Showing papers on "Diffusion flame published in 1979"

Fire radiation—A review

Abstract: Non-luminous radiation theories are reviewed and compared to Hottel's emmissivity charts for typical homogeneous combustion situations. Both narrow-band statistical and exponential wide-band models are considered. The results are then extended to luminous flames and the issue of whether flames can be regarded as gray is discussed quantitatively for various flame gases. Experimental investigations of the heat transfer components to burning fuel surfaces show that radiation is dominant at scales of 0.2–0.3 m and above. Comparative measurements of various non-charring plastic fuels show that the flame absorption-emission coefficient is the principal fuel property controlling the fuel's burning rate at hazardous scales. The measurements also indicate that the actual volumetric heat release rate is the same for different fuels burning as buoyant turbulent diffusion flames at similar scales. Concerning flame structure it is shown that the presence of luminous soot must be locally supported by chemical heat release in normal fire situations. It is also suggested that the observed proportionality of radiant heat output to fuel supply rate for geometrically similar buoyant diffusion flames is due to the insensitivity of the characteristic Kolmogorov microscale to changes in fuel flow rate. The review also discusses numerous important unresolved fire research topics.

Effects of turbulent structure and local concentrations on soot formation and combustion in C2H2 diffusion flames

Abstract: The present work is a continuation of previous work to study and model effects of turbulence on soot formation and combustion in turbulent flames. The main objectives have been to investigate effects of turbulent structure and of local chemical and thermodynamic conditions on soot formation and combustion. The following experimental conditions have been investigated for a free jet C2H2 diffusion flame: - Effects of variations of Reynolds number. - Effects of variable dilution of the fuel with nitrogen. - Effects of preheating of the fuel. - Effects of water—vapor addition to the fuel. Experimental results are given for mean quantities, fluctuations, the flatness factor and intermittency, and probability density distributions for the soot appearance. Based on a generalized eddy—dissipation concept, predictions were made for all experimental conditions. Variations of the Reynolds number had a strong effect on the amount of soot, with soot concentration increasing as the Reynolds number was reduced. Addition of nitrogen and water vapor to the fuel, as well as preheating the fuel, reduced the amount of soot. These observations could in general be accounted for by the mathematical model, taking into consideration the effects of changes of the turbulent structure due to the various input conditions.

Flame Spread Over Combustible Surfaces for Laminar Flow Systems Part II: Flame Heights and Fire Spread Rates

Abstract: In the previous paper (Part I), results were given for unburnt fuel fraction and wall heat flux for boundary layer flows with combustion. In the present paper, a highly simplified model of the combusting plume region and the previous unburnt fuel results are used to obtain analytical results for the ratio of flame to gasification heights in terms of the parameters r and B. It is found that these analytical results can be correlated by a simple formula which depends only on the ratio r/B. These results are compared with the predictions of other theoretical work and with available experimental measurements. The flame height results and the heat flux results from Part I are then combined with a simplified physical model to obtain an explicit result for laminar flame spread on thermally thick surfaces. Using this explicit result the effects of varying parameters such as ambient oxygen concentration, heat of combustion and heat of pyrolysis on the flame spread rate are described. The importance of cho...

Phenomenological Models of Soot Processes in Combustion Systems

Abstract: : The fuel sooting trends in pre-mixed and diffusion flames follow opposite directions due to the different influences of temperature. In pre-mixed flames as the temperature rises the rate of oxidative attack on the soot precursors increases faster than the rate of precursor formation through pyrolysis. Thus the higher the temperature the less is the tendency to soot. In the pre-mixed flame all aliphatics form some acetylene which is the monoelement for precursor formation and all fuels essentially follow the same nucleation route. Aromatics oxidize partially through carboxyl radicals formed during attack on the ring and thus aromatic mixtures burn 'richer' than corresponding aliphatic mixtures of the same stoichiometry. The dominant factors in diffusion flames are the temperature and fuel structure. Here the higher the temperature, the greater is the rate of fuel pyrolysis and the propensity to soot. Thus fuel structures which during pyrolysis form strongly conjugated (polar) species most readily soot. A pure acetylenic polymerization route is too slow to predict the early nucleation of soot particles and the presence of ions cannot predict the difference of isomers to soot in diffusion flames. Thus it has been postulated that only conjugated species which have polar resonance structures can undergo the fast reactions necessary.

