Lift-off Height Research Articles

Numerical and experimental study of flame propagation and quenching of lean premixed turbulent low swirl flames at different Reynolds numbers

This paper presents a joint experimental and large eddy simulation (LES) study of lean premixed low swirl stabilized methane/air flames at different Reynolds numbers (Re∼20,000–100,000). The aims are to investigate the sensitivity of the structures and dynamics of low swirl flames to the inflow boundary conditions and to evaluate the capability of an LES flamelet model in predicting the stabilization and local extinction of the flames. Chemiluminescence measurements are carried out for Re=20,000–50,000 and further detailed oxygen concentration and temperature fields are measured using rotational coherent anti-Stokes Raman spectroscopy (RCARS) for Re=20,000 and 30,000 along the centerline of the burner and at various radial positions at different heights above the burner. The data are used first for validation of the combustion LES model employed in the numerical simulations, and then the RCARS and LES results are used to delineate the effect of ambient air entrainment on the flame structure at various burner exit velocities. A three-scalar flamelet model based on a level-set G-equation shows excellent predictions of the lift-off positions and the structures of the flames, including quenching at the trailing edge of the flame. The results show that the flame lift-off height varies only slightly when the burner exit velocity is increased, which is consistent with a shear-layer flame stabilization mechanism reported previously. The volume of the flame decreases substantially with increasing burner exit velocity at relatively low Reynolds numbers, as a result of flame quenching at the trailing edge of the flame caused by entrainment of the ambient air into the fuel/air stream and the flame itself. At high Reynolds numbers the flame structures become fairly self-similar with the flame volume nearly independent of the Reynolds number.

An investigation of methane and propane vertical flares

Flame lengths, temperature, species and lift-off heights have been experimentally and numerically investigated and compared for buoyant methane and propane jet diffusion flames at Reynolds number of 5700. The flow field has been modeled using the Reynolds-Averaged Navier–Stokes equation incorporating the k–ε realizable turbulence closure model. Combustion was modeled using the unsteady Eulerian-flamelet model based on the mixture fraction approach and the heat loss by radiation was accounted for using the Discrete Ordinates Method. The GRI mech. 3.0 and the CRECK reaction mechanisms were used to model the kinetics of the methane and propane reactions, respectively. Comparison of the predicted flame length and temperature revealed good agreement with experimental data. Post-flame measurements of NOx and CO revealed greater quantities of both pollutants in the methane flame. Furthermore, investigation of the effect of the burner nozzle thickness on the flame lift-off heights showed that the lift-off height decreased as the nozzle thickness was increased, with the methane flame displaying higher lift-off heights.

Chemical Flame Length of a Methane Jet into Oxidant Stream

The present paper reports a systematic study on different methods of obtaining the chemical length of a turbulent non-premixed CH 4 jet flame. The study is performed based on numerical simulations by the RANS modeling over wide ranges of the fuel jet velocity (25 ∼ 100 m/s), the coflow oxygen concentration (6 ∼ 100 % by volume) and the coflow temperature (300 ∼ 2000 K). To validate the modeling, some calculations are compared with previous measurements of Dally et al. (Proc. Combust. Inst. 29, 1147–1154, 2002) and Kim et al. (Energy Fuels 21, 1459–1467, 2007). Primary results reveal that, among seven different definitions of chemical flame length, the best performance is made by those defined from the yield ratio of CO to CO 2 at y C O / y C O2= 0.02 and the carbon monoxide ratio at R C O = X C O /X C O m a x = 0.01. It is also found that only use of these two definitions exhibit the flame lift-off behavior and to gain the lift-off height properly. Besides, this paper explains the importance of taking the near-zero contour of the CO concentration as the base on which the border of the complete combustion region is defined for chemical flame length.

