high to avoid excessive fluctuations of the lift-off height and the reattachment to the burner.     

Large eddy simulation of a lifted turbulent jet flame

The flame index concept for large eddy simulation developed by Domingo et al. [P. Domingo, L. Vervisch, K. Bray, Combust. Theory Modell. 6 (2002) 529–551] is used to capture the partially premixed structure at the leading point and the dual combustion regimes further downstream on a turbulent lifted flame, which is composed of premixed and nonpremixed flame elements each separately described under a flamelet assumption. Predictions for the lifted methane/air jet flame experimentally tested by Mansour [M.S. Mansour, Combust. Flame 133 (2003) 263–274] are made. The simulation covers a wide domain from the jet exit to the far flow field. Good agreement with the data for the lift-off height and the mean mixture fraction has been achieved. The model has also captured the double flames, showing a configuration similar to that of the experiment which involves a rich premixed branch at the jet center and a diffusion branch in the outer region which meet at the so-called triple point at the flame base. This basic structure is contorted by eddies coming from the jet exit but remains stable at the lift-off height. No lean premixed branches are observed in the simulation or and experiment. Further analysis on the stabilization mechanism was conducted. A distinction between the leading point (the most upstream point of the flame) and the stabilization point was made. The later was identified as the position with the maximum premixed heat release. This is in line with the stabilization mechanism proposed by Upatnieks et al. [A. Upatnieks, J. Driscoll, C. Rasmussen, S. Ceccio, Combust. Flame 138 (2004) 259–272].     

Autoignited laminar lifted flames of propane in coflow jets with tribrachial edge and mild combustion

Characteristics of laminar lifted flames have been investigated experimentally by varying the initial temperature of coflow air over 800 K in the non-premixed jets of propane diluted with nitrogen. The result showed that the lifted flame with the initial temperature below 860 K maintained the typical tribrachial structure at the leading edge, which was stabilized by the balance mechanism between the propagation speed of tribrachial flame and the local flow velocity. For the temperature above 860 K, the flame was autoignited without having any external ignition source. The autoignited lifted flames were categorized in two regimes. In the case with tribrachial edge structure, the liftoff height increased nonlinearly with jet velocity. Especially, for the critical condition near blowout, the lifted flame showed a repetitive behavior of extinction and reignition. In such a case, the autoignition was controlled by the non-adiabatic ignition delay time considering heat loss such that the autoignition height was correlated with the square of the adiabatic ignition delay time. In the case with mild combustion regime at excessively diluted conditions, the liftoff height increased linearly with jet velocity and was correlated well with the square of the adiabatic ignition delay time.     

Emission characteristics and heat release rate surrogates for ammonia premixed laminar flames

Ammonia is a carbon-free fuel that has the potential to meet increasing energy demand and to reduce CO₂ emissions. In the present work, the characteristics of pollutant emissions in ammonia premixed laminar flames are investigated using one-dimensional simulations, and heat release rate (HRR) surrogates for ammonia combustion are proposed. Both atmospheric and high-pressure conditions were considered, and four representative mechanisms for ammonia combustion were employed. It is shown that the total emission of NO and NH₃ achieves a minimum around an equivalence ratio (?) of 1.1 under atmospheric conditions, and there is no noticeable emission of NO and NH₃ for ? = 1.1 ~ 1.5 under high-pressure conditions. Three HRR surrogates, [NH₃][OH], [NH₂][O], and [NH₂][H], were proposed based on the analysis of HRR and elementary reaction profiles. The performance of HRR surrogates was found to vary with equivalence ratios. For example, with the Miller mechanism, [NH₃][OH], [NH₂][O], and [NH₂][H] have the best performance under atmospheric conditions at ? = 1.15, 0.95 and 1.05, respectively, and under high-pressure conditions at ? = 1.11, 0.87 and 0.96, respectively. Similar conclusions can also be drawn with other mechanisms. These findings provide valuable insights into emission control and flame identification of ammonia combustion.     