An experimental study of flammability limits using counterflow flames

Abstract: Measurements of the lean and rich limits of flammability of methane-air mixtures and detailed experimental studies of the structure of premixed flames near the lean limit of flammability were made using a counterflow flame (a double flame) stabilized in the forward stagnation region of a porous cylinder. The flammability limits were determined by measuring a discernible break in the distance between the two flame zones with change in composition of the test mixture. The equivalence ratios of the lean and rich limits were found to be 0.47 and 1.72, respectively, which are somewhat broader than those obtaained by the standard technique. From previous results of structure analysis of rich fuel-air flames and the present results of lean fuel-air flames, it becomes clear that, in a double flame inside the flammability limits, the two reaction zones are completely separated, the reactions in the inner flame zone are completed, and the inner flame has the characteristics of a self-sustaining flame. On the other hand, in a double flame outside the flammability limits, the two reaction zones are not completely separated and the inner flame is closely correlated with the outer diffusion flame, so that the inner flame is not a self-sustaining flame. The flammability limit, defined as the limit of flame propagation, can be considered as the limit at which the supplies of reactants from the burned side must be necessary for a premixed flame to be established. The flammability limits determined in the present study depend only on the physicochemical properties of the mixture. The propagation velocity at the flammability limits was found to have a finite value.

Acceleration Effects on the Stability of Flame Propagation

Abstract: We consider the effect of acceleration on the diffusional thermal instability of a plane flame front in a premixed gaseous mixture. We recall that the diffusional thermal instability occurs if the Lewis number L of the mixture is less than a critical value $L_0 $. We show that the diffusional thermal instability is inhibited if the acceleration in the direction of the cold unburned gas is sufficiently high. If the acceleration is sufficiently low, plane flames are unstable, and perturbations of the flame front may evolve to fronts exhibiting cellular structure. Thus we are able to explain the phenomenon that for some mixutres, cellular fronts are observed only for upward propagating flames, while downward propagating plane flames remain stable.

Turbulence effects on flame speed and flame structure

Abstract: Turbulence effects on methane-air flames stabilized in grid turbulence were investigated through measurements of flame speed and mean and fluctuating flame temperature profiles. Published turbulent flame speed correlations were able to correlate the experimental flame speed data but were contradictory in indicating flame structure and combustion mechanisms. A simple, one-dimensional, wrinkled laminar flame model was used to predict characteristic flame temperature fluctuation levels. Comparison of these predictions with measured temperature fluctuations indicated that the majority of the flames studied were wrinkled laminar flames. However, a wrinkled laminar flame structure was inappropriate for the most intensely turbulent flames examined.

Measurements of a conserved scalar in turbulent jet diffusion flames

Abstract: The light scattering technique is used to measure the statistics of a conserved scalar in a hydrogen jet diffusion flame. Measurements of the means, root mean square fluctuations and probability density functions are made for flames at two Froude numbers. The results are presented as conventional and as Favre quantities. Conventional means are near sampled composition measurements but Favre means drop considerably faster on the centreline. The probability density functions reveal that in the buoyant turbulent flame the profiles are intermittent and far from Gaussian downstream. Favre rather than conventional probability density functions in the flames are found to resemble isothermal data.

Raman spectroscopic study of a laminar hydrogen diffusion flame in air

Abstract: Spontaneous Raman spectroscopy has been employed for time-averaged, spatially-resolved measurements of temperature and species concentration in an axisymmetric, laminar hydrogen diffusion flame in quiescent air. Temperatures were obtained from vibrational Q-branch raman spectra of N2, O2, and H2 and the rotational Raman spectra of N2 and H2, and concentrations of H2, and N2 were determined. The results are compared to existing numerical nonequilibrium calculations for the conditions of this experiment. Significant differences between experimental and predicted temperature and concentration profiles are observed. In particular, the flame is larger in both diameter and length and the flame zone is thicker than predicted. Some possible sources of the discrepancies are discussed.

A Model of the Effects of Pressure and Crossflow Velocity on Composite Propellant Burning Rate

Abstract: : Several variations of a model for prediction of burning rate versus pressure behavior of unimodal oxidizer composite propellants in the absence of crossflow were developed and evaluated against a set of data for a series of four formulations. Three variants, including one in which an average oxidizer- burning-surface intersectional area concept is employed and two in which allowance is made for geometry and stoichiometry changes as the propellant recedes past an oxidizer crystal, were found to give excellent agreement with data. The former variant was extended to treat multimodal oxidizer formulations, yielding predictions in excellent agreement with data for two additional formulations containing bimodal oxidizer. In the initial development of the erosive burning aspect of the model, columnar diffusion flame bending was assumed to the sole mechanism leading to burning rate augmentation by crossflow. This assumption led to severe underprediction of erosive burning effects. Accordingly, the model was revised through addition of a flow profile analysis for prediction of cross-flow-induced turbulence augmentation of transport properties governing heat feedback from gas flames as well as flame-bending. This revised model was found to yield good agreement with erosive burning data for five of the six formulations tested, but gave higher predicted rates than observed values for the sixth propellant.

