Showing papers in "Combustion Science and Technology in 1984"

Lift-Off Heights and Visible Lengths of Vertical Turbulent Jet Diffusion Flames in Still Air

Abstract: The lift-off heights and visible-flame lengths of jet diffusion flames in still air have been determined for hydrogen, propane, methane and ethylene. The flame lift-off height varies linearly with the jet exit velocity and is independent of the burner diameter for a given gas. The results support the assumption that if the burner exit flow is choked the burner can be approximated by an equivalent convergent-divergent nozzle at whose exit the flow has expanded to ambient pressure. The data for different gases can be collapsed onto a single curve if they are plotted in terms of the appropriate non-dimensional groupings. These results and previous results for blow-out stability suggest that diffusion flames blow out when the base is lifted to between 0.65 and 0.75 times the height at which stoichiometric concentration is reached at the jet axis. It can be deduced from the experimental results that, at the base of the flame. the ratio of turbulent burning velocity to laminar burning velocity varies a...

The Structure and Concentration Decay of High Pressure Jets of Natural Gas

Abstract: The concentration field in compressible sonic jets of natural gas over the pressure range from 2 to 70 bar has been studied using gas chromatography. At these pressures the jet expands immediately downstream of the nozzle in order to equilibriate with ambient conditions. The behaviour of such underexpanded jets is shown to be similar to classical free jets provided that an appropriate scaling factor is employed to describe the effective size of the jet source. This notional source is referred to as the pseudo-diameter and a theroretical analysis for its derivation is presented. The experimental data support this approach, and the decay constants for the jet concentration field agree with previous measurements in incompressible jets. Comparison with results obtained from a high pressure jet of ethylene at 8 bar indicate that the effect of jet fluid density in concentration decay is the same as in classical free jets.

Flamelet Crossing Frequencies and Mean Reaction Rates in Premixed Turbulent Combustion

Abstract: Abstraet–The one-point, one-time description contained in the Bray-Moss-Libby model of premixed turbulent combustion is first generalized to a two-point, two-time formulation which includes information on the time and length scales of the scalar field within turbulent flames. This formulation is then specialized to a one-point, two-time description which is treated in detail so as to yield expressions for the autocorrelation of the progress variable, for the time scale of the scalar field, for the mean crossing frequency of the flamelets and, finally, for the mean rate of chemical reaction. Although the latter expression closely resembles results developed earlier on an intuitive basis, its derivation in the present study permits an assessment of various assumptions and intermediate results by comparison with experimental measurements. Such comparisons as are currently possible are shown to lend support to the analysis and to enhance the prospects of future exploitation of the general formulation.

A Comprehensive Chemical Kinetic Reaction Mechanism for Oxidation and Pyrolysis of Propane and Propene

Abstract: Abstract—A detailed chemical kinetic reaction mechanism is developed to describe the oxidation and pyrolysis of propane and propene. The mechanism consists of 163 elementary reactions among 4l chemical species. New rate expressions are developed for a number of reactions of propane, propene, and intermediate hydrocarbon species with radicals including H, 0, and OH. The mechanism is tested by comparisons between computed and experimental results in shock tubes and the turbulent flow reactor. The resulting comprehensive mechanism accurately reproduces experimental data for pressures from 1 to 15 atmospheres, temperatures from 1000 to 1700 K, and fuel-oxidizer equivalence ratios from 0.066 to pyrolysis conditions. The mechanism also predicts correctly laminar flame properties for propane and propene, and detonation properties for propane.

Who do Diffusion flames Emit smoke

Abstract: The causes of smoke emission from free gaseous laminar diffusion flanles are investi-gated. Soot concentration, temperature and velocity profiles were measured around the smokepoint for a range of fuels. Smoke was emitted when the temperature in the burnout region droppedbelow about 1300 K.A smoking flame stopped emitting smoke when the temperature in the burnout region wasslightly raised locally by absorption of energy from a COz laser beam.The temperature drop in the flame is strongly coupled to soot concentration through radiationlosses. Soot formation rates are in turn influenced by heat losses and residence times near thefuel nozzle tip. Thus conditions near the nozzle influence the smoke point. The results are alsodiscussed in context with turbulent diffusion flame smoke emission.

