Fuel-lean Flames Research Articles

Measurement of laminar flame speeds through digital particle image velocimetry: Mixtures of methane and ethane with hydrogen, oxygen, nitrogen, and helium

A digital particle image velocimetry technique that is appropriate for the experimental derivation offundamental flame properties was implemented. The technique allows for the determination of the instantaneous flowfield and is essential for fluid mechanics measurements in reduced gravity environments. Measurements of laminar flame speeds were conducted in the stagnation flow configuration just before a flame undergoes a transition from planar to Bunsen flame. Results obtained for lean CH4/air and C2H2/air flames were found to be in close agreement with prefious laser Doppler velocimetry data. Subsequently, measurements were conducted for the CH4 and C2H2 flames by independently varying the equivalence ratio and flame temperature to distinguish between temperature and concentration effects. The laminar flame speeds were also calculated using the GRI-Mech 3.0 mechanism. It was convincingly shown that under high-O2 and low-temperature conditions, the experimental laminar speeds are over predicted by the simulations especially for C2H6 flames. Additional experiments were conducted by adding H2 to lean C2H6/air flames and by diluting those mixtures by either He or N2 to vary the flame temperature. While for the He dilution case, the predictions noticeably overpredict the experiments, for N2 dilution, close aggrement was observed. Analyses of the flame structures revealed that for those fuel-lean flames, the burning rate largely depends on the competition of the two-body branching and three-body termination reaction between H and O2. It was not possible to point to possible kinetic deficiencies other than referring to uncertainties associated with the rates and collision efficiencies of three-body reactions. The high-O2 low-temperature region is of interrest not ouly to lean-premixed combustion, but also to flame ignition, and requires further exploration.

Tri-brachial flames around two interactive droplets in flows with fuel vapor

The flame structures around two equal-sized and interactive fuel droplets immersed in high-temperature convective flows are examined. When the flame is sustained between the two droplets while the flow is pure air, an anchor-shape flame structure, namely the tri-brachial flame, is clearly observed. When far-field temperature or ambient equivalence ratio is increased, the flame propagates upstream and the fuel-lean premixed flame withers. However, if the ambient equivalence ratio is high to some extent such as 0.3, two wings of the flame, viz. the fuel-lean and the fuel-rich premixed flames, extend outward from the stoichiometric point. Once the flame encircles the leading droplet, it evolves into a double-flame structure consisting of a front premixed flame and an aft diffusion flame. Accordingly, the impact of the preceding two environmental factors on the droplets' flame structures is outlined.

Gas-phase modeling of impinging flames used for the flame surface modification of polypropylene film

Contact-angle measurements, atomic force microscopy (AFM), and X-ray photoelectron spectroscopy (XPS or ESCA) were used to characterize flame-treated biaxially oriented polypropylene (PP) films. While the surface of PP treated in a fuel-lean flame is highly oxidized, no watersoluble low-molecular-weight oxidized material (LMWOM) is formed by the flame treatment. A new computational model, SPIN, was used to determine the chemical composition of the impinging flames used to modify the PP. The SPIN model indicates that the species primarily responsible for the surface oxidation of the PP are OH, HO2, H2O2, and O2. Because the concentration of atomic O in the flame is low, there is little scission of the PP chains and no formation of LMWOM. AFM indicates that a 'nodular' surface topography is generated during the flame oxidation of the PP. The surface topographical features generated by flame treatment are probably the result of the agglomeration of intermediate-molecular-weight materials.

Double, Triple, and Tetra-Brachial Flame Structures around a Pair of Droplets in Tandem

The multiple flame structures for a pair of droplets in tandem in hot convective flows are analyzed numerically, by employing a body-fitted technique in association with a non-orthogonal curvilinear coordinate system. For selected droplet spacing and Reynolds numbers, three flame structures, including the double, triple, and tetra-brachial flames, are identified at specific ambient temperatures and droplet size ratios. As the droplet size ratio is small, the numerical results indicate that a double-flame structure composed of the front fuel-lean premixed flame and the rear diffusion flame develops at lower ambient temperature. Increasing the droplet size ratio results in the two flames experiencing the separation, then merger, and eventually destruction. Hence a triple flame consisting of both fuel-rich and fuel-lean premixed flames, as well as a diffusion flame appears. When the ambient temperature is higher, for the double-flame the front flame departs from the rear one as a result of the increased flame speed. On the other hand, because the flame moves upstream and the blowing effect of the trailing droplet grows, the triple-flame converts into a tetra-brachial flame at the larger droplet size ratio. The formation of these flames is attributed to the coupling of fuel vapor, cooling effect, and oxidizer depletion stemming from the leading droplet, and these observations can provide a more accurate modeling on droplets behaviors in spray.