Extinction of laminar partially premixed flames

Flame extinction represents one of the classical phenomena in combustion science. It is important to a variety of combustion systems in transportation and power generation applications. Flame extinguishment studies are also motivated from the consideration of fire safety and suppression. Such studies have generally considered non-premixed and premixed flames, although fires can often originate in a partially premixed mode, i.e., fuel and oxidizer are partially premixed as they are transported to the reaction zone. Several recent investigations have considered this scenario and focused on the extinction of partially premixed flames (PPFs). Such flames have been described as hybrid flames possessing characteristics of both premixed and non-premixed flames. This paper provides a review of studies dealing with the extinction of PPFs, which represent a broad family of flames, including double, triple (tribrachial), and edge flames. Theoretical, numerical and experimental studies dealing with the extinction of such flames in coflow and counterflow configurations are discussed. Since these flames contain both premixed and non-premixed burning zones, a brief review of the dilution-induced extinction of premixed and non-premixed flames is also provided. For the coflow configuration, processes associated with flame liftoff and blowout are described. Since lifted non-premixed jet flames often contain a partially premixed or an edge-flame structure prior to blowout, the review also considers such flames. While the perspective of this review is broad focusing on the fundamental aspects of flame extinction and blowout, results mostly consider flame extinction caused by the addition of a flame suppressant, with relevance to fire suppression on earth and in space environment. With respect to the latter, the effect of gravity on the extinction of PPFs is discussed. Future research needs are identified.     

The behaviour of lifted oxy-fuel flames in burners with separated jets

Oxy-fuel combustion in separated-jet burners has been proven to increase thermal efficiency and to have a potential for NOx emission reduction. This paper presents an investigation into confined, turbulent, oxy-flames generated by a burner consisting of a central natural gas jet surrounded by two oxygen jets. The study is focused on the identifying the influence of burner parameters on the flame characteristics and topology, namely stability, lift-off height and flame length. The effects of the natural gas and oxygen jet exit velocities, the distance separating the jets and the deflection of oxygen jets towards the natural gas jet are examined. The OH chemiluminescence. Results show that the lift-off heights increase when jet exit velocities and the distance separating the jets are increased. The deflection of oxygen jets decreases the lift-off height and increases the volume of flame in the transversal plane. The flame length increases principally with the oxygen exit velocity and the separation distance, and decreases considerably when the angle of oxygen jets is increased.     

Conditional analysis of lifted hydrogen jet diffusion flame experimental data and comparison to laminar flame solutions

Simultaneous point measurements of temperature, mixture fraction, major species, and OH concentrations in a lifted turbulent hydrogen jet flame are reprocessed to obtain the Favre average and conditional mean profiles. Large discrepancies between the Favre average and the ensemble average temperature, H₂O, and OH mole fractions are found at the lifted flame base, due to density weighting of fairly large samples of unreacted mixtures. Conditional statistics are used to reveal the reaction zone structure in mixture fraction coordinates. The cross-stream dependence of conditional reactive scalars, which is most notable at the lifted flame base and decreases to negligible levels with increasing streamwise positions, could be attributed to radial differences in both the Damköhler number and the level of partial premixing. Conditional results indicate that the lifted flame is stabilized at the outer region of the jet characterized by low strain rates and lean mixtures. Comparison of the measured conditional mean OH vs H₂O with a series of stretched laminar partially premixed flame and diffusion flame calculations reveals that strong partial premixing takes place at the lifted flame base and the strain rates vary from a=14,000 to 100 s⁻¹. The level of partial premixing and the strain rate decrease with increasing downstream locations. The range of estimated scalar dissipation rates (χ≈1–0.13 s⁻¹) at a further downstream location (x/D=33.3) is in agreement with reported values and the flame composition reaches an equilibrium condition at x/D=194.4. These results combined with previously reported data provide a benchmark data set for evaluation and refinement of turbulent combustion models for lifted hydrogen jet flame predictions.     