Entrainment in the Near and Far Field of Fire Plumes

Abstract: This paper describes entrainment measurements made in fire plumes with a new technique. Measurements were in plumes rising from natural gas diffusion flames stabilized on 0.10, 0.19 and 0.50 m diameter burners and the heat release rates ranged from 10 to 200 kW. The heights examined ranged from elevations starting very close to the burner surface to distances about five times the average flame heights. Experiments indicate the presence of three regions: a region close to the burner surface where plume entrainment rates are independent of the fuel flow (or heat release) rates; a far field region above the flame top, where a simple point source model correlates the data reasonably well; and an intermediate region where entrainment appears to be similar to that of a turbulent plume.

Sooting Correlations for Premixed Flames

Abstract: —Using a Bunsen-type burner in which the name temperature was controlled by nitrogen addition, the tendency of hydrocarbon fuels to form soot has been determined systematically. The effective sooting equivalence ratio reported is based on the stoichiometric relation in which the carbon and hydrogen in the fuel are converted to carbon monoxide and water. respectively, rather than the conventional equivalence ratio or the C/O ratio. The temperature effect is clearly shown by plotting the log of the effective equivalence ratio versus the reciprocal of the (adiabatic) flame temperature, and this dependency is similar for all fuels examined and is relatively weak. At a given flame temperature, an increase in either the number of carbon atoms or the C/H ratio in the fuel molecule increases the tendency to soot. These effects are attributed to the increased pyrolysis rate and the decreased oxidative attack, respectively. No significant effects are found with regard to the structure of the fuel molecule....

Estimation of Soot Thresholds for Fuel Mixtures

Abstract: Experimental soot threshold results obtained using binary and ternary hydrocarbon mixtures were analyzed to obtain two empirical rules for estimating the onset of soot formation for fuel blends in atmospheric pressure, laminar premixed and diffusion flames. The results suggest that the onset of soot formation for a fuel mixture of known composition can be accurately calculated from soot threshold data for the pure compounds.

Some Experimental Aspects of Turbulent Diffusion Flames and Buoyant Plumes from Fire Sources Against a Wall and in a Corner of Walls

Abstract: Measurements of temperature distribution and visible flame height were made on the fire plumes from square burners in a semi-infinite space, against a wall and in a corner of walls, Flame height is formulated as a function of Q * = Q/poC p T og3/4 D 3/4;52 and the maximum temperature on each height was formulated with the concept of virtual heat source. According to the comparison of the results on the fires against a wall and, n a corner of walls with those in a semi-infinite space, the existence of walls near a fire source is thought to suppress the growth of eddy scale in the plume. The characteristics of natural fuels as sources of turbulent diffusion flame are also discussed.

Soot Particle Growth in Premixed Toluene/Ethylene Flames

Abstract: Surface growth of soot was examined in several premixed flat flames. The fuel was a mixture of toluene and ethylene, with more than half of the carbon supplied by the toluene. We found: (I) Acetylene in the burned gases supplied most of the mass for surface growth, and the growth mechanism in toluene/ethylene flames was the same as in the ethylene flames that we examined previously; i.e., the rate constants that we found in the earlier work predicted the surface growth rates in these flames. (2) In comparing ethylene and toluene/ethylene flames, the fluorescence intensity from polycyclic aromatic hydrocarbons (PAH) was very different for flames with the same C/O ratios but similar for flames with the same soot volume fraction. (3) For a given C/O ratio the amount of soot produced in the particle inception stage in the toluene/ethylene flames was much higher than in the ethylene flames. However, because the acetylene concentrations were higher in the ethylene flames, the rate of surface growth per...

Flame Spread Modeling

Abstract: A survey of the characteristics and predictions of the different theoretical models of the spread of flames over the surface of a solid combustible in opposed or concurrent oxidizing flows shows that, at present, there is a good understanding of what are the controlling mechanisms of the flame spread process andl of what is the necessary formulation to develop a rigorous analysis of the phenomenon. It also shows, however, that the problem is very complicated and difficult to solve mathematically particularly if an analytical solution is sought, and that this complexity is what has prevented so far the development of an analysis capable of describing accurately the flame spread process under realistic conditions where material properties, finite rate kinetics, turbulence and radiation effects can determine the characteristics of the process. Although some of the analyses presently available are capable of predicting quantitatively or at least qualitatively rates of flame spread under certain limit...