Molecular growth and oxygenated species formation in laminar ethylene flames

Ethylene constitutes a key intermediate in the oxidation of practical hydrocarbon fuels. Past studies of ethylene oxidation in shock tubes, flow/stirred reactors, and flames have provided an understanding of several key aspects of the fuel oxidation chemistry. However, the current trend, toward leaner premixed combustion in practical devices results in significant changes in the combustion regime, and studies under fuel-lean flame conditions have become essential. The changes in the main reaction channels are significant, and the chemistry of oxygenated species comes to the fore. The present work compares the main reaction paths in fuel-rich and fuel-lean ethylene flames, with particular emphasis on the relative importance of addition, abstraction, and isomerization paths in the context of the formation of oxygenated and aromatic species. It is shown that in the fuel-lean flame, ethylene is mainly consumed by addition reactions, while in the rich flame environment, abstraction reactions leading to vinyl radical formation dominate. The balancing of the destruction chemistry of the latter and the importance of the rate and product distribution of the dolecular oxygen attack are discussed in detail. It is noted that the channel C2H4+O=CH2CHO +H is the dominant route to vinoxy under lean flame conditions, and the present work supports the acetyl channel (CH2CHO→CH3CO→CH3+CO) for vinoxy destruction at elevated temperatures. Potential benzene formation paths are assessed, and it is concluded that vinyl radical addition to vinylacetylene and propargyl radical recombination constitute the major benzene formation paths in rich ethylene flames. The present work thus further emphasises the need to consider multiple benzene formation channels.

Surface modification of polypropylene film using N2O-containing flames

Contact-angle measurements and X-ray photoelectron spectroscopy (XPS or ESCA) were used to characterize polypropylene (PP) films that were exposed to laminar premixed air: natural gas flames containing small quantities of nitrous oxide. During combustion, the nitrous oxide generates gas-phase nitrogen oxides that lead to the affixation of nitrogen-containing functional groups to the PP surfaces. Treatment of PP in nitrous oxide-containing flames also leads to an increase in surface oxidation and markedly improves wettability when compared with standard flame treatments. The chemical form of the nitrogen affixed to the PP surface is strongly dependent on the flame equivalence ratio. Fuel-lean flames tend to affix highly oxidized forms of nitrogen such as nitrate and nitro groups, while fuel-rich flames tend to affix less-oxidized nitrogen groups such as nitroso, oxime, amide, and amine. A computational model, SPIN, was used to elucidate the chemistry of the flame as it impinges upon the cooled PP surface. The SPIN modeling indicates that the principal reactive gas-phase species at or near the PP surface are O2, OH, H, NO, NO2, HNO, and N2O. A number of possible reactions between these species and the PP can account for the formation of the various nitrogen functional groups observed.