Self-excitation in laminar lifted free-jet propane flames diluted with nitrogen

Laminar lifted propane free-jet flames diluted with nitrogen were experimentally investigated to determine distinctive self-excitation regimes in the flame stability map and to elucidate the individual flame characteristics. Extremely low-frequency (<0.1 Hz) self-excitation was caused by conductive heat loss from the premixed wings to the trailing diffusion flame and could be explained by a proposed mechanism for edge flame extinction during triple-flame propagation as well as flame-front propagation. A newly observed heat-loss-induced flame blow-out mechanism was related to conductive heat loss from the premixed wings to the trailing diffusion flame. Additional self-excitation prior to flame blow-out was caused by buoyancy and also significantly affected by the conductive heat loss from the premixed wings to the trailing diffusion flame. This was explained in terms of triple-flame propagation and flame-front propagation. Self-excitation obtained in laminar lifted propane free-jet flames diluted with nitrogen was characterized by functional dependencies of the Strouhal number with related parameters.     

Effects of multi-component diffusion and heat release on laminar diffusion flame liftoff

Numerical simulations were conducted of the liftoff and stabilization phenomena of laminar jet diffusion flames of inert-diluted C₃H₈ and CH₄ fuels. Both non-reacting and reacting jets were investigated, including multi-component diffusivities and heat release effects (buoyancy and gas expansion). The role of Schmidt number for non-reacting jets was investigated, with no conclusive Schmidt number criterion for liftoff previously arrived at in similarity solutions. The cold-flow simulation for He-diluted CH₄ fuel does not predict flame liftoff; however, adding heat release reaction lead to the prediction of liftoff, which is consistent with experimental observations. Including reaction was also found to improve liftoff height prediction for C₃H₈ flames, with the flame base location differing from that in the similarity solution - the intersection of the stoichiometric and iso-velocity (equal to 1-D flame speed) is not necessary for flame stabilization (and thus liftoff). Possible mechanisms other than that proposed for similarity solution may better help to explain the stabilization and liftoff phenomena.     

Autoignited laminar lifted flames of methane, ethylene, ethane, and n-butane jets in coflow air with elevated temperature

The autoignition characteristics of laminar lifted flames of methane, ethylene, ethane, and n-butane fuels have been investigated experimentally in coflow air with elevated temperature over 800 K. The lifted flames were categorized into three regimes depending on the initial temperature and fuel mole fraction: (1) non-autoignited lifted flame, (2) autoignited lifted flame with tribrachial (or triple) edge, and (3) autoignited lifted flame with mild combustion.For the non-autoignited lifted flames at relatively low temperature, the existence of lifted flame depended on the Schmidt number of fuel, such that only the fuels with Sc > 1 exhibited stationary lifted flames. The balance mechanism between the propagation speed of tribrachial flame and local flow velocity stabilized the lifted flames. At relatively high initial temperatures, either autoignited lifted flames having tribrachial edge or autoignited lifted flames with mild combustion existed regardless of the Schmidt number of fuel. The adiabatic ignition delay time played a crucial role for the stabilization of autoignited flames. Especially, heat loss during the ignition process should be accounted for, such that the characteristic convection time, defined by the autoignition height divided by jet velocity was correlated well with the square of the adiabatic ignition delay time for the critical autoignition conditions. The liftoff height was also correlated well with the square of the adiabatic ignition delay time.     

Particle combustion rates in premixed flames of polydisperse metal—air aerosols

A new approach and experimental technique are proposed to determine times of metal particle combustion in flames of polydisperse aerosols. Laminar flames are produced in air at 1 atm, using aerosol jets formed by an electrostatic particulate method. The flame radiation intensities as a function of vertical coordinate are measured and compared with the flame radiation profiles reconstructed using experimental data and simplified models. The experimental data used include particle size distributions, flame velocities, and temperatures of metal ignition and combustion. The simplified models describe the particle ignition delay, combustion time, and particle flame radiation intensity as a function of particle diameter, D. Variable parameters of the models describing particle radiation intensities and combustion times are adjusted to achieve the best fit between the reconstructed and measured flame radiation profiles. A set of parameters providing the best agreement between the reconstructed and measured profiles is selected for several aerosol flames produced by powders of different sizes of the same material. These parameters are assumed to adequately describe particle combustion times and radiation intensities for the chosen material. The experimental radiation profiles for both aluminum and magnesium aerosol flames with particles of different sizes were found to be in very good agreement with the respective reconstructed profiles. For both metals, particle radiation intensities were well described by a D³-type expression. The combustion times for magnesium aerosol particles were well described by the traditional D²-law with the evaporation constant close to those reported earlier for single particles. Aluminum aerosol particle combustion was better described by a D¹-law and combustion times of fine (<80 μm) aluminum particles in the aerosol were somewhat longer than the reported earlier combustion times for single aluminum particles.     