On the Initiation of a Spherical Flame Kernel

Abstract: Using matched asymptotic expansions for large activation energies of the reaction rates, we have studied the structure and stability properties of steady spherical premixed flames which, in this geometry, need not be supported by any source (or sink) of fresh (or burned) gases. The radius of such unsupported flames strongly depends on the differential diffusion of heat and reactant and is interpreted as a flame initiation criterion which corresponds to an unstable equilibrium. This radius must be exceeded for the flame to self-propagate outward. We then focus attention on two means to modify, or even suppress, such a critical flame size, viz.: (i) the effect of a central heal source of constant power, and (ii) the addition of small amounts of a hydrogen-like reactive species with large molecular diffusivity. Multiple steady solutions-including stable, non-propagating flames-or the disappearance of equilibrium solutions are exhibited as are qualitative changes in flame dynamics.

Sooting Counterflow Diffusion Flames with Varying Oxygen Index

Abstract: Varying oxygen concentrations was found to have pronounced effects on the sooting characteristics of diffusion flames. Ethylene and propane flames were stabilized around a porous ceramic cylindrical burner mounted in a vertically oriented, low turbulence wind tunnel. Laser light scattering and extinction, laser Doppler velocimetry and thermocouple temperature measure- ments were obtained along the forward stagnation line. Measurements of soot particle size, number density and volume fraction showed that with an increasing oxygen index the soot volume fractions in both ethylene and propane flames increased at all locations along the stagnation stream- line. For ethylene flames (when the oxygen index increased from 0.18 to 0.28). the maximum soot volume fraction increased by almost an order of magnitude. Propane was found to produce about one half the volume of soot of an ethylene flame at the same oxygen index. The increase in soot volume fraction with increasing oxygen index was attributed to bot...

Soret and Dilution Effects on Premixed Flames

Abstract: The structure of a wrinkled premixed flame is analysed theoretically. By assuming the reactive mixture diluted in an inert gas and a weak cross-diffusion coupling between the heat and mass fluxes, the effect of the change by the reaction of the physical gas properties (thermal conductivity, specific heat, number of molecules) and Soret and Dufour diffusions have been investigated in the limit of high activation energies. It turns out that Soret mass diffusion cannot be neglected in the study of the flame front dynamics. In contrast, the change of physical properties as well as Dufour heat diffusion generally appears to play a minor role.

Conditional Sampling of Turbulence Intensities and Reynolds Stress in Premixed Turbulent Flames

Abstract: Abstrac–-A two-color LDA technique has been developed for conditional sampling of turbulence intensity and Reynolds stress in premixed turbulent flame. To sample conditionally in the unburned reactants, silicone oil aerosol which burned and evaporated by combustion is used to seed the flow. Conventional AI2O3 seed particles are used to sample unconditionally through the flame. A conditional analysis method is developed to deduce conditioned data in the burned products from the unconditioned data. Comparison of the conditioned and unconditioned data obtained in a lean ethylene/air v-flame shows that the so-called "flame-generated turbulence" and the increase in Reynolds stress within the flame brush are caused primarily by intermittent measurements in the reactants and in the products. Using an intermittency model for the flame turbulence, the contributions due to intermittency are shown to be proportional to the difference between the conditioned mean velocities. The conditioned fluctuation inten...

Nonluminous Radiation in Turbulent Buoyant Axisymmetric Flames

Abstract: A theoretical and experimental study of nonluminous radiation properties of turbulent, buoyant, axisymmetric, methane diffusion flames burning in still air is reported. Spectral radiance was measured along radial paths within and near the luminous portion of three flames whose structure was studied earlier in this laboratory. Predictions of spectral radiance were found by solving the equation of radiative transfer along the same paths. A narrow-band model was used with flow structure predicted by the conserved scalar formalism of a k-e-g turbulence model. The computations were based on mean properties as well as a stochastic method which allowed for turbulent fluctuations. In general, the stochastic method overestimated the measured spectral radiances by 10-30 percent, while the mean property method was in reasonably good agreement with measurements. Therefore, while turbulent fluctuations have a significant influence on non-luminous radiation properties of the present flames, the effect is compa...

