Stephen R. Turns

Oxides of nitrogen emissions from turbulent jet flames: Part I—Fuel effects and flame radiation

Measurements of oxides of nitrogen emission indices, flame radiant fractions, and visible flame dimensions were made for turbulent jet diffusion flames covering a wide range of flow conditions. Objectives of the study were to explain the observed scaling of NOx emissions with flow variables and to understand the interrelationships among NOx, flow conditions, and flame radiation. The flames were vertical and stabilized with hydrogen pilot flames on straight tube burners. Flow conditions were varied by changing the initial jet velocity and/or the burner tube diameter. Four burner sizes were used, with diameters ranging from 2.18 to 6.17 mm; and four fuel types, having a wide range of sooting tendencies, were employed: methane, ethylene, propane, and a 57%CO43%H2 (by volume) mixture. The ranges of Reynolds numbers and Froude numbers explored were 3,130–88,500 and 218 to 2.8 × 106, respectively. The effects of flow parameters and fuel type on radiant losses are shown to be important in determining the NOx emissions from simple jet flames. For high-temperature flames (T > 2050 K), overall NOx production rates for all four fuels were found to scale with characteristic flame temperatures deduced from the measured radiant fractions in a manner consistent with Zeldovich kinetics. This successful scaling of NOx production rates with global flame temperatures and residence times is consistent with, but does not prove, the view that much of the NOx emitted by jet flames is formed in large-scale eddies at the flame tip. NOx production rates higher than expected from the thermal mechanism alone are observed for the hydrocarbon fuels at lower flame temperatures (<2050 K), with the NOx production rates ranking in the same order as sooting tendencies. This suggests that gas-molecular radiation is more relevant than broadband radiation from soot for determining temperatures in NO formation zones. Prompt NO and/or other soot-NO interactions may also be important for the hydrocarbon fuels in this temperature regime. Previously reported Reynolds and Froude number dependencies for NOx production rates are examined and found to be consistent with flame heat loss characteristics.

Understanding NOx formation in nonpremixed flames: Experiments and modeling

Experimental studies and models of NOx formation in simple nonpremixed flames at atmospheric pressure are reviewed. Laminar flames are briefly discussed, while the bulk of the review focuses on axisymmetric, turbulent, jet flames. The issue of the scaling of NOx emission indices with nozzle exit diameter, initial jet velocity, and fuel type is a major theme of the article. The failure of a simple leading-order scaling is investigated and interpreted in terms of the various interrelated parameters affecting NO formation: the relative importance of various NO-forming chemical pathways, departures of O-atom concentrations and temperatures from their equilibrium values, flame strain, and flame radiation.

Oxides of nitrogen emissions from turbulent jet flames: Part II—Fuel dilution and partial premixing effects ☆

Measurements of NOx emission indices, flame radiant fractions, and visible flame length were made for turbulent, nonpremixed, jet flames for which various amounts of inert diluent or air were mixed with the fuel. The objective of the study was to explore further the role of flame radiation in NOx production in jet flames. Vertical free jet flames were stabilized on a 4.12 mm diameter straight-tube burner. Four fuels, CH4, C2H4, C3H8, and a 95% CO5% H2 mixture (by mass); three inert diluents, N2, Ar, and CO2; and air premixing were employed in parametric tests. Complementary dilution experiments were run with laminar jet flames using the three hydrocarbon fuels and N2. For the turbulent flames, the results showed that the effects of dilution and premixing were strongly dependent on fuel type. Flame temperatures and NOx emissions increased when the more sooting fuels (C3H8 and C2H4) were diluted or partially premixed, resulting in increased NOx emissions. The opposite trend was observed for the nonluminous COH2 flames. Using the results reported here and from Part I [1] of this study, the effects of residence time, flame temperature, and departure from equilibrium on NOx emissions, regardless of what parameter affected the change, were well characterized by regressing characteristic NOx production rates as a function of nonadiabatic characteristic flame temperatures and global residence times. Separate regressions for the hydrocarbon and COH2 flames showed a weaker dependence of NOx on temperature for the hydrocarbons, suggesting that the prompt NO mechanism is quite active in these flames. The laminar flame experiments demonstrated the importance of the relative locations of NOx-producing regions and soot-containing (strongly radiating) regions of the flame.

