Non-premixed Jet Research Articles

Investigation of conditional source-term estimation applied to a non-premixed turbulent flame

A turbulent combustion model, Conditional Source-term Estimation (CSE) is applied to a non-premixed turbulent jet methane flame. The conditional chemical source terms are determined on the basis of first order closure and the conditional averaged species concentrations are obtained by inverting an integral equation. The Tikhonov method is implemented for regularisation. Detailed chemistry is tabulated using the trajectory generated low-dimensional manifold method. Radiation due to the gaseous species is included. Reynolds Averaged Navier–Stokes calculations are performed using two different turbulence models. The objectives of the paper are (i) assessment of the impact of the main numerical parameters in CSE and (ii) comparison of the CSE numerical predictions with available experimental data and results from previous simulations for the selected flame. The number of CSE domains and the number of points in each CSE domain are shown to have a significant impact on the results if not selected appropriately. The present CSE calculations always converge to unique and stable predictions. The corrected k–ε model yields mixture fraction profiles in good agreement with the experimental data values for axial locations in the first half of the flame. Farther downstream, the RNG k–ε model performs better. Overall, the current predictions for the mixture fraction are in good agreement with the experimental data. The predicted temperatures using CSE and the k–ε turbulence model with a modified value of Cε1 = 1.47 are found to be in very good agreement with the experimental data. Further, the current CSE results are of comparable quality with previous simulations using the flamelet model and conditional moment closure. Future work may include further investigation on optimal determination of the regularisation parameter and alternative regularisation techniques, soot modelling within the CSE formulation, and improved formulation of radiation.

The effect of CO2 addition on the flame behavior of a non-premixed oxy-methane jet in a lab-scale furnace

The dilution effect of carbon dioxide (CO2) in an oxidizer was studied experimentally in a non-premixed oxy-methane flame. A lab-scale furnace with a slot-type burner was used as a model of an industrial furnace. The objectives of the current study are to investigate the flame stabilization, to analyze the flame spectra, and to study the flame behavior of non-premixed oxy-methane flames as the CO2 mole fraction is varied in an oxidizer. The flame stabilization was estimated by changing the flow velocity at the fuel jet nozzle exit (uF) and the oxygen nozzle exit (uOx) from uF=7–50m/s for methane (CH4) and from uOx=10–120m/s for oxygen (O2), respectively. The experimental results indicate that the flame stabilization of non-premixed oxy-methane flames decreased with increases in the CO2 mole fraction in the oxidizer. The flame spectra of non-premixed oxy-methane flames had a maximum intensity at the wavelength (λ) of λ=308nm. The reaction zone of non-premixed oxy-methane flames was broadened by the addition of CO2 to the oxidizer.

Large eddy simulation of soot evolution in an aircraft combustor

An integrated kinetics-based Large Eddy Simulation (LES) approach for soot evolution in turbulent reacting flows is applied to the simulation of a Pratt & Whitney aircraft gas turbine combustor, and the results are analyzed to provide insights into the complex interactions of the hydrodynamics, mixing, chemistry, and soot. The integrated approach includes detailed models for soot, combustion, and the unresolved interactions between soot, chemistry, and turbulence. The soot model is based on the Hybrid Method of Moments and detailed descriptions of soot aggregates and the various physical and chemical processes governing their evolution. The detailed kinetics of jet fuel oxidation and soot precursor formation is described with the Radiation Flamelet/Progress Variable model, which has been modified to account for the removal of soot precursors from the gas-phase. The unclosed filtered quantities in the soot and combustion models, such as source terms, are closed with a novel presumed subfilter PDF approach that accounts for the high subfilter spatial intermittency of soot. For the combustor simulation, the integrated approach is combined with a Lagrangian parcel method for the liquid spray and state-of-the-art unstructured LES technology for complex geometries. Two overall fuel-to-air ratios are simulated to evaluate the ability of the model to make not only absolute predictions but also quantitative predictions of trends. The Pratt & Whitney combustor is a Rich-Quench-Lean combustor in which combustion first occurs in a fuel-rich primary zone characterized by a large recirculation zone. Dilution air is then added downstream of the recirculation zone, and combustion continues in a fuel-lean secondary zone. The simulations show that large quantities of soot are formed in the fuel-rich recirculation zone, and, furthermore, the overall fuel-to-air ratio dictates both the dominant soot growth process and the location of maximum soot volume fraction. At the higher fuel-to-air ratio, the maximum soot volume fraction is found inside the recirculation zone; at the lower fuel-to-air ratio, turbulent fluctuations in the mixture fraction promote the oxidation of soot inside the recirculation zone and suppress the accumulation of a large soot volume fraction. Downstream, soot exits the combustor in intermittent fuel-rich pockets that are not mixed during the injection of dilution air and subsequent secondary fuel-lean combustion. At the higher fuel-to-air ratio, the frequency of these fuel-rich pockets is increased, leading to higher soot emissions from the combustor. Quantitatively, the soot emissions from the combustor are overpredicted by about 50%, which is a substantial improvement over previous works utilizing RANS to predict such emissions. In addition, the ratio between the two fuel-to-air ratios predicted by LES compares very favorably with the experimental measurements. Furthermore, soot growth is dominated by an acetylene-based pathway rather than an aromatic-based pathway, which is usually the dominant mechanism in nonpremixed flames. This finding is the result of the interactions between the hydrodynamics, mixing, chemistry, and soot in the recirculation zone and the resulting residence times of soot at various mixture fractions (compositions), which are not the same in this complex recirculating flow as in nonpremixed jet flames.