Experimental investigations and large-eddy simulation of low-swirl combustion in a lean premixed multi-nozzle combustor

This paper presents laser diagnostic experiments and large-eddy simulations (LES) of low-swirl lean premixed methane/air flames in a multi-nozzle combustor including five nozzles with the same structure. OH planar laser-induced fluorescence is used to observe flame shapes and identify main reaction zones. NOx and CO emissions are also recorded during the experiment. The flows and flames are studied at different equivalence ratios ranging from 0.5 to 0.8, while the bulk inlet velocity is fixed at 6.2 m/s. Results show that the neighboring swirling flows interact with each other, generating a highly turbulent interacting zone where intensive reactions take place. The flame is stabilized above the nozzle rim, and its liftoff height decreases with increasing equivalence ratio. The center flow is confined and distorted by the neighboring flows, resulting in instabilities of the center flame. Mean OH radical images reveal that the center nozzle flame is extinguished when equivalence ratio is equals to 0.5, which is successfully predicted by LES. In addition, NOx emissions show log-linear dependency on the adiabatic flame temperature, while the CO emissions remain lower than 10 ppm. NOx emissions for multi-nozzle flame are less sensitive to the flame temperature than that for single nozzle. These results demonstrate that the low-swirl multi-nozzle concept is a promising solution to achieve stable combustion with ultra-low emissions in gas turbines.

Doubly Conditional Source-term Estimation (DCSE) applied to a series of lifted turbulent jet flames in cold air

The objective of the present study is to evaluate the applicability of a new partially premixed combustion model, Doubly Conditional Source-term Estimation (DCSE), applied to a series of lifted jet flames using Reynolds Averaged Navier Stokes (RANS) calculations. The partially premixed implementation involves double conditioning on two variables, mixture fraction and progress variable. The present work represents the first application of DCSE for lifted turbulent flames. Conditionally averaged chemical source terms are closed by conditional averaged scalars which are obtained by inverting an integral equation. Three different jet conditions are simulated, over a range of Reynolds number between 7,000 and 19,500. Two-equation k–ε turbulence model is used to solve the flow field. The predicted mean centerline and radial profiles of methane concentration are in good agreement with the experimental values. The present DCSE calculations are successful in predicting the lift-off heights and reproducing many aspects of the flame structure in agreement with the experimental observations. The stabilization point is located on the fuel lean side, the mixture fraction isolines are deflected around the flame base, significant entrainment of air by the fuel jet upstream of the flame stabilization point is observed and the calculated flame width becomes larger with increasing Reynolds number. Some discussion is given related to triple flames. Future work is suggested related to the scalar dissipation rate modelling and DCSE implementation in Large Eddy Simulation (LES).

Critical observations on the modeling of nonreacting and reacting diesel sprays

This paper begins by discussing the structure of nonreacting and reacting diesel sprays. After a general discussion of approaches to model nonreacting diesel sprays, results from some recent work employing the Reynolds-averaged Navier–Stokes equations to model the nonreacting spray are presented. It is shown through detailed comparison with measured results that, under conventional high-pressure, high-temperature chamber and high-pressure injection conditions, the vaporizing diesel spray behaves like a gas jet and can be modeled as such without noticeable loss of accuracy in reproducing the structure of the spray. Moving on to consider a reacting diesel spray, turbulence–chemistry interaction models for the reacting spray are reviewed. Challenges in determining the suitability of one model over another, based on information available in the literature, are highlighted. The Reynolds-averaged Navier–Stokes simulation results of reacting diesel sprays in which an unsteady-flamelet-generated manifold model is employed for turbulence–chemistry interactions are discussed in detail. It is shown that the model, like others reported in the literature, can predict the ignition delay and the flame lift-off heights with reasonable accuracy. The model has also been extended to compute the concentrations of soot and nitrogen oxides in reacting diesel sprays. Nitrogen oxides are modeled using the mechanism from GRI-Mech 3.0 and soot is modeled using a kinetic mechanism coupled with a tracer particle approach to estimate the residence times within the jet. As part of this review, other recent studies on modeling pollutants in the reacting diesel spray are also reviewed in this work. Finally, recent studies employing large-eddy simulations of diesel sprays are reviewed. Initial simulations of the reacting diesel spray using a large-eddy simulation approach coupled with an unsteady-flamelet-generated manifold model are presented. The paper closes with a summary and suggestions for further work in modeling diesel sprays.