Ignition of Aluminum Slurry Droplets

The ignition of aluminum slurry fuel droplets was studied in the post-flame region of a flat-flame burner. Individual droplets having initial diameters ranging from 500 to 1100 microns were supported on silicon carbide filaments and rapidly exposed to a hot gas environment. Burner operating conditions were varied to provide oxygen mole fractions ranging from 0.10 to 0.25 and gas temperatures from 1250 to 1800 K at atmospheric pressure. Flame environments both with and without water vapor were considered. Using motion picture photography, ignition times were measured and ignition limits were determined. Experiments showed that AI/JP-10 slurry droplets ignited at ambient gas temperatures higher than 1300-1450 K within 0.2-1.2 s. The ignition processes of JP-10 bun -off and agglomerate heat-up were modeled analytically, and the theoretical predictions of ignition times were compared with experimental data. Ignition limits also were studied analytically. In both cases, theoretical calculations and experimenta...

Measurements of Oxides of Nitrogen Emissions from Turbulent Propane Jet Diffusion Flames

Measurements of global emission indices for NO and NO,were made in unconfined,vertical,propane jet diffusion flames. Initial fuel jet velocities were varied to provide a range of jet Reynolds numbers from 10000 to 60000. The measurements showed that scaling of NO, emission indices with velocity changes as the flame character changes from transitional-turbulent (Re ≈ 10000) to fully turbulent (Re > 20000). At the lower Re condition. the observed scaling approaches that reported in the literature for similar conditions. Arguments are presented to show that radiation losses from the flames may account for the observed scaling. Significant amounts ofN02 were found in the flame plumes. The variation of NO2/NOx, ratios with jet velocity and with radial position was consistent with the hypothesis that NO2is formed in the cooler post-flame regions. in agreement with the findings of others for CO and CO/H2 fueled diffusion flames.

A Study of the Influence of Oxygen Index on Soot, Radiation, and Emission Characteristics of Turbulent Jet Flames

Many combustion applications benefit from the use of oxygen-enriched air or pure oxygen as an oxidizer. To provide guidance in the use of oxygen-enriched or oxy-fuel combustion, research was conducted to understand how key parameters affect the radiation and emissions characteristics of jet flames for a range of oxygen indices from 21% (air) to 100% (pure O 2 ). Experiments were conducted on simple jet flames created by fuel issuing from a 3-mm i.d. straight tube into a low-velocity oxidizer stream for a variety of conditions. In addition to O 2 content of the oxidizer, fuel jet velocity and fuel type were varied. Laser light extinction was used to measure mean soot volume fractions as functions of axial distance. Other measurements include axial profiles of total radiant heat flux and NO x and CO emission indices. Results of these experiments were compared with predictions from a modified version of the two-stage Lagrangian (TSL) model of Broadwell and Lutz. The TSL code was modified by the incorporation of a detailed soot model from Frenklach et al. and the addition of a radiation submodel that accounts for both gas-band and soot blackbody radiation. In general, the TSL model captures the trends observed in the experiments for various parameters as functions of oxygen index, fuel type, and fuel-jet velocity, although absolute values are not predicted with engineering accuracy. The model was also exercised to predict the soot contribution to the total radiant heat transfer with soot volume fractions adjusted to be more representative of experimentally measured values.

Experiments on axisymmetrically pulsed turbulent jet flames

An experimental study of the effects of strong axisymmetric pulsing on a free, vertical turbulent jet diffusion flame is presented. The jet flame was pulsed over the frequency range of 2-1300 Hz with amplitudes ranging from 0.13-0.89 of the centerline jet velocity. Conditionally averaged centerline velocity measurements and flame photographs were obtained to characterize the pulsed jet flames. The centerline evolution of the pulse waveforms was examined and found to be dependent on the pulse frequency; the centerline decay of the pulse amplitude increased with increasing pulse frequency. The visual dimensions of the pulsed flame were also frequency dependent. Significant changes in the local flame structure were observed, which exhibited a constant local nondimensional frequency of about 0.2, consistent with a preferred-mode coupling between the pulsing and the jet flame structure. Several features of the results also suggest a coupling between the forcing and low frequency structure in the outer preheat layer surrounding the jet flame.