99/00811 Scientific and research activities of Oil and Gas Institute in Krakow

A digital particle image velocimetry technique that is appropriate for the experimental derivation offundamental flame properties was implemented. The technique allows for the determination of the instantaneous flowfield and is essential for fluid mechanics measurements in reduced gravity environments. Measurements of laminar flame speeds were conducted in the stagnation flow configuration just before a flame undergoes a transition from planar to Bunsen flame. Results obtained for lean CH4/air and C2H2/air flames were found to be in close agreement with prefious laser Doppler velocimetry data. Subsequently, measurements were conducted for the CH4 and C2H2 flames by independently varying the equivalence ratio and flame temperature to distinguish between temperature and concentration effects. The laminar flame speeds were also calculated using the GRI-Mech 3.0 mechanism. It was convincingly shown that under high-O2 and low-temperature conditions, the experimental laminar speeds are over predicted by the simulations especially for C2H6 flames. Additional experiments were conducted by adding H2 to lean C2H6/air flames and by diluting those mixtures by either He or N2 to vary the flame temperature. While for the He dilution case, the predictions noticeably overpredict the experiments, for N2 dilution, close aggrement was observed. Analyses of the flame structures revealed that for those fuel-lean flames, the burning rate largely depends on the competition of the two-body branching and three-body termination reaction between H and O2. It was not possible to point to possible kinetic deficiencies other than referring to uncertainties associated with the rates and collision efficiencies of three-body reactions. The high-O2 low-temperature region is of interrest not ouly to lean-premixed combustion, but also to flame ignition, and requires further exploration.

99/00812 Structural dependence of heterolytic bond dissociation energy of carbon-carbon g bonds in hydrocarbons

A digital particle image velocimetry technique that is appropriate for the experimental derivation offundamental flame properties was implemented. The technique allows for the determination of the instantaneous flowfield and is essential for fluid mechanics measurements in reduced gravity environments. Measurements of laminar flame speeds were conducted in the stagnation flow configuration just before a flame undergoes a transition from planar to Bunsen flame. Results obtained for lean CH4/air and C2H2/air flames were found to be in close agreement with prefious laser Doppler velocimetry data. Subsequently, measurements were conducted for the CH4 and C2H2 flames by independently varying the equivalence ratio and flame temperature to distinguish between temperature and concentration effects. The laminar flame speeds were also calculated using the GRI-Mech 3.0 mechanism. It was convincingly shown that under high-O2 and low-temperature conditions, the experimental laminar speeds are over predicted by the simulations especially for C2H6 flames. Additional experiments were conducted by adding H2 to lean C2H6/air flames and by diluting those mixtures by either He or N2 to vary the flame temperature. While for the He dilution case, the predictions noticeably overpredict the experiments, for N2 dilution, close aggrement was observed. Analyses of the flame structures revealed that for those fuel-lean flames, the burning rate largely depends on the competition of the two-body branching and three-body termination reaction between H and O2. It was not possible to point to possible kinetic deficiencies other than referring to uncertainties associated with the rates and collision efficiencies of three-body reactions. The high-O2 low-temperature region is of interrest not ouly to lean-premixed combustion, but also to flame ignition, and requires further exploration.

Effect of potassium hydroxide on PAH formation during toluene incineration

A laboratory-scale liquid-injection incinerator was used to incinerate toluene at 750°C and PAH emission was analyzed. The results indicate that the emission of solid-phase PAHs considerably increased when potassium hydroxide was added into a fuel-lean toluene flame; more potassium hydroxide resulted in more solid-phase PAHs. But this phenomenon was not observed for the fuel-rich toluene flame. The enhanced PAH formation is proposed as by-product of the partial oxidation or incomplete burning of soot in the secondary flame zone of the fuel-lean flame. It is suggested that reducing soot formation via potassium hydroxide addition may increase PAHs.

Evidence for a New Way of Producing NO via NNH in Fuel-Rich Flames at Atmospheric Pressure

Concentration profiles of NO X have been measured along the axes of laminar, premixed, flat flames of H 2 + O 2 + N 2 and CH 4 + O 2 + N 2 burning at 1 atm, with final flame temperatures in the range of 2500 to 1850 K. While the Zel’dovich mechanism [1]explains the rate of production of NO in the burnt gases of the fuel-lean flames studied, it was found that, downstream of the reaction zone in fuel-rich flames, NO is produced faster than predicted by the Zel’dovich mechanism alone, especially in flames cooler than 2100 K. Careful examination of the observations suggests there is an additional mechanism forming NO in the burnt gases of these fuel-rich flames. The route suggested by Bozzelli and Dean [2]via NNH as an intermediate, i.e., N 2+H=N 2H (1) N 2H+O→NO+NH, (2)accounts, both qualitatively and quantitatively, for the downstream measurements of the rate of generation of NO exceeding the rate of the Zel’dovich mechanism in these flames. The measurements here indicate that for reaction (1) at ∼2000 K: ΔH 1 0 = 23 ± 7 kJ mol −1 and ΔS 1 0 = −94.4 ± 2.0 J mol −1 K −1. Reaction (2) has a constant rate coefficient equal to 2 × 10 −10 ml molecule −1 s −1, to within a factor of 2.5, from 1800 to 2500 K.