Strained Premixed Laminar Flames with Two Reaction Zones

Abstract: The method of activation energy asymptotics (AEA) is used to study the structure and extinction characteristics of premixed laminar flames in two counterflowing reactant streams such that a layer of products is bounded by two reaction zones. The analysis includes the influences of realistic density variations and of large rates of strain. Account is taken of Lewis numbers different from unity and of different temperatures in the two reactant streams. It is found that abrupt extinction occurs either upon merging of the two reaction zones or prior to merging depending on the Lewis number and on the temperature difference of the two streams. In the adiabatic case Lewis numbers suitably greater than unity, i.e., greater than a critical value, are found to be needed for extinction to occur prior to merging. The critical Lewis number decreases as the temperature difference increases such that at suitably large differences extinction prior to merging occurs for Lewis numbers less than unity. Density non...

Burke—Schumann Flame with Streamwise and Preferential Diffusion

Abstract: The classical diffusion flame model of Burke and Schumann is extended to include the effects of both streamwise and preferential diffusion. The study first identifies a Lewis-number-weighted coupling function as the proper one to usc in formulations involving non-unity Lewis numbers. A perturbation technique is then developed for the realistic situation of ncar-unity Lewis number flows. Results show that, depending on the effective Lewis numbers of fuel and oxidizer, the flame temperature may vary either monotonically or non-monotonically along the flame sheet, and thereby suggesting the possibilities of tip opening, and tip intensification, as well as partial and complete flame blow-off.

Conditionally-Sampled Velocity and Turbulence Measurements in a Spark Ignition Engine

Abstract: Laser Doppler velocimeter measurements have been made in a homogeneously-charged spark ignition engine. With ignition at the side wall of the disc-shaped combustion chamber, the fluid motion in the direction of flame propagation was measured at the center of the chamber. A simultaneous ionization probe measurement was used to identify the time of flame arrival at the velocimeter probe volume. Phase-averaged measurements recorded from many engine cycles were conditionally sampled according to flame arrival time. The results presented show an increase in the unburned gas turbulence from compression, and strongly suggest that cyclic variation in burn duration is caused by cyclic variation in the bulk turbulence intensity ahead of the flame.

Turbulent Diffusion Flames with Large Buoyancy Effects

Abstract: In order to contribute to the modelling of room fires, experimental results are presented concerning turbulent diffusion flames obtained with a 30 cm diameter burner and corresponding to initial Froude number F in the range 2 × 10–6 < F < 1.2 × 10–5, and calorific powers Q in the range 16 kW < Q < 38 kW. The results concern main species mole fractions, temperature and flow velocity. The spectral analysis of the signal obtained by laser velocimetry shows that the flow is periodic or quasi-periodic in the lower part of the flame, where the quasi-totality of the combustion is achieved. Although the k-e-g models imply a fully turbulent flow, the comparison of our experimental results with predictions based on the models developed by F. Tamanini shows that the Wallburn model gives a satisfactory approximation of the mean velocity distribution along the flame axis. However these models underestimate the lateral expansion of the flame.

A Perspective on Compartment Fire Growth

Abstract: A review was made of research related to fire growth in compartments. Mention is made of the zone and field model approaches that have been used to describe many aspects of compart- ment developing fires. Primarily the review is organized by phenomena associated with com- partment fires. These include fluid mechanic, heat transfer and combustion processes. Each phenomenon is discussed and work is presented to illustrate predictive techniques. Limitations and deficiencies in our understanding are discussed. A previously unavailable analysis of radiative transfer in an enclosure for a two-layer participating gas is also presented.

Determination of the Refractive Indices of Soot Particles from the Reflectivities of Compressed Soot Pellets

Abstract: Abstract-Measurements of the polarized components of the reflectivity of a compressed propane soot pellet are presented which demonstrate that, for wavelengths in the infrared, the reflection is specular and is described by the Fresnel equations By matching these equations to the full angular variations of the specular reflectivities, an effective refractive index of the soot/air mixture is determined at each wavelength To deduce the refractive indices of the particles themselves, theories for the optical properties of composites are considered and a new relationship based upon the effective medium theory is introduced The void fraction near the surface of the pellet must be known A technique is presented for measuring this voidage by performing image analyses of scanning electronmicrographs of the pellet surface This value of void is found to be substantially less than the bulk void The inferred indices of the particles are seen to be sensitiveto the value of the void fraction but somewhat

Comparison of Flame Development in a Spark-Ignition Engine Fueled with Propane and Hydrogen

Abstract: Results are presented of a study in which schlieren photographs of the flame in a special visualization spark ignition engine were taken, with first propane and then hydrogen as fuel, under conditions where the flow field, total energy released and geometric constraints were closely comparable. It is shown that the rate of development of the flame from its inception is much faster for hydrogen, that the cycle-to-cycle variations for hydrogen are substantially less, and that the thickness of the hydrogen turbulent flame is about one-third to one-quarter that of propane. Application of a wrinkled laminar flame model to these results shows that the characteristic burning time of the hydrogen flame is one-sixth that for propane, and that the characteristic scale of the wrinkling for hydrogen is 0.6 that for propane. Both the appearance of the two flames, and their turbulent flame speeds in relation to their laminar flame speeds, indicate that the two flames fall in different turbulent flame regimes (...