Disruptive Burning of Aluminum/arbon Slurry Droplets

Abstract Abstract–The disruptive combustion of probe-supported and free-falling Al/C/JP-lO slurry droplets (200 to 1OOO, μm in diameter) was investigated experimentally. The disruption processes and disruption times were analyzed using backlit or natural-light film records taken with a high-speed motion picture camera. Parametric studies were conducted to investigate the influence of the relative proportions of aluminum and carbon in the solids and of surfactant and stabilizing additive levels; and disruption times were quantified with respect to droplet diameter, total solids loading. ambient gas temperature, and oxygen mole fraction. Slurry composition parameters were found to affect the type and strength of the observed disruptive phenomena presumably by altering the porosity and plasticity of the surface shells of solids formed during the early stages of gasification.

Nitric Oxide Emissions from Laminar Diffusion Flames: Effects of Air-Side versus Fuel-Side Diluent Addition

Abstract Flue gas recirculation (FGR) is a well-known method used to control oxides of nitrogen (NO x ) in industrial burner applications. Recent small- and large-scale experiments have shown that introducing the recirculated flue gases with the fuel results in a much greater reduction in NO x , per unit mass of gas recirculated, compared to introducing the flue gases with the combustion air. At present, however, there is no definitive understanding of why introducing the recirculated gases with the fuel is more effective than conventional FGR. The objective of the present investigation is to ascertain to what degree chemical kinetics and/or molecular transport effects can explain the differences in NO x reduction observed between fuel-side and air-side introduction of flue gases by studying laminar diffusion flames. Numerical simulations of counterflow diffusion flames using full kinetics were performed and NO x emission indices calculated for various conditions. Studies were conducted in which N 2 diluent was added either on the fuel- or air-side of the flame for conditions of either fixed initial velocities or fixed fuel mass flux. Results from these simulation studies indicate that a major factor in diluent effectiveness is the differential effect on flame zone residence times associated with fuel-side versus air-side dilution. Experiments using laminar jet flames were conducted in which either the air or fuel stream was diluted with N 2 . The experiments showed that fuel-side dilution results in somewhat greater NO x emission indices than air-side dilution. The higher flame temperatures measured with fuel dilution appear to be the principal cause of the higher emissions. The results of both the numerical simulations and the experiments suggest that, although molecular transport and chemical kinetic phenomena are affected by the location of diluent addition depending on flow conditions, the dramatically greater effectiveness of fuel-side over air-side introduction of recirculated flue gases in practical applications likely results also from differences in turbulent mixing and heat transfer.

Turbulent jet flames in a crossflow : Effects of some jet, crossflow, and pilot-flame parameters on emissions

Abstract A study was conducted to document and understand the pollutant emissions characteristics of high-velocity jet flames in a crossflow of air. Measured were unburned hydrocarbon, carbon monoxide, and NOx emission indices; the ratio of NO2 to NOx; and flame dimensions. Four fuels of differing sooting propensities were used: C2H4, C3H8, CH4, and a 95 wt.% CO/5 wt.% H2 mixture. A straight-tube burner was employed (4.12 mm i.d.) with jet velocities ranging from approximately 15–100 m/s, depending on the fuel type. Tests were conducted at two crossflow velocities, 2.3 and 4.3 m/s, yielding jet-to-crossflow velocity and momentum ratios within the ranges of 4–37 and 9–940, respectively. The crossflow flames are shorter than comparable straight-jet flames, indicative of enhanced entrainment and mixing. Comparatively high levels of unburned hydrocarbons and CO emissions, along with high ratios of NO2 to NOx, obtain for crossflow flames as a consequence of fuel being swept from the jet in the nearfield, which is possibly supplemented by local flame quenching brought about by the rapid mixing with the crossflow. NOx emissions tend to follow the same trends as those of straight-jet flames, for most conditions. Ethylene flames, however, are a major exception. For these flames, NOx emission indices depend strongly on both the jet-to-crossflow momentum ratio and the magnitude of the crossflow velocity. Radiant fraction measurements suggest that the NOx levels are coupled strongly with how the crossflow affects in-flame soot and, consequently, radiation losses.