The effect of CO addition on the flame behavior of a non-premixed oxy-methane jet in a lab-scale furnace

The characteristics of the flame behavior of a non-premixed oxy-methane jet were investigated by increasing the carbon oxide (CO) mole fraction in a fuel jet. In the current study, a lab-scale furnace was used with a slot-type burner to modify industrial furnaces. The flow velocity was fixed at uF=25m/s for methane and uOx=25m/s for oxygen. The mole fraction of CO gas in a fuel jet (XCO) was varied from XCO=0% to 30% over 10 steps. To observe the flame behavior of a non-premixed oxy-methane jet, flame stabilization and flame luminescence were experimentally measured. The objective of the current study is to investigate the effect of CO addition in a fuel jet on the flame stabilization, flame spectra, and flame behavior of a non-premixed oxy-methane flame. From the experimental results, the flame stabilization area expanded as the CO mole fraction in the fuel jet decreased. The intensity of light emission in the rage of Δλ=303–313nm increased as the CO mole fraction decreased. In addition, the flame length became longer and the flame slope was more declined as lee CO was added.

Measurements and statistics of mixture fraction and scalar dissipation rates in turbulent non-premixed jet flames

The present study adds to the limited database of mixture fraction and scalar dissipation rate measurements within turbulent flames. One noteworthy aspect is that the current measurements are two-dimensional images, which provide the physical structure of the scalar dissipation rate field. Two turbulent carbon monoxide flames are studied at two different jet-exit velocities, where the latter flame is near blowoff. The effects of increasing levels of local extinction (by increasing the jet velocities) on the structure and statistics of the mixture fraction fluctuation and scalar dissipation rate fields are investigated in detail. In general, the topology of the scalar dissipation rate layers is similar to that found previously in turbulent non-reacting flows, with some notable exceptions. For the lower jet-velocity case, the dissipation layers were somewhat smooth, isolated, and preferentially aligned in the axial direction, while for the higher jet-velocity case (near global extinction), the scalar dissipation rate field was characterized by an increased number of shorter, thinner, and highly wrinkled dissipation layers. It appears that as the flame nears blowoff, the scalar dissipation rate field exhibits an increasing level of isotropy, resembling turbulent non-reacting flows. In addition to providing new insights on the physical nature of the dissipation fields, the two-dimensional statistics of the mixture fraction and scalar dissipation rate are examined in detail, in the form of conditional averaging and through the construction of probability density functions. It is found, that although the two flames have apparent differences in physical structure, the corresponding statistical properties of the scalar dissipation rate were similar. Probability density functions (pdfs) of the scalar dissipation rate exhibited large departures from strict log-normality and the pdfs of the scalar dissipation rate were found to be significantly different if only the values of the scalar dissipation conditioned on the stoichiometric contour were considered compared to the entire flowfield. This has implications for modeling assumptions that seek to describe the entire turbulent flowfield by a single pdf. The nature of the scalar dissipation rate in the current lower-Reynolds number turbulent flames, as revealed through the statistics, is compared to previous, well-known data sets obtained in the Sandia series of piloted jet flames as well as other piloted non-premixed jet flames at higher Reynolds numbers. The present flame conditions are significantly different from previous work, yet the general structure, statistical trends, and general effects of increasing levels of local extinction agree favorably with the previous experimental data.