On Quenching of Curved Flames

Abstract: (1984). On Quenching of Curved Flames. Combustion Science and Technology: Vol. 40, No. 5-6, pp. 257-268.

Interpretation of Optical Measurements of Flame Generated Particles

Abstract: The dynamic equation for an aerosol undergoing simultaneous particle formation and coagulation is solved by the moment method, Solutions are obtained that quantify the evolution of the particle volume distribution function (PVDF) for both a size independent and a free molecular collision function. The solutions show that an equilibrium is rapidly established between the source and the coagulation terms and that the particle number concentration N is then proportional to the square root of the source strength. During the interval of equilibrium there is a widening of the PVDF, the moment ratio ƒN = D63 3/D30 3, and other width parameters. These quantities may exceed the asymptotic values that apply in the absence of particle formation. The time evolution of the PVDF of soot particles in a laminar diffusion flame is derived using the profile of soot volume fraction that has been observed in recent hi situ optical experiments. This analysis yields the evolution of ƒN, which is necessary to determine...

Flame Propagation in Aerosols of Fuel Droplets, Fuel Vapor and Air

Abstract: Thermal flame propagation into a quiescent combustible mixture consisting of liquid fuel droplets, fuel vapor and air is modelled using the heat release from (i) combustion of single droplets after ignition, and (ii) homogeneous reaction of the fuel vapor in the ambient gas. The purpose of the model is to show how it is possible to calculate a burning velocity acceleration effect due to the presence of liquid droplets in the mixture, in support of a limited number of previous experimental observations. The calculated results for a variety of fuels are in agreement with available experimental data which, however, is mostly outside the range of droplet diameters for which the burning velocity attains a maximum value.

Stabilization Heights in Lifted Methane-Air Jet Diffusion Flames Diluted with Nitrogen

Abstract: Abstract– Measurements of stabilization heights in lifted turbulent methane-air jet diffusion flames are performed in order to test a theory on local quenching of dilfusion flamelets. The global residence time (d/u0 the nozzle diameter, u0, the nozzle exit velocity) is scaled with Zq the instantaneous scalar dissipation rate at quenching. This quantity had been identified before to be charncteristic for the extinction of diffusion flamelets. Values for Zg, are deduced from measurements of laminar diffusion flames in a counter flow geometry. The scaling is performed for constant values of Zq the conserved scalar variable at sioichiometric conditions. Both the fuel and the air stream are diluted to obtain constant values of Zq. Scaling with Zq makes the curves for different residence times collapse into one curve with however, considerable experimental scatter. From this is concluded that Zq, is the essential kinetic parameter describing quenching effects in turbulent diffusion flames.

Laminar Burning Velocities of Chlorinated Hydrocarbon-Methane-Air Flames

Abstract: Laminar burning velocities of chlorinated methanes, trichloroethylene, and chloro-benzene in mixtures with methane and air were measured using the Bunsen cone method. The effects of mixture composition and precombustion temperature on the measured flame velocities were determined.

Coke Formation in the Combustion of Isolated Heavy Oil Droplets

Abstract: The combustion and coking of isolated, free and suspended droplets of a residual oil have been studied. The combustion history of the free and suspended droplets was found to be characterized by qualitatively similar phenomena. The coke particle was seen to form within the final 9 percent of the droplet burning period and the mass of the coke particle was about 3 percent of the mass of the initial residual oil droplet. Changing the temperature history of a droplet by varying the initial droplet size between 260 and 570 microns or by diluting the residual oil with a more volatile No. 2 fuel oil did not affect the mass fraction of the residual oil in the initial droplet which was converted to coke. The contribution to the coke particle from both the asphaltene and maltenc (residual oil from which the asphaltenes have been removed) components of the oil was found to be significant, and therefore fuel asphaltene content alone will not provide a satisfactory basis for determining the relative coking t...