HCO concentration in flames via quantitative laser-induced fluorescence

Quantitative laser-induced fluorescence (LIF) measurements of the concentration of HCO are made in three 25-torr methane-oxygen-nitrogen flames: fuel lean ( φ =0.81), near stoichiometric ( φ =1.07), and fuel rich ( φ =1.28). LIF is excited in the (000)-(000) band of the B-X system near 258 nm. The LIF signal from the flame is calibrated against nonflame measurements of a known HCO concentration produced by laser photolysis of acetaldehyde. The LIF signal is adjusted for the variation in the fraction of the population of the laser-excited level as the measured temperature changew with position in the flame and for the measured variation in quenching. The resulting concentration measurements agree well with model predictions for the fuel-lean and near-stoichiometric flame. The measurements in the fuel-rich flame are significantly larger than the model predictions: however, these measurements are subject to increased uncertainty due to the large, broadband background in the rich flame.

Nitric oxide formation and reburn in low-pressure methane flames

Absolute NO concentrations are measured in low-pressure methane-oxygen-nitrogen flames of varying stoichiometry, using laser-induced fluorescence. These measurements are combined with previously measured absolute CH radical concentrations to test the chemical mechanism of prompt NO formation and NO reburn. The GRI-Mech 2.11 chemical mechanism underpredicts prompt NO by 35–60%. Increasing the rate coefficient for the CH+N2 reaction brings the predicted [NO] into agreement with the measurements. Reburn chemistry is studied by seeding NO at concentrations between 0.1% and 3% in rich and lean flames. The model agrees with the observed reduction of NO in a fuel-lean (=0.80) flame and underpredicts NO reburn in a fuel-rich (=1.27) flame. The model predicts significantly more reduction of the peak CH concentration by reburn of seeded NO in the rich flame than is measured.

SHORT COMMUNICATION

The response of laminar premixed fuel-lean methane-air counterflow flames to the combined effects of stretch and cooling by a fine spray of water and NaCl-water as well as surfactant-water solution is studied experimentally. Results show that flame extinction is promoted both by increased flame stretch and by a higher concentration of liquid. An unexpected finding is that at a given concentration an NaCl-water solution is significantly more effective than pure water in causing flame extinction. On the other hand, at the concentration studied, results from a spray of water containing a surfactant ∥Synperonic PE/L62) are indistinguishable from those with pure water.

Flame-ion chemistry of the lanthanide metals Ce, Pr and Nd

A pair of premixed, H 2O 2Ar flames of fuel-rich (FR) and fuel-lean (FL) composition, both at atmospheric pressure and 2425 K, were doped with about 10 −6 mol fraction of the lanthanide metals La, Ce, Pr and Nd; from a previous study, La was used as a benchmark. The metals produce solid particles in the flames and gaseous metallic species. The latter include metallic atoms A near the flame reaction zone, but only the monoxide AO, the oxide hydroxide OAOH and, in some cases, the dioxide AO 2 further downstream at equilibrium. Metallic ions (< 1% of the total metal) were observed by sampling the flames through a nozzle into a mass spectrometer. All of the observed ions can be represented by four hydrate series: (a) major signals of AO +· nH 2O ( n = 0–3) for La, Ce, Pr and Nd; (b) small signals of AO 2H +· nH 2O ( n = 0–2) for Ce, Pr and Nd; (c) still smaller signals of AO 2 +· nH 2O ( n = 0, 1) for Ce, Pr and Nd in the FL flame only; and (d) tiny signals of AOH +· nH 2O ( n = 0, 1) for Pr and Nd in the FR flame only. The actual structures of some of these ions may not correspond to simple hydrates: e.g. AO +·H 2O = A(OH) 2 + = protonated OAOH; AO 2H +·H 2O = A(OH) 3 +, etc. Since hydrogen flames contain essentially no natural ionization, a major objective was to consider probable ionization mechanisms for the metals. The primary reactions include both chemi-ionization, and thermal (collisional) ionization of AO whose ionization energy is low (about 5 eV). Some of the ions are formed by secondary ion/molecule reactions including three-body hydration, proton transfer, electron (charge) transfer, H atom abstraction by radicals and oxidation. In addition, the chemical ionization of the metallic species by H 3O + was investigated. The flame-ion chemistry of these metals is discussed in detail.