Flame characteristics of a non-premixed oxy-fuel jet in a lab-scale furnace

The effect of varying the fuel and oxidizer composition on flame characteristics in a non-premixed oxy-methane flame was experimentally investigated in a lab-scale furnace with a slot-type burner. The fuel composition was varied in the range of XF,CH4 = 70∼100%, XF,H2 = 0∼15%, XF,CO = 0∼10%, and XF,CO2 = 0∼30% while the oxidizer composition was changed in the range of XOx,O2 = 70∼100 and XOx,CO2 = 0∼30%. Flow velocity at the nozzle exit was fixed at uF = 25 m/s for the fuel jet and uOx = 25 m/s for the oxidizer. The objective of the current study was to investigate the effect of varying the fuel and oxidizer composition on flame stabilization, flame luminescence, flame slope, liftoff height, flame length, and furnace inside temperature in a non-premixed oxy-fuel flame. The addition of H2 and CO (carbon monoxide) to a fuel jet had the positive effect of broadening the flammable limit. The hydroxide radical chemiluminescence (OH*) intensity was related to adiabatic temperature. The flame slope was affected by global equivalence ratio and the mass flow rate between the fuel and the oxidizer. The liftoff height decreased with the increase in edge flame velocity while the flame length decreased with the increase in the Péclet number of unburned gas.

Combustion of hythane diluted with CO2

With increasing concern about energy shortage and environmental protection, improving engine fuel economy and reducing exhaust emissions have become major research topics in combustion and engine development. Hythane (a blend of hydrogen H2 and natural gas NG) has generated a significant interest as an alternative fuel for the future. This paper describes an experimental study of the effects of CO2 addition on the stability of a turbulent jet diffusion NG-H2 flame. The mole fraction of hydrogen (% H2) in NG-H2 mixture was varied from 0% to 50%. The equivalence ratio of the hythane/CO2/air mixture was kept at stoichiometry. The results show that the lift-off height increases with the addition of CO2 at various % H2 content in hythane. However, we observe that with 20% H2, we can obtain a stable flame diluted with 40% CO2, while for 0% H2, the flame is blown out above 20% CO2. This means that the limits of flame blowing out are pushed with the additions of H2. Moreover, the results show that for %H2 content in NG-H2 fuel up to 10%, the addition of CO2 could produce lifted flame if the % CO2 is low. At higher % CO2 dilution, flame would remain attached until blow-out. This is mainly due to the fact that the dilution leads to ejection velocities very high but reactivity of the mixture does not change so the flame tends to stretch.

音響振動がTriple Flameの曲率および浮き上がり高さに及ぼす影響

音響振動がTriple Flameの曲率および浮き上がり高さに及ぼす影響

Patterns of barbell acceleration during the snatch in weightlifting competition

The purpose of this study was to determine the association between weightlifting performance and vertical barbell acceleration patterns. Barbell kinematic time-series data were tracked from 18 snatches from six weightlifters during a regional weightlifting competition. These data were used to calculate vertical barbell accelerations. Time-series data were normalised to 100% of lift phase, defined as the time interval between barbell lift-off and maximum height of the barbell during each snatch lift. The time-series data were then entered into a pattern recognition algorithm that extracted principal patterns and calculated principal pattern scores. Body mass-normalised lift weight, which was used to quantify weightlifting performance, was significantly correlated (r = 0.673; P = 0.033) with a pattern that captured a difference in peak vertical barbell acceleration between the transition and the second pull phase. This correlation indicated that barbell acceleration profiles of higher weight snatch lifts were characterised by smaller decreases in acceleration during the second knee bend and smaller peak acceleration during the second pull phase. Weightlifting coaches and sports scientist should monitor and track vertical acceleration of the barbell, with focus on acceleration profiles that limit (1) deceleration during the transition phase between the first and second pull and (2) peak acceleration during the second pull phase of the snatch.