Hydrogen-enriched non-premixed jet flames: Compositional structures with near-wall effects

The species concentrations of non-premixed hydrogen and syngas flames were examined using results obtained from direct numerical simulation technique with flamelet generated manifold chemistry. Flames with pure H2 and H2/CO mixtures are discussed for an impinging jet flame configuration. Single-point data analyses are presented illustrating the effects of fuel composition on species concentrations. In general, scatterplots of all species show the effects of fuel variability on the flame compositional structures. The behaviours of major combustion products and key radicals species indicate the effects of CO concentration on the 2/CO syngas combustion. In particular, high concentration of CO tends to induce local extinction in the 2/CO flames in which critical chemical reactions of the fuel mixture such as CO + OH become important. The unsteady fluctuations of species profiles in the wall jet region characterise the complexity of the distributions of compositional structures in the near-wall region with respect to the effects of CO concentration on the combustion of hydrogen-enriched fuels.

Stability characteristics of non-premixed turbulent jet flames of hydrogen and syngas blends with coaxial air

The stability characteristics of attached hydrogen (H2) and syngas (H2/CO) turbulent jet flames with coaxial air were studied experimentally. The flame stability was investigated by varying the fuel and air stream velocities. Effects of the coaxial nozzle diameter, fuel nozzle lip thickness and syngas fuel composition are addressed in detail. The detachment stability limit of the syngas single jet flame was found to decrease with increasing amount of carbon monoxide in the fuel. For jet flames with coaxial air, the critical coaxial air velocity leading to flame detachment first increases with increasing fuel jet velocity and subsequently decreases. This non-monotonic trend appears for all syngas composition herein investigated (50/50 → 100/0% H2/CO). OH∗ chemiluminescence imaging was performed to qualitatively identify the mechanisms responsible for the flame detachment. For all fuel compositions, local extinction close to the burner rim is observed at lower fuel velocities (ascending stability limit), while local flame extinction downstream of the burner rim is observed at higher fuel velocities (descending stability limit). Extrema of the non-monotonic trends appear to be identical when the nozzle fuel velocity is normalized by the critical fuel velocity obtained for the single jet cases.

On preheating and dilution effects in non-premixed jet flame stabilization

The impact of preheating and dilution on methane/air non-premixed flame stability are studied experimentally. Six preheating levels are considered for initial reactant temperature between 295K and 850K in a round jet configuration. Four diluent gases are added on the air-side, either CO2, N2, Ar or a (CO2+Ar) mixture having the same molar heat capacity as N2. For undiluted flames, jet transition velocities between attached and lifted states are investigated depending on initial reactant temperature. The hysteresis zone defined by these stability limits is shifted towards higher jet velocities with preheating. Whereas jet and coflow temperatures were identical in similar previous experiments, the present work allows examination of the thermal effects from either fuel or oxidizer streams. Flame stability is described based on the propagative aspects of the flame leading-edge, by analogy with the temperature dependency of the laminar burning velocity of a stoichiometric mixture. Results show that the jet temperature has a major influence on the lifting of an attached flame, whereas the coflow temperature remains important for the reattachment of a lifted flame. In addition, flame stability experiments have been performed at high levels of both preheating and dilution. Stability maps of critical dilution ratios at lifting have been obtained with preheating. It appears that the ability of a diluent to break flame stability keeps the same relative order as at ambient temperature. It is even enhanced with preheating because higher temperature widens the gap between diluent molar heat capacities Cp. The Cp approach is however not sufficient to interpret the temperature dependency of the relative influence of the different dilution effects. Furthermore, the role played by the jet flow regime on attached flame stability in dilution-induced lifting experiments is highlighted when dilution is coupled with preheating.