Ion chemistry of second-row transition metals in hydrocarbon flames: cations and anions of Y, Zr, Nb, and Mo

A pair of laminar, premixed, CH4–O2 flames above 2000 K at atmospheric pressure, one fuel-rich (FR) and the other fuel-lean (FL), were doped with ~10−6 mol fraction of the second-row transition metals Y, Zr, Nb, and Mo. Since these hydrocarbon flames contain natural ionization, metallic ions were produced in the flames by the chemical ionization (CI) of metallic neutral species, primarily by H3O+ and OH− as CI sources. Both positive and negative ions of the metals were observed as profiles of ion concentration versus distance along the flame axis by sampling the flames through a nozzle into a mass spectrometer. For yttrium, the observed ions include the YO+•nH2O (n = 0–3) series, and Y(OH)4−. With zirconium, they include the ZrO(OH)+•nH2O (n = 0–2) series, and ZrO(OH)3−. Those observed with niobium were the cations Nb(OH)3+ and Nb(OH)4+, and the single anion NbO2(OH)2−. For molybdenum, they include the cations MoO(OH)2+ and MoO(OH)3+, and the anions MoO3− and MoO3(OH)−. Not every ion was observed in each flame; the FL flame tended to favour the ions in higher oxidation states. Also, flame ions in higher oxidation states were emphasized for these second-row transition metals compared with their first-row counterparts. Some ions written as members of hydrate series may have structures different from those of simple hydrates; e.g., YO+•H2O = Y(OH)2+ and ZrO(OH)+•H2O = Zr(OH)3+, etc. The ion chemistry for the production of these ions by CI in flames is discussed in detail. Keywords: transition metals, ions, flame, gas phase, negative ions.

Laser-induced fluorescence measurements of NO and OH mole fraction in fuel-lean, high-pressure (1–10 atm) methane flames: Fluorescence modeling and experimental validation

A method for quantifying laser-induced fluorescence (LIF) signals and planar laser-induced fluorescence (PLIF) images of the OH and NO radicals in high pressure flames is presented. The fluorescence signal per unit radical mole fraction is modeled as a function of temperature, pressure, overall flame stoichiometry and laser spectral bandwidth. A recently developed model ( JQSRT, 51, 511; Appl. Phys. B, 57, 249) 1,2 for electronic quenching cross-sections of OH and NO is utilized to express the fluorescence yield as a function of these parameters. The models are confirmed using single-point measurements in the burnt gas region of a flat flame burner at up to 10 atm. The measurements are performed at points in the flame where the temperature, pressure, OH and NO mole fraction are all known. For fuel-lean flames at elevated pressure, interference from the O 2 Schumann-Runge system was found with NO A ← X(0, 0) fluorescence measurements. This interference must be considered when selecting an appropriate NO transition in this type of environment.

Oxidation of carbon black particles in a premixed flame under pressure

A dilute stream of carbon black particles was introduced into the postflame gases of premixed, high-pressure, both fuel-lean and rich ethylene-air laminar flames. The rate of oxidation of carbon black particles was measured using scattering and absorption techniques to determine the particle size, concentration, and number density as a function of distance and, hence, burn-off. The reactivity of the carbon black was observed to be strongly dependent upon the total pressure in fuel-lean flame, while no such dependence was noted for rich flames.