Rate constants for the reactions of atomic boron with O2, SO2, CO2, and N2O

Abstract: The gas phase reactions of atomic boron with O2, SO2, CO2, and N2O have been studied using a diffusion flame technique. Atomic boron is produced in a microwave discharge of 1% B2H6 in helium and diffuses into a chamber containing the oxidizer also diluted in helium. The temperature was approximately 300 °K. The rate constants were determined from atomic absorption measurements of boron density in the flame. The rate constants for the O2 and SO2 reactions are (9±7) ×10−12 cm3 molecule−1sec−1 and (7±5) ×10−12 cm3 molecule−1sec−1, respectively. An upper limit of 5×10−13 cm3 molecule−1sec−1 is estimated for the rate constants of the CO2 and N2O reactions.

A schlieren study of combustion in a rapid compression machine simulating the spark ignition engine

Abstract: A rapid compression machine in a configuration to simulate a spark ignition engine with a simple cylindrical combustion chamber was operated to generate high-speed schlieren movies of combustion of a propane-air mixture over a range of flow conditions. These included the flows produced by the piston alone, and by intake processes with both shrouded and nonshrouded valves. Hot-wire anemometer measurements of turbulence were made at several locations in the chamber and for two wire orientations in order to examine homogeneity and isotropy. The behavior of these flows is discussed. Burning velocity was calculated from the flame propagation velocity observed from the film and pressure records of combustion following the method of Fiock and King. For combustion with no intake generated flow, a “roll-up vortex” along the cylinder walls was observed which appeared to have only a minor influence on the flame. The burning velocity for flames produced with swirl using the shrouded valve, correlates to the intensity of turbulence according to the relations observed by Abdel-Gayed and Bradley 2 where the ratio of turbulent to laminar flame speed and the ratio of turbulent intensity to laminar flame speed are the correlation parameters. Burning velocity for combustion with the nonshrouded valve showed poorer correlation which was attributed to the problem of interpreting the hot-wire signal when no well-defined mean flow exists. The results suggest that a deficiency in the heat release for the first part of the flame travel yields a propagation velocity lower than would be expected for the extent of the observed flame. The source of this deficiency appears to be one or more of the following: o a. Incomplete reaction in the flame front. b. Finite flame thickness c. Deviations from spherical propagation as the flame and expansion flows interact with the chamber walls.

The dynamics and radiant intensity of large hydrogen flames

Abstract: Measurements of the spectral radiant intensity and structure of large turbulent hydrogen diffusion flames, with mass flow rates from 5.4 to 68kg/sec, are presented. These measurements are compared with predictions of the reacting, buoyant flowfield and radiant intensities, utilizing the flowfield and infrared radiation computer codes originally developed to determine rocket and jet aircraft exhaust plume characteristics. The predictions of radiant intensity are shown to be within the data fluctuations (±25%) and the predicted length of the diffusion flame is consistent with the data. However, the predicted flame width is somewhat smaller than that obtained from thermovision measurements of the flame (1.9–3.3 microns). This is attributed to the failure of the turbulence model used in the flowfield code to accurately treat the large density gradients across the flame. It is shown that, for the conditions of these tests, the radiant intensity does not scale linearly with mass flow rate. The ratio of radiant intensity to rate of heat release, F, is demonstrated to vary from 0.153 at very low velocities (small Froude numbers) to 0.086 at the high velocities (large Froude numbers). This trend of decreasing F with increasing velocity is consistent with the experimental observations of Brzustowski et al. 1 The reasonable agreement between the theoretical predictions and data suggests that these codes, and selected laboratory experiments, may be used to develop scaling relations for radiant intensity and flame structure which include effects of optical depth, mass flow rate, velocity, pipe radius, and Froude number.