Kinetic and fluid dynamic modeling of ethylene jet flames in diluted and heated oxidant stream combustion conditions

This paper reports the results of a numerical investigation of several turbulent nonpremixed ethylene jet flames, either undiluted or diluted with hydrogen, air or nitrogen, which have been experimentally studied using a jet in hot coflow (JHC) burner. The fuel jet issues into a hot and highly diluted coflow at two O2 levels (3% or 9%) and a temperature of 1100 K. These conditions emulate those of moderate or intense low oxygen dilution (MILD) combustion. The attention is mainly focused on assessing the performance of different models for the turbulence–chemistry interaction. The model predictions are compared to experimental measurements. The Eddy dissipation concept (EDC), flamelet and PDF transport models were used in combination with a two-equation turbulence model (k–ε) and two different kinetic mechanisms (GRI-Mech 3.0 and POLIMI). The effect of the oxygen concentration in the coflow is well captured by all models. A modified version of the EDC model, recently proposed for the modeling of methane MILD combustion conditions, is also used. This model significantly improves the predictions of the EDC in the case of the ethylene flames studied in this paper. The agreement between measurements and predictions of the flamelet model is generally not as good as the PDF transport and modified EDC models, whose predictions are in good agreement with the measurements, and are improved especially for the apparent liftoff heights and the peak flame temperature. The effect of the fuel composition on the CH2O formation is also discussed, showing generally good agreement.

Infrared Imaging and Spatiotemporal Radiation Properties of a Turbulent Nonpremixed Jet Flame and Plume

Radiation transfer from turbulent nonpremixed jet flames and plumes is important in many applications such as energy-efficient combustion systems, temperature sensitive pollutant control, and detection, control, and suppression of accidental fires. Combined spatial and temporal correlations of scalar values such as temperature and species concentrations affect the emitted radiation intensity. Spatiotemporal correlations and radiation intensity measurements downstream of the reacting parts of flames (plumes) have received limited attention. Motivated by this, planar time-dependent narrowband radiation intensity measurements are acquired of a turbulent nonpremixed flame and its plume using an infrared camera. Temporally and spatially correlated instantaneous realizations of local scalars and path integrated intensity values are calculated using a stochastic time and space series analysis, a narrowband radiation model, and the radiative transfer equation. The time-dependent infrared images reveal intermittent, low intensity regions in the plume characteristic of buoyancy-dominated transport. High radiation intensity structures are observed in the flame characteristic of momentum dominated flow and vorticity driven mixing. Normalized intensity fluctuations are nearly constant in the flame region, but increase by up to a factor of three in the plume. Normalized temporal correlations, power spectral density functions, and spatial correlations of the intensity are independent of the spatial location throughout both the flame and the plume. Spatial correlations of the radiation intensity exhibit approximately linear decay to half an integral length scale followed by an exponential decrement. The radiation intensity fluctuations remain spatially correlated up to separation distances two times larger than the integral length scale. Space–time cross correlations of the intensity fluctuations are measured for the first time and are shown to be more isotropic in comparison to the product of the spatial and temporal correlations. This suggests that a correction factor should be applied to the space–time correlation model in future stochastic calculations to account for the anisotropy. The infrared imaging technique, illustrated in this paper, is promising to be a useful qualitative and quantitative nonintrusive technique for studying both reacting and nonreacting flows.

Volumetric PIV and 2D OH PLIF imaging in the far-field of a low Reynolds number nonpremixed jet flame

Cinematographic stereoscopic PIV with temporal and spatial resolution ranging from 2.6 to 5.5 Kolmogorov scales, which is sufficient to accurately represent most of the dissipation structures, is used in conjunction with Taylor’s frozen flow hypothesis to generate quasi-instantaneous pseudo-volumes of the three-component velocity field in the far-field of a nonpremixed jet flame at the jet exit Reynolds number (Red) of 8000. The 3D data enable the computation of the nine components of the velocity gradient tensor and other important kinematic quantities. The volumetric PIV is combined with single-shot simultaneous OH PLIF imaging to mark the instantaneous reaction zone at one plane in the reconstructed volume. The combined datasets enable the investigation of the relationship between the reaction zone and the fully-3D representations of strain, vorticity, kinetic energy dissipation and dilatation, and of the impact of heat release on the structure of turbulence. In this Red = 8000 flame, it is observed that sheet-like layers of vorticity and dissipation tend to coincide and are aligned with the OH layers, an effect that is believed to be due to the stabilizing effect of heat release on this relatively low Reynolds number jet flame. Furthermore, the spatial organization of the strain field is predominantly driven by the presence of the flame rather than turbulence. Finally, intense dissipation is mostly due to the laminar shear caused by the presence of the flame rather than to the strain generated by vortical structures as typically observed in nonreacting jets.