Field measurements of soot volume fractions in laminar partially premixed coflow ethylene/air flames

The conditions under which soot is formed vary widely and depend upon several factors, including pressure, temperature, fuel type, combustor geometry, and extent of premixing. Although it is known that partially premixed flames (PPFs) can become either more or less sooting than their nonpremixed or premixed counterparts, the impact of partial premixing on soot formation across a large equivalence ratio and flow range is still inadequately understood. Comprehensive experimental data are relatively sparse for this important configuration. Herein, we report on soot formation in various ethylene/air PPFs utilizing full-field light extinction. The dimensionless extinction coefficient Kext is an important calibrated constant for the determination of the soot volume fraction for this measurement technique. We find that a value of Kext=7.1 provides results that are in good agreement with benchmark literature data for a nonpremixed flame. We examined the soot microstructures for two flames established at ?=∞ (i.e., nonpremixed) and 5. In both cases, the primary particles were found to be nearly spherical. In case of the nonpremixed flame the average primary soot particle diameter was ∼35 nm, but for the ?=5 flame it was ∼20 nm. However, the parameter responsible for the value of Kext is the average aggregate size and not that of the primary particles. The aggregate sizes are similar for the two flames. We consider this as verification of a constant Kext value over the entire equivalence ratio range. The addition of air to the fuel stream produces an initial increase in the flame height. Further air addition gradually decreases the flame height, which is followed by a more rapid decrease with larger premixing. Likewise, the peak soot concentration first increases with small amounts of air addition (or partial premixing of the fuel stream) and reaches a maximum value at ?∼24. With further air addition, as ? decreases below a value of 20, the soot volume fraction considerably decreases.     

Soot formation in laminar diffusion flames

Laminar, sooting, coflow diffusion flames at atmospheric pressure have been studied experimentally and theoretically as a function of fuel dilution by inert nitrogen. The flames have been investigated with laser diagnostics. Laser extinction has been used to calibrate the experimental soot volume fractions and an improved gating method has been implemented in the laser-induced incandescence (LII) measurements resulting in differences to the soot distributions reported previously. Numerical simulations have been based on a fully coupled solution of the flow conservation equations, gas-phase species conservation equations with complex chemistry, and the dynamical equations for soot spheroid growth. The model also includes the effects of radiation reabsorption through an iterative procedure. An investigation of the computed rates of particle inception, surface growth, and oxidation, along with a residence time analysis, helps to explain the shift in the peak soot volume fraction from the centerline to the wings of the flame as the fuel fraction increases. The shift arises from changes in the relative importance of inception and surface growth combined with a significant increase in the residence time within the annular soot formation field leading to higher soot volume fractions, as the fuel fraction increases.     

Combustion and heat release characteristics of hydrogen/air diffusion flame on a micro-jet array burner

Hydrogen (H₂) is considered as a carbon-free alternative fuel. The heat release characteristics of H₂ flame as a key parameter in its combustion process are unclear. In this study, the combustion and heat release characteristics of H₂/air diffusion flame on a micro-jet array burner were experimentally and numerically investigated. It is shown that the OH distribution and flame length based on Bilger mechanism are in good agreement with the experimental results. Furthermore, the intensity and distribution of OH and heat release rate can be adjusted by the thermal power and equivalence ratio. A uniform flame with intensive heat release rate can be achieved at a thermal power of 0.1 kW. R41: H + O₂ = OH + O and R43: H + O₂ + M = HO₂ + M are the main reactions with oxidizer consumption to form reactive radicals. R40: OH + H₂ = H₂O + H and R47: OH + OH = O + H₂O with OH consumption are the main heat release reactions at the upstream and downstream of the flame.     