Behaviors of tribrachial edge flames and their interactions in a triple-port burner

In a triple-port burner, various non-premixed flames have been observed previously. Especially for the case with two lifted flames, such configuration could be suitable in studying interaction between two tribrachial flames. In the present study, the flame characteristics have been investigated numerically by adopting a reduced kinetic mechanism in the triple-port burner. Four different types of flame configurations, including two attached flames, inner lifted/outer attached flames, inner attached/outer lifted flames, and twin lifted flames, were successfully simulated depending on the flow conditions. The representative edge propagation speed of a single lifted flame or an upstream lifted flame in the case of twin lifted flames increased as the liftoff height became higher. In the twin lifted flames, the inner lifted flame was affected appreciably when the other flame was located further upstream such that the lifted flame located further downstream encountered the axial velocity acceleration induced by the gas expansion from the lifted flame located upstream, while thermal effects were not observed since the temperature of the incoming flow toward the lifted flame was not affected. A unique flip–flop behavior between the inner and outer flames, observed experimentally previously, was successfully captured in the simulation such that the inner lifted flame became attached to the nozzle as the liftoff height of the outer lifted flame grew higher with an increase in the outer air velocity.

Simulation of Non-Autoignited and Autoignited Laminar Non-Premixed Jet Flames of Syngas in Heated Coflow Air

The characteristics of non-autoignited and autoignited flames of syngas (CO/H2 mixture) fuels with heated coflow air have been investigated numerically. The simulations accounting for the differential diffusion have been performed by adopting several kinetic mechanisms to test the models’ ability in predicting the flame behaviors observed previously. The results agreed well with the observed nozzle-attached flame characteristics in case of non-autoignited flames. For autoignited lifted flames in high temperature regime, a unique autoignition behavior can be predicted having HO2 and H2O2 radicals in a broad region between the nozzle and stabilized lifted flame edge. Species’ profiles as a function of temperature showed a similarity with a homogeneous autoignition case. The autoignited liftoff height was correlated well in terms of the jet velocity multiplied by the square of ignition delay time, in agreement with previous experimental data. The simulation results tested with various kinetic mechanisms, however, could not predict the non-monotonic liftoff height behavior with jet velocity for autoignited flames observed experimentally at relatively low temperature. The initial temperature in the simulation should be higher at about 80 K as compared to experimental conditions to have comparable liftoff heights. The result suggested a necessity to improve kinetic mechanisms at atmospheric pressure conditions.

Effects of coflow temperature and composition on ethanol spray flames in hot-diluted coflow

Ethanol pressure-swirl sprays issuing in a hot-diluted oxidizer coflow with different temperature and composition were studied. The bulk coflow temperature was varied together with the oxygen volume fraction. The bulk coflow temperature was changed from 1480K to 1225K and the oxygen volume fraction from 7.1% to 10.1%. The liquid mass flow rates were chosen to yield spray flames with nearly identical Weber number. Laser Doppler anemometry, phase Doppler anemometry and coherent anti-Stokes Raman scattering were applied in the spray region and the coflow inlet. The current measurements provide a thorough description of the spray structure, droplet dispersion and gas temperature fields as well as a comprehensive database useful for validation of numerical models. Visual observations of the flame luminescence reveal that the lift-off height depends on the liquid mass flow rates as well as the coflow conditions. The lift-off height is shown to increase for lower coflow temperatures and higher liquid mass flow rates. It is found that lift-off behavior depends on the droplet convective, vaporization and chemical time scales prior to ignition. Phase Doppler anemometry results indicate that the droplet mean size and velocity distribution close to the atomizer are not influenced by the coflow conditions. A flame-front develops at the outer region of the spray where a low density of large droplets are present. A significant number of peak temperatures samples above 2000K is observed at this location. Decrease of the coflow temperature leads to a reduction of the local peak temperatures. Closer to the center axis, the local mixture composition becomes increasingly rich and the heat-release is lower than in the outer region.

Predictions of lift-off height of turbulent methane and propane flames issuing in cold surroundings using conditional moment closure coupled with an extinction model

Recently, an extinction model was successfully integrated with a conditional moment closure (CMC) approach to predict lift-off height accurately for the lifted jet diffusion flame discharging to hot surroundings Roy et al. (2014). This method, referred as CMCE, is a low-cost model based on the RANS and first-order CMC approach. The proposed methodology may not be superior compared to involved LES and second-order CMC combination, but enables to simulate industrial problems like gas turbine processes in a reasonable CPU time. Though the credibility of this novel approach was well established by validating its predictions against the experimental data, the applicability of this method and constants involved in this model, for other test cases was questionable. In this brief communication, CMCE model is further validated against lift-off studies of two different fuels issuing in cold surroundings, without changing the model constants used in the earlier work. The results are encouraging and a good match is found for lift-off heights for both the test cases and for varying jet velocities. Large variations are found between the predictions of lift-off heights by CMC and CMCE models.