Combustion of Laminar Non-Premixed Acetylene Jet at Two Different Altitudes

In this article, the influence of pressure variation due to altitude change on flames is studied. A series of largely buoyancy-driven acetylene jet fire experiments was conducted at two places in China, namely Lhasa and Hefei, with altitudes of 3658 m and 50 m, respectively. A range of parameters, including flame height, flame and plume centerline temperatures, transmittance through smoke, and irradiance were measured and compared. The flame height at low altitude is higher than that at high altitude. The centerline temperatures of the flame and plume regions exhibit opposite variations with pressure. The soot mass concentration resolved by transmittance measurement follows a weaker power-law dependence on pressure under substandard atmospheric pressure than under pressure greater than standard atmospheric pressure. The radiation heat flux produced from flames of the same burning rate displays a positive proportionality to pressure.

Lifted flame behavior of a non-premixed oxy-methane jet in a lab-scale slot burner

The characteristics of oxy-methane flames in atmospheric conditions (298K and 1bar) were studied in a lab-scale slot burner. As a diagnostic tool, OH∗, CH∗, and C2∗ measurements were used to analyze the flame structure. To estimate the flame stabilization, the flow velocity was varied from uF=7–50m/s for methane gas and uOx=10–170m/s for oxygen gas. From the experimental results, the flame stability was enhanced when the slot width became narrower. The flame length increased with increasing fuel quantity, while the lift-off height increased with increase in the fuel jet and oxygen velocity. Using the concept of effective diameter, an empirical formula for flame length was suggested for use in a combustor design.

Simultaneous LII and PIV measurements in the soot formation region of turbulent non-premixed jet flames

An experimental study was performed to investigate the soot–turbulence interaction in the soot-formation region of turbulent non-premixed co-flowing ethylene/N2 jet flames. Simultaneous velocity and soot volume-fraction (fv) fields were obtained using two-component particle image velocimetry and laser-induced incandescence, respectively. Measurements were made for jet exit Reynolds numbers between 8500 and 12,300, and the measurement location was 10 jet diameters downstream, near the beginning of the yellow luminous region where soot is first formed. In agreement with previous studies, the peak mean fv in the production region is inversely related to the bulk strain rate. The simultaneous data show that soot is formed to the inside of the stoichiometric surface (inferred from stoichiometric velocity), but the formation region moves outside to regions of lower velocity and strain rate as the bulk strain rate is increased. Soot structures form in low strain rate regions, but their upstream edge is seen to become stretched out and aligned at a preferred angle (near 45 degrees) owing to alignment with the instantaneous principal extensive strain rate. Statistical analysis shows that the soot exists, on average, in fluid with axial velocity of about 3m/s and strain rate of 700s−1, regardless of the jet exit velocity. The radial profiles of the covariance between fv and radial velocity are consistent with a model where the soot is formed at a preferred radial location (near the reaction zone) and then is transported by turbulent fluctuations to regions of lower fv.

On the structure of the near field of oxy-fuel jet flames using Raman/Rayleigh laser diagnostics