Large-eddy simulation of a lifted methane jet flame in a vitiated coflow

The impact of burned gases on flame stabilization is analyzed under the conditions of a laboratory jet flame in vitiated coflow. In this experiment, mass flow rate, temperature, and the exact chemical composition of hot products mixed with air sent toward the turbulent flame base are fully determined. Autoignition and partially premixed flame propagation are investigated for these operating conditions from simulations of prototype combustion problems using fully detailed chemistry. Using available instantaneous species and temperature measurements, a priori tests are then performed to estimate the prediction capabilities of chemistry tabulations built from these archetypal reacting flows. The links between autoignition and premixed flamelet tables are discussed, along with their controlling parameters. Using these results, large-eddy simulation of the turbulent diluted jet flame is performed, a new closure for the scalar dissipation rate of reactive species is discussed, and numerical predictions are successfully compared with experiments.     

Lifted flames on fuel jets in co-flowing air

The three principal theories for the stabilization of lifted flames on turbulent jets of fuel are reviewed in the light of the most recent flame imaging experiments in the literature. Most of these experiments have been conducted with a small co-flow of air, but the observations are relevant to lift-off with higher ratios of co-flowing air to fuel jet velocity. The similarity solutions for jets in co-flow are developed, and data from a variety of fluid dynamic sources are assessed to yield the governing parameters for mean flow, turbulence and mixture fraction. New data for lifted flames on a methane jet in diffusing streams of co-flowing air are then presented. These data provide essential information on the intermittency, and on the properties of the jet conditioned on the presence of turbulent fluid. However, the co-flow lifts the flame to stabilize in better-mixed regions than in its absence. The ‘premixture’ model is confirmed for this situation, in which the lift-off heights were more than 20 jet diameters and where there is little intermittency at the stabilization radius. Nevertheless, mixing data for this geometry in the absence of a flame show that, with lift-off heights less than 20 jet diameters, the base of the flame would have been in the outer regions of the jet where the mixture of fuel in air only reaches stoichiometric proportions intermittently, with the passage of large eddies. Trading on many papers from the recent literature where this was the case, both experimental and computational insights as to the processes in this region are reviewed. A question remains about how ignition is maintained in these experiments with low turbulent lift-off. It is hypothesized that the mechanism is the diffusive heating of the slowly moving surrounding air which then provides an energy store for the incoming eddies. Further time-resolved observations of reaction zone and high temperature gas structure are required to test this model.     

The effects of heat loss on the burning velocity of cellular premixed flames generated by hydrodynamic and diffusive-thermal instabilities

The effects of heat loss on the burning velocity of cellular premixed flames are investigated by two-dimensional unsteady calculations of reactive flows based on the compressible Navier-Stokes equation and on the diffusive-thermal model equation. Hydrodynamic and diffusive-thermal instabilities are taken into account as contributing to the intrinsic instability of premixed flames. A sufficiently small disturbance is superimposed on a planar flame to obtain the relation between the growth rate and the wavenumber, i.e., the dispersion relation. As the heat loss becomes larger, the growth rate decreases and the unstable range narrows. This is because hydrodynamic instability caused by thermal expansion weakens for nonadiabatic flames. To investigate the characteristics of cellular flames, the disturbance with the linearly most unstable wavenumber, i.e., the critical wavenumber, is superimposed. As the superimposed disturbance evolves, the cellular-flame front forms due to the intrinsic instability. The lateral movement of cellular flames is observed at low Lewis numbers, and the behavior of cellular-flame fronts becomes more unstable for nonadiabatic flames. As the heat-loss parameter increases, the burning velocity of a cellular flame normalized by that of a planar flame increases at Lewis numbers lower than unity. By contrast, when the Lewis number is not less than unity, the burning-velocity increment decreases by increasing the heat loss. Diffusive-thermal instability thus has a pronounced influence on the unstable behavior and burning velocity of nonadiabatic cellular flames.