Cars investigations in flames

Abstract: Coherent anti-Stokes Raman spectroscopy (CARS) investigations in a variety of flames are described. These studies were directed toward finding solutions to previously identified or newly encountered shortcomings of CARS and to provide a foundation on which to base future development. CARS is generated by mixing a 10 pps, frequency-doubled neodymium “pump” laser with a spectrally broadband, laser-pumped, “Stokes” dye laser. This approach obviates the requirement to frequency scan the dye laser and generates the entire CARS spectrum with each pulse permitting “instantaneous” measurements of medium properties. A crossed-beam, phase-matching technique, termed BOXCARS, is used which leads to greatly enhanced and unambiguous spatial resolution in contrast to the conventionally employed collinear phase-matching approaches. Using this technique, moderate resolution (∼ 1.25 cm−1) CARS spectra from hot N2, obtained by scanning the spectrum in premixed laminar flames, show excellent agreement with computer generated model predictions. Lower resolution (2.7 cm−1) collinear phase-matched CARS spectra of flame N2 have been obtained in a single 10 nanosecond pulse using an optical multichannel analyzer. These single pulse spectra also display good agreement with predicted spectra and demonstrate the feasibility of single pulse thermometry. Measurements in a highly sooting, laminar propane diffusion flame revealed the existence of a coherent spectral interference arising from electronic resonance CARS generation from C2, produced by the laser vaporization of the soot. Reduction of the Stokes laser bandwidth and use of polarization filters permitted low distortion N2 CARS spectra to be obtained. These spectra, when computer fitted, allowed determination of the temperature marking the first measurement in highly sooting flames by a remote, spatially precise diagnostic technique. CARS species sensitivity was examined in a study of flame CO detectability levels. Very good agreement between CARS CO spectra and the computer model was obtained at the 4 percent CO level. With the fluctuations in the experimental apparatus it was difficult to detect CO below the 1–2 percent level. The computer calculations indicate that CO would be barely detectable to about 0.5 percent using conventional CARS approaches.

Process and apparatus for stoichiometric combustion of fuel oil

Abstract: A process and apparatus for the combustion of liquid fuel provides an extremely intense blue/violet flame having a temperature in excess of 3000° F. with combustion under near perfect stoichiometric conditions without the formation of soot. The liquid fuel is atomized and mixed with air within a nozzle and enters a flame tube surrounding the nozzle as a conical stream where it is further atomized by jets of air directed to converge on the stream and mixed with secondary combustion air to obtain the desired combustion mixture. Yet further atomization of the liquid fuel-air mixtures within the flame tube can be obtained as a consequence of at least partial vaporization of the liquid fuel therein through the heat of the flame tube. An advantageous relationship exists between the size, angle and point of convergence of the air jets with the atomized conical stream, the flame tube diameter and length and the location of the nozzle therein and the fuel feed.

Numerical study of laminar wall quenching

Abstract: Laminar flame quenching at the cold wall of a combustion chamber has been studied, using a numerical model to describe the reactive flow. The model combines an unsteady treatment of the fluid mechanics and a detailed chemical kinetic reaction mechanism. Fuels considered included both methane and methanol. The one-dimensional case of flame propagation perpendicular to the wall was studied. Two reference cases are described in detail for flame quenching at 10 atmospheres pressure and a wall temperature of 300/sup 0/K with stoichiometric mixtures of methane-air and methanol-air. In each case a conventional laminar flame propagates toward the wall, approaching to within a distance determined by the thermal flame thickness. Chemical kinetic factors, particularly differences between the temperature dependence of radical recombination reactions and conventional chain branching and chain propagation reactions, are shown to be responsible for quenching the flame near the wall. The flame stagnates, but fuel remaining near the wall diffuses out of the boundary region and is rapidly oxidized away from the wall. Subsequent model calculations demonstrate the effects of variations in pressure, fuel-air equivalence ratio, wall temperature, and type of fuel. Computed results from these methane and methanol flame quenching models indicate that the total unburned hydrocarbonmore » content is considerably smaller than is commonly beleived and that thermal wall quenching may not be the major source for hydrocarbon emissions from internal combustion engines at near-stoichiometric conditions.« less

Laser induced sodium fluorescence measurements in a turbulent propane diffusion flame

Abstract: A cw laser induced fluorescence technique has been employed to make spatially (0.15 mm3) and temporally (DC to 10 kHz) resolved measurements on trace quantities of Na ( Lack of simultaneous measurements of temperature and/or another scalar introduces an uncertainty of up to ±30% in the fluorescence yield and hence in the local atomic Na number density. Estimates of total Na, necessary if the data are to be interpreted in terms of a conserved scalar, are inaccessible in this flame, except perhaps near the flame front, without detailed local composition and temperature information. The fluorescence data have been analysed to obtain statistical information, including power spectra and probability density functions (pdfs). The latter have been compared qualitatively to pdfs of temperature obtained by other workers in similar flames. Future experimental work applying LIFS in combustion is discussed briefly.