An experimental study on turbulent non-premixed jet flames is presented with focus on CO2-diluted oxy-fuel combustion using a coflow burner. Measurements of local temperatures and concentrations of the main species CO2, O2, CO, N2, CH4, H2O and H2 were achieved using the simultaneous line-imaged Raman/Rayleigh laser diagnostics setup at Sandia National Laboratories. Two series of flames burning mixtures of methane and hydrogen were investigated. In the first series, the hydrogen molar fraction in the fuel was varied from 37% to 55%, with a constant jet exit Reynolds number ReFuel of 15,000. In the second series the jet exit Reynolds number was varied from 12,000 to 18,000, while keeping 55% H2 molar fraction in the fuel. Besides local temperatures and concentrations, the results revealed insights on the behaviour of localized extinction in the near-field. It was observed that the degree of extinction increased as the hydrogen content in fuel was decreased and as the jet Reynolds number was increased. Based on the distribution of the temperature, a fully burning probability index able to quantify the degree of extinction along the streamwise coordinate was defined and applied to the present flame measurements. A comparison of measured conditional mean of mass fractions and laminar flame calculations underlined the significant level of differential diffusion in the near-field that tended to decrease farther downstream. The results also showed high local CO levels induced by the high content of CO2 in the oxidizer and flame products. A shift of maximum flame temperature was observed toward the rich side of the mixture fraction space, most likely as a consequence of reduced heat release in the presence of product dissociation. Main characteristics of laser Raman scattering measurements in CO2-diluted oxy-fuel conditions compared to air-diluted conditions are also highlighted. Most data, including scalar fluctuations and conditional statistics are available upon request.

Turbulence-radiation interactions in large-eddy simulations of luminous and nonluminous nonpremixed flames

Turbulence-radiation interactions (TRI) are explored in large-eddy simulations (LES) of luminous and nonluminous nonpremixed jet flames. The simulations feature a transported filtered density function (FDF) method for subfilter-scale fluctuations in composition and temperature, and a fully coupled photon Monte Carlo (PMC) method for radiative transfer with line-by-line (LBL) spectral resolution. The model is exercised to isolate and quantify individual contributions to TRI for conditions that range from small optically thin flames to relatively large optically thick flames, including spectral molecular gas radiation and broadband soot radiation. The results provide new physical insight into TRI and guidance for modeling. In all cases, emission TRI are responsible for a significant fraction of the radiative emission, and that fraction increases with increasing optical thickness. For simulations where 84% of the turbulence kinetic energy is resolved, contributions of subfilter-scale fluctuations to emission TRI exceed those of resolved-scale fluctuations. The largest contributions to emission TRI are the absorption coefficient–temperature correlation and the temperature self-correlation. Absorption TRI are evident only for relatively high optical thicknesses. In all cases, the contributions of subfilter-scale fluctuations to absorption TRI are negligible.

Comparisons of Lagrangian and Eulerian PDF methods in simulations of non-premixed turbulent jet flames with moderate-to-strong turbulence-chemistry interactions

Transported probability density function (PDF) methods have been applied widely and effectively for modelling turbulent reacting flows. In most applications of PDF methods to date, Lagrangian particle Monte Carlo algorithms have been used to solve a modelled PDF transport equation. However, Lagrangian particle PDF methods are computationally intensive and are not readily integrated into conventional Eulerian computational fluid dynamics (CFD) codes. Eulerian field PDF methods have been proposed as an alternative. Here a systematic comparison is performed among three methods for solving the same underlying modelled composition PDF transport equation: a consistent hybrid Lagrangian particle/Eulerian mesh (LPEM) method, a stochastic Eulerian field (SEF) method and a deterministic Eulerian field method with a direct-quadrature-method-of-moments closure (a multi-environment PDF-MEPDF method). The comparisons have been made in simulations of a series of three non-premixed, piloted methane–air turbulent jet flames that exhibit progressively increasing levels of local extinction and turbulence-chemistry interactions: Sandia/TUD flames D, E and F. The three PDF methods have been implemented using the same underlying CFD solver, and results obtained using the three methods have been compared using (to the extent possible) equivalent physical models and numerical parameters. Reasonably converged mean and rms scalar profiles are obtained using 40 particles per cell for the LPEM method or 40 Eulerian fields for the SEF method. Results from these stochastic methods are compared with results obtained using two- and three-environment MEPDF methods. The relative advantages and disadvantages of each method in terms of accuracy and computational requirements are explored and identified. In general, the results obtained from the two stochastic methods (LPEM and SEF) are very similar, and are in closer agreement with experimental measurements than those obtained using the MEPDF method, while MEPDF is the most computationally efficient of the three methods. These and other findings are discussed in detail.