On the Hysteresis Phenomenon of Turbulent Lifted Diffusion Methane Flame

This paper reports an experimental investigation on the flow characteristics upstream of a lifted turbulent diffusion flame in the presence of a coflow. Three fuel nozzles made of a long pipe with different outlet geometry were examined. One pair of these nozzles has the same orifice diameter but different normalized lip thickness, and another pair has the same normalized lip thickness but different orifice diameter. The strength of the co-airflow was also varied to assess its impact on the liftoff height of the jet diffusion flame. Previously published studies reported the existence of a hysteresis phenomenon in the liftoff height of a turbulent diffusion flame in the presence of a high co-airflow. That is, as the fuel velocity decreases, the lifted flame base would first move upstream (toward the burner) to a local minima followed by a downstream movement before its reattachment. The results of the present study, however, showed that such a phenomenon does not appear for a fuel pipe having a very small lip thickness. The present results also revealed that in the hysteresis region, the flame base sits where the turbulence intensity experiences its local maxima in the upcoming unburnt mixture. This corroborates the premixed stability theory which is based on turbulence intensity. Based on this, a correlation was found between the flame liftoff height in the hysteresis region and the fuel and co-airflow velocity at the nozzle exit. This relationship predicts successfully the liftoff height trend as a function of the fuel jet and co-airflow velocity and nozzle geometry. Away from the hysteresis region, however, the flame base location tends more toward the outside of the local turbulence intensity maxima. This indicates the limitations of the premixed stability theory in predicting the flame behavior in this region where the effect of the flow large-scale structures becomes important.

RANS modelling of a lifted H2/N2 flame using an unsteady flamelet progress variable approach with presumed PDF

An unsteady flamelet/progress variable (UFPV) approach is used to model a lifted H2/N2 flame in a RANS framework together with presumed PDF. We solve the unsteady flamelets both in physical space and in mixture fraction space. We show that in the former case, the scalar dissipation rate profile strongly varies in time (while it is assumed to be fixed in time in the latter). However, this does not result in significant qualitative differences in the corresponding flamelet libraries. The progress variable is carefully defined, including both the main combustion product (H2O) and a key radical species in ignition process (HO2). The presumed-PDF model is proposed in terms of the non-normalised progress variable, without assuming its statistical independence with mixture fraction. We introduce a modelled transport equation for the mean progress variable which is consistent with the basic underlying UFPV assumption, derived from the Lagrangian flamelet model. The influence of different model parameters on the results for the mean temperature and mean species mass fractions and their fluctuations is discussed. Good results are obtained for the conditions of the considered lifted flame where detailed experimental data is available. However, at low coflow temperature the modelled flame lift-off height is shorter than expected.

Simulation of turbulent lifted methane jet flames: Effects of air-dilution and transient flame propagation

Turbulent lifted methane jet flames with various air-dilution levels and a range of inlet velocities are simulated. A partially premixed combustion model based on premixed flamelets with presumed joint Probability Density Function (PDF) is used. The joint PDF is obtained using a copula to include the statistical correlation between mixture fraction, Z, and progress variable, c. The non-premixed combustion effect is included using a simple algebraic model. Both steady and unsteady RANS simulations are performed. The steady simulations show that the computed lift-off heights agree well with measured values for a wide range of jet velocities and air-dilution level. Both of the Z–c correlation and non-premixed combustion effects are found to be important to get the correct lift-off height. Their individual and combined effects are analysed systematically. The unsteady RANS results indicate that multi-stage flame development, namely the initial expansion, flame brush development, its propagation and final stabilisation, is captured reasonably well in simulations. The various stages of temporal evolution of the flame brush edge is captured well and the agreement with experimental measurements is good.