Ionisation in premixed fuel-lean flames of H2, O2and N2. Part 1.—Naturally occurring positive ions

Mass spectrometric measurements have been made of the axial concentration profiles of the positive ions in six fuel-lean flames of H2, O2 and N2, incorporating no additives. The principal positively charged species is the hydronium ion, H3O+, whilst the sole negative species is the free electron, e–. The reaction H + H + OH ⇌ H3O++ e–(1) is found to be responsible for establishing H3O+ and e– concentrations close to their values for equilibrium well downstream in the burnt gases. The equilibrium constant, K1, is measured to be (2.16 ± 0.80)× 10–11× exp(–11 591/T) atm–1. The forward step of reaction (1) determines the initial rate of formation of H3O+, and has a rate coefficient, k1, of (3.3 ± 1.6)× 10–36 exp(–13 800/T) cm6 molecule–2 s–1. The activation energy for the forward step of reaction (1) is equal to the endothermicity of reaction (1), namely 97 ± 10 kJ mol–1. These expressions for k1 and k1 yield a rate constant for the reverse reaction (–1), of 1.9 ± 2.3 × 10–7 cm3 molecule–1 s–1, approximately independent of temperature. The proton affinity of water in the gas phase H2O + H+→ H3O+ has also been derived to be 674 ± 14 kJ mol–1, at 0 K.

Atomization efficiencies in halocarbon-loaded acetylene-air flames—I. Monochlorides of readily volatilized elements

Carbon tetrachloride vapour was introduced with a carrier air stream into the mixing chamber of an acetylene-air flame while nebulizing aqueous solutions of several metal salts. The atomic absorption signal was measured under increasing flow rate of the halocarbon vapour at constant fuel-to-oxidant ratio, and the latter parameter was also varied in separate experiments. By applying the theory developed by Sugden and Bulewicz, the exclusive formation of monochlorides in the gaseous phase (presented here, in Part I) and the additional formation of hydroxychlorides and dichlorides (presented in Part II) could be elucidated. From the decrease of the signal measured for the alkali elements of known monochloride dissociation constants and dissociation energies, the temperature and the chlorine concentration present in the observed flame zone could be calculated. It is inferred that only 19% of the total halogen introduced is converted to HCl and Cl species in a slightly fuel-rich flame. The signal depression is stronger in a fuel-lean flame of higher monoatomic chlorine concentration for those elements which have a relatively efficient atomization under these flame conditions. An increase of the electron concentration resulting from the introduction of the halocarbon in an alkali-containing flame was deduced from the experimental findings, in agreement with earlier observations.

Aspects of laminar premixed flame extinction limits

Numerical methods have been used to examine (1) the combined effects of stretch and upstream heatloss on the properties of a stoichiometric, laminar, pre-mixed methane-air flame, and (2) the combined effects of stretch and an extended, radiative heat loss on the properties of a fuel-lean methane-air flame, equivalence ratio Σ=0.57. Appropriate opposed flow configurations were investigated in both cases. In the first case, fresh gas was discharged towards the stagnation plane from identical, coaxial, plug flow nozzles maintained at constant temperature and separated by a fixed distance. Continuation methods were used in tandem with Newton solutions to study the flame behaviour as the stretch rate (or nozzle inlet velocity) was varied. High and low stretch extinction limits were observed as positions of vertical tangency on the steady-state solution curves of flame property vs stretch rate. The complete series of solutions between these positions formed a limit cycle with both stable and unstable branches. The chemical behaviour at the limits and along the unstable branch is discussed. The effect of radiative losses on the fuel-lean flame was studied in the unburnt-to-unburnt (UTU)configuration as above, but also in the unburnt-to-burnt (UTB) configuration where the ignited unburnt stream is opposed by a stream of its own equilibrium combustion products. If the flame is adiabatic, the product stream is input at the adiabatic equilibrium flame temperature. Such a system shows no abrupt extinction limits. On the other hand, if the radiative “sink” temperature is below some critical value, then, irrespective of the magnitude of the heat loss rate on a volumetric basis, the unburnt-to-burnt system will show abrupt extinctions if the stretch rate is increased sufficiently. Because of extremely slow numerical convergence, the numerical investigation of these limits for the methane-air system is not yet complete. However, such an investigation is in progress, and this combination of stretch and radiative loss effects is clearly of importance in relation to behaviour at composition limits of flammability.