Modeling scalar dissipation and scalar variance in large eddy simulation: Algebraic and transport equation closures

Scalar dissipation rates and subfilter scalar variances are important modeling parameters in large eddy simulations (LES) of reacting flows. Currently available models capture the general behavior of these parameters, but these models do not always perform with the degree of accuracy that is needed for predictive LES. Here, two direct numerical simulations (DNS) are used to analyze LES dissipation rate and variance models, and to propose a new model for the dissipation rate that is based on a transport equation. The first DNS that is considered is a non-premixed auto-igniting C2H4 jet flame simulation originally performed by Yoo et al. [Proc. Combust. Inst. 33, 1619–1627 (2011)]10.1016/j.proci.2010.06.147. A LES of this case is run using algebraic models for the dissipation rate and subfilter variance. It is shown that the algebraic models fail to adequately reproduce the DNS results. This motivates the introduction of a transport equation model for the LES dissipation rate. Closure of the equation is addressed by formulating a new adapted dynamic approach. This approach borrows dynamically computed information from LES quantities that, unlike the dissipation rate, do not reside on the smallest flow length scales. The adapted dynamic approach is analyzed by considering a second DNS of scalar mixing in homogeneous isotropic turbulence. Data from this second DNS are used to confirm that the adapted dynamic approach successfully closes the dissipation rate equation over a wide range of LES filter widths. The first reacting jet case is then returned to and used to test the LES transport equation models. The transport equation model for the dissipation rate is shown to be more accurate than its algebraic counterpoint, and the dissipation rate is eliminated as a source of error in the transported variance model.

One-dimensional-turbulence simulation of flame extinction and reignition in planar ethylene jet flames

A series of three simulations of temporally evolving, planar, nonpremixed ethylene jet flames are performed using the one-dimensional-turbulence (ODT) model. The simulations are performed at a fixed Reynolds number, but varying Damköhler numbers, under conditions that result in significant flame extinction and reignition. Results are compared to corresponding direct numerical simulations (DNSs), which exhibit 40, 70, and nearly 100% peak flame extinction among the three cases. The planar, temporal configuration is ideal for comparison and validation of the ODT model, and identical thermodynamic, transport, and kinetic models were used. The ODT model captures the jet evolution and heat release effects. Line of sight profiles are qualitatively similar between the ODT and DNS. Good agreement is found for scalar dissipation statistics. While ODT captures the flame extinction process, the level of flame reignition is underpredicted, and conditional mean temperature profiles are depressed during reignition compared to the DNS. This is in contrast to previous ODT studies using CO/H2 mixtures (syngas). As a one-dimensional model, ODT is unable to capture multi-dimensional edge flame propagation during reignition, but is able to capture reignition via flame folding. Given the fidelity of the fine-scale transport and reaction processes built into ODT, and good flow modeling observed, more reduced combustion models will be challenged to capture nonpremixed flame reignition.

Broadband Combustion Noise Simulation of Open Non-Premixed Turbulent Jet Flames

Numerical broadband combustion noise simulations of open non-premixed turbulent jet flames applying the Random Particle-Mesh for Combustion Noise (RPM-CN) approach are presented. The RPM-CN approach is a hybrid Computational Fluid Dynamics/Computational Aeroacoustics (CFD/CAA) method for the numerical simulation of turbulent combustion noise, based on a stochastic source reconstruction in the time domain. The combustion noise sources are modeled on the basis of statistical turbulence quantities, for example achieved by a Reynolds averaged Navier-Stokes (RANS) simulation, using the Random Particle-Mesh (RPM) method. RPM generates a statistically stationary fluctuating sound source that satisfies prescribed one- and two-point statistics which implicitly specify the acoustic spectrum. Subsequently, the propagation of the combustion noise is computed by the numerical solution of the Linearized Euler Equations (LEE). The numerical approach is applied to the DLR-A, the DLR-B and the H3 flames. The open non-premixed turbulent jet flames differ in the mean jet exit velocity, therefore in their respective Reynolds number, and in the fuel composition. Computed radial profiles of the reacting flow field are compared to experimental data and discussed. Computed sound pressure level spectra of the DLR-A and DLR-B flames and acoustic intensity level spectra of the H3 flame at different microphone locations are presented and compared to measurements.