Tubular Burner Research Articles

Extinction and cellular instability of premixed tubular flames

Extinction and cellular instabilities are investigated in premixed tubular flames. A tubular burner has recently been designed and constructed to be capable of operating under a wide range of conditions that allow more accurate and detailed study of tubular premixed and non-premixed flames. Extinction measurements of tubular flames can be carried out through the entire equivalence ratio spectrum for various types of fuels without introducing serious turbulent effects. In this study, the extinction conditions of premixed tubular flames generated by H2/O2 or CH4/O2 mixed with various inert diluents (Ar, CO2, N2) and enriched oxygen are measured. In addition, the onset conditions for flame instability at different cell numbers are reported. Petal shaped cellular flames with symmetric regions of local extinction are observed for lean H2 tubular flames with low Lewis numbers. Wrinkled flames with no local extinction are observed for nearly equidiffusive lean CH4 tubular flames. The combined effects of curvature, diffusive-thermal instability and hydrodynamics are thought to be responsible for the cellular structure. The lack of extinction in the nearly equidiffusive lean CH4 tubular flames indicate that strong diffusive-thermal effects due to non-unity Lewis number are required to produce local extinction in the cellular structure of tubular flames.

Phenomenology of methane flame propagation into compositionally stratified, gradually richer mixtures

Flame propagation in compositionally stratified, rich methane air mixtures was investigated experimentally. Compositional stratification was established in a counterflow, tubular burner employing very mild strain and the equivalence ratio in the unburnt mixture was measured with line-Raman imaging of methane. Particular emphasis was placed in the determination of differences with stratified combustion in lean mixtures that was considered in earlier work. It was established that departure from “quasi-homogeneity”, i.e. from flame propagation that can be viewed as a series of flamelets corresponding to the local equivalence ratio, was more pronounced in most cases of rich mixtures than in lean combustion. In certain cases, we were able to record flame speeds in stratified media that were smaller than the speeds of “quasi-homogeneous” propagation, something that was never observed in lean stratified combustion. A significant extension of the rich flammability limit was observed in stratified mixtures and flames propagated all the way into mixtures of equivalence ratio equal to 1.7, which may point to a possibility of hydrogen production through flame propagation into “super”-flammable mixtures. Also, in the rich mixtures, the propagating flame fronts did not develop the instabilities that were observed in lean mixtures in the vicinity of the flammability limit. This was rationalized based on well-established theoretical results on flame stability.

Laminar premixed methane/air flame extinction characteristics influenced by co-flow water mists

Based on the tubular burner, the burning velocities, flame stretch and inhibition rules influenced by co-flow water mists were studied using a high-speed schlieren system. Moreover, the variation rules of the flame critical extinction in our burner equipment were also obtained by analyzing the process and mechanism of flame extinction and inhibition. It is shown that the flame stretch is related to the fuel concentration, co-flow fluxes and water mist diameters. For droplets with a larger diameter, the smaller the co-flow fluxes, the more obvious the flame stretch. When the water mist loading rate is rather smaller, for fuel-rich premixed flame with Le>1, the flame with larger burning rate tends to backfire more easily. Under the same water mist conditions, for fuel-lean premixed flame with Le<1, the smaller the gas concentration, the easier the flame is extinct.

An experimental study of the lower flammability limit of LPG/halocarbon mixtures using the tubular burner apparatus

The motivation for this project is to develop a hydrocarbon-based (ER12) refrigerant that has a lower flammability limit in excess of 5% refrigerant in air. Optimisation of the hydrocarbon content in the mixture, favourable refrigeration properties, minimal toxicity and benign environmental impact form additional constraints. A variety of compounds were considered as potential flammability reducing additives. The flammability of the various mixtures was evaluated using the tubular flame burner technique. The compounds tested in the initial screening study included CO2, CF3I, CHF2Cl (R22), C3HF7 (R227ea), C3F8 (R218), C4F10 (R3110), C4F8O, SF6, CBr2F2, 1-bromopropane, 2-bromopropane, CH2Br2 and CH2BrCl. Strong flammability reduction performance was observed for CBr2F2, CH2Br2, CF3I, C4F10, SF6 and C3F8. Less effective compounds included CO2, C3HF7 (R227ea), CHF2Cl (R22), C4F8O, and CH2BrCl. Estimation of potential non-flammable ER12/additive compositions based on the flammability tests revealed the best additives to be CBr2F2, CH2Br2 and CH2Br2 each with an ER12 content above 20 mass percent. It is concluded that a practical single additive/ER12 mixture with higher than 20% ER12 content is unlikely. Compounds containing bromine and iodine possess superior chemical suppression effectiveness with the strength of that effect increasing in proportion to the level of bromination/iodination. Highly fluorinated compounds have excellent suppression efficiency due to the relatively low atomic weight of fluorine. The presence of hydrogen in an additive compound dramatically reduces the suppression efficiency. Inclusion of oxygen and sulphur in compounds can potentially give very good flammability reduction effects. The major environmental and parameters of atmospheric lifetime, ozone depletion potential (ODP) and global warming potential (GWP) were compiled for the screened and targeted additive compounds, with some of the more favourable species shown to possess relatively high atmospheric lifetime and global warming potential despite having zero ozone depleting effect. The toxicological and safety aspects of the compounds are also discussed as an additional means of evaluation of potential additives.

Experimental and numerical investigation of non-premixed tubular flames

Non-premixed tubular flames are established using the opposed tubular burner for the first time, which enables the quantitative study of curvature effects on flame behavior. A detailed structural investigation using the spontaneous Raman scattering technique is conducted on flames with constant curvature using 15% H 2 diluted by N 2 against air at various stretch rates. The measured temperature and major species concentrations agree well with the numerical prediction, and this validates the numerical model. Comparing with the numerical results of the opposed-jet flat flame, the tubular flame data show that the curvature weakens the effects of preferential diffusion for the diluted hydrogen flames where the curvature is concave towards the fuel stream. To further prove this discovery, near-extinction non-premixed tubular flames using different types of fuels are also studied. Hydrogen flames ( Le < 1) automatically change their orientation of curvature at high stretch rates and cellular structures are observed. Methane flames ( Le ≈ 1) do not show any preference on the direction of curvature before extinction is reached and no cellular structures are observed. Propane flames ( Le > 1) prefer concave curvature, which is opposite to the hydrogen flames, and cellular structures appear throughout the burner’s operational range. These phenomena confirm that for flames with Le < 1, convex curvature promotes combustion, enhances the flame temperature and therefore retards extinction; vise versa for flames with Le > 1. Curvature has minimal effects on tubular non-premixed flames with Lewis number close to unity. The comparison between experimental and calculated extinction conditions also supports the above conclusions.

Stretch rate of tubular premixed flames

Following Seshadri and Williams solution describing the flow field for the opposed jet burner, the analytical solution is given for the flow field of two other burners: the opposed tubular burner and the tubular burner. Under plug flow boundary conditions, it is shown that the stretch rate at the stagnation surface of the opposed tubular burner is k = π V / ( R 2 − R 1 ) for the case of equal velocities and equal densities (i.e., ρ 1 = ρ 2 and V 1 = − V 2 = V ). For the tubular burner, the stretch rate at the center of the burner is k = π V / R 2 . The comparison of the numerical simulation and analytical solution is carried out to verify the analytical solution.

Flame stabilization with a tubular flame

A new technique to stabilize a flame in a high-velocity stream with use of a tubular flame has been proposed. To elucidate the validity of this technique, an experiment has been conducted by mounting a tubular flame burner on the nozzle. Flame stability limits and temperature distributions around the burner port have been determined, and experiments have been extended to the ducted combustion to measure pressure fluctuations and to analyze the burned gases. Results show that the tubular flame can successfully stabilize the main flame up to 130 m/s, which is the upper limit of the present supply facility. The main flame is well anchored at the exit of the nozzle, and the tubular flame efficiently supplies heat and radicals to the main flame. In the ducted combustion, the pressure fluctuations are reduced significantly. The exhaust gas analyses, however, indicate that an almost chemical equilibrium condition can be achieved at 50 m/s, but not at 90 and 130 m/s. Since the energy input relative to the main flame is just 6.1% at 130 m/s, the present tubular burner is not enough to burn all the unburned gas completely at high velocities, although the main flame can be anchored. The slit length and/or the slit width of the tubular flame burner should be larger to overcome this shortage. From the above results, it is concluded that the tubular flame has a potential for stabilizing a flame in a high-speed stream.

Thermal Stability of Flame-Synthesized Anatase TiO2 Nanoparticles

The thermal stability of anatase TiO2 nanoparticles, produced by flame synthesis, is investigated in the current study. Phase-pure anatase particles of ∼4 nm in size can be reproducibly synthesized by using a tubular burner and rotating sampler. Transmission electron microscopy (TEM), glancing incidence X-ray diffraction (XRD), and near-edge X-ray absorption fine structure (NEXAFS) were utilized to characterize the particles before and after annealing to various temperatures. TEM investigations reveal a primary particle size of ∼4 nm with standard deviations around 1 nm. XRD and selected area electron diffraction (SAED) indicate that TiO2 nanoparticles are phase-pure anatase. Our results indicate that the particles remain kinetically trapped upon annealing up to 773 K in air for 2 h. At 973 K, increases in average size, rutile content, and particle shape are observed, consistent with recent reports in the literature. NEXAFS measurements indicate that the O K-edge features of the nanoparticles show similar...

Experimental and numerical investigation of premixed tubular flames

Fundamental flame response in a stretched and curved flow field is investigated in a unique opticallyaccessible tubular burner. Time-averaged, one-dimensional spatially resolved temperature and major species measurements are obtained in a set of stretched, φ =0.175 premixed H 2 /air tubular flames using visible laser-induced Raman spectroscopy. The very lean H 2 /air flames are formed under relatively high stretch rates, 90≤ κ ≤215 s −1 , with 227 s −1 being the extinction condition. In tubular flames, both stretch and thermal-diffusive effects are dependent on both aerodynamic straining and flame curvature. Thermal-diffusive effects in highly curved (∼2 mm flame radius) tubular flames significantly influence the flame structure, leading to flame temperature increases of ∼120 K over the planar unstretched flame temperature (∼1180 K). The standard program for modeling stretched planar flames (Oppdif) is modified for the cylindrical geometry of the tubular flame. Comparisons of the Raman measurements with numerical simulations for tubular premixed flames, using complex chemistry and detailed transport properties, show excellent agreement at low rates of stretch (i.e., κ ≤127 s −1 ). At higher flame stretch, hence increased curvature, numerical simulations using the currently available transport data and chemical kinetic mechanisms incorrectly predict the flame structure. The experimental observations show extinction occurring (∼227 s −1 ), while numerical simulations overpredict the extinction limit (∼750 s −1 ). Simulations using four different H 2 /air chemical kinetic mechanisms show that the flame structure is very sensitive to the particular mechanism and the molecular diffusion coefficients. Evaluation of molecular diffusion coefficients indicates that the thermodiffusive properties of the deficient reactant species, H 2 , strongly affect the tubular flame structure. Thus, the products in the flame zone ( φ =0.34) are enriched from the initial reactant mixture ( φ =0.175) by flame curvature and the rapid diffusion of H 2 .

Experimental and computational study of the effect of C2F6 on premixed methane flames

The objective of this work was to specify the mechanism of flame inhibition by compounds that do not contain bromine. Mole fraction profiles have been measured by molecular beam-mass spectrometer technique in a stoichiometric CH 4 /O 2 /Ar low-pressure flame seeded with and without 1% hexafluoroethane (C 2 F 6 ). For both flames, stable and reactive species concentrations have been computed and compared with experimental results in order to validate a detailed mechanism. Extension of the mechanism to atmospheric pressure was carried out by substituting the rate constant of a few reactions by high-pressure values. Mole fraction profiles computed in 1-atm-pressure stoichiometric methane-air flame were in very good agreement with Bechtel's experimental results. Thus validated, the mechanism was used to compute the reduction in flame velocity that results from the addition of increasing amounts of the inhibitor to a methane-oxygen-nitrogen flame. Calculated flame velocities are in very close agreement with experimental values obtained with a premixed flame stabilized on an open tubular burner. This shows that inhibition efficiency of fluorinated compounds can be predicted very accurately.

Methanol as a Soot Reducer in a Turbulent Swirling Burner

The combined effect of using methanol as a fuel additive together with a prototype multifuel injector has been evaluated with regard to soot formation in a tubular laboratory burner with a turbulent swirl stabilized diffusion flame. Kerosene, ERBS fuel, and Blending Stock with approximately 14, 12.8, and 10.3 weight percent of hydrogen, respectively, were characterized in terms of soot loading at the axial positions Z/D = 2.5 and 4.0 and normalized radius r/R = ± 0.67. Mixtures of ERBS fuel and Blending Stock with 15 and 7.5 weight percent of methanol were also characterized in the same way. Measurements with the plain fuels showed a drastic reduction in soot formation, on the order to 150-fold decrease, due to the new injector design. Further reductions by a factor of 2 and 1.5 were accomplished with mixtures of 15 and 7.5 weight percent of methanol, respectively.

Experimental study of the thermal–acoustic efficiency in a long turbulent diffusion‐flame burner

A two-year study of noise production in a long tubular burner is described. The research was motivated by an interest in understanding and eventually reducing core noise in gas turbine engines. The general approach is to employ an acoustic source/propagation model to interpret the sound pressure spectrum in the acoustic far field of the burner in terms of the source spectrum that must have produced it. In the model the sources are assumed to be due uniquely to the unsteady component of combustion heat release; thus only direct combustion-noise is considered. The source spectrum is then the variation with frequency of the thermal-acoustic efficiency, defined as the fraction of combustion heat release which is converted into acoustic energy at a given frequency. The thrust of the research was to study the variation of the source spectrum with the design and operating parameters of the burner.

Soot Formation in a Turbulent Swirl Stabilized Laboratory Combustor

The effect of pressure and fuel composition on soot formation was studied using a high pressure, tubular laboratory burner with a swirl stabilized diffusion flame. A specially-designed water flushed probe was used to extract samples of soot and other combustion products for analysis. Measurements showed the expected increase of soot concentration with pressure only until a critical value of the Reynolds number was reached. After this point, no further increase of soot loading was observed with pressure rise. This behavior was found to be present when either kerosene or propane was used in the burner. At a suitably high Reynolds number, several fuels, closely related to JP-4, were tested in the burner. These fuel blends were constructed by adding cyclic compounds to the JP-4 so as to change its hydrogen-to-carbon ratio. It was found that the amount of soot formed scaled with this ratio, although the sooting effect of different blends with the same ratio were not identical.

Effect of Turbulence on Pollutant Formation in a Tubular Burner

Effect of Turbulence on Pollutant Formation in a Tubular Burner

The nucleation of flame carbon by ions and the effect of electric fields

Applied electric fields have been shown to affect all the three phases in the process of carbon formation in diffusion flames: nucleation, growth in the pyrolysis zone, and deposition. It is possible to resolve these effects by using different burner and electric-field configurations and it is necessary to do so before conclusions regarding nucleation can be drawn. With a tubular burner as one electrode and a steam-cooled plate above it as the other it is found that the particles are positively charged, i.e., deposition occurs mainly on whichever is the negative electrode; also that the form of aggregate greatly increases in volume but not in mass with the current density. Using a counterflow diffusion flame system with the burner mouths as electrodes, it is possible to draw ions of either sign from the flame through the pyrolysis zone. The extent of growth of particles as determined by their size and mass is controlled entirely by their period of residence in the pyrolysis zone and therefore by the field strength, since they acquire their charge very early in their history. The mechanism of charging is discussed. Regarding nucleation, the number rate of generation of particles depends on both the sign and magnitude of the ion flux through the pyrolysis zone. It is concluded that positive ions, either flame ions or Cs + ions produced by the thermal ionization of CsCl added to the flame, act as nuclei for the formation of carbon particles. Negative charges do not appear to act as nuclei but tend to reduce carbon formation by neutralizing positive-ion nuclei.

Dynamic characteristics of solid propellant combustion

This paper describes the experimentally determined response of a solid propellant combustion zone to normally incident, one-dimensional pressure waves. A self-excited, side-vented, tubular burner in which end-burning grains are used is the experimental vehicle. During a firing in this burner, the early growth period of the oscillations and also the decay are exponential with time. The exponential growth and decay constants are used to calculate the acoustical admittance of the combustion zone and these admittances are reported as a function of pressure (200 to 1,600 psi) and frequency (500 to 3,500 cps) for four different propellants. The experimental results are also reported in terms of the fractional increase in energy content of a wave reflecting from a combustion zone. Because the oscillations have exponential growths and decays, observations made during these portions of a firing might be expected to have relevance to a theory, such as that of Hart and McClure, which is mathematically linear. Accordingly, the experimental results are compared in a tentative and general way with this theory and the comparison shows that the experimental admittances are somewhat larger than the theoretical predictions. In the treatment of the experimental data, the mean flow field in the burner is neglected, and the rather questionable assumption is made that the damping of the oscillations in the burner is the same during the decay period (after the propellant is consumed) as it was during the rise. Because of these approximations, the results must be regarded as a little uncertain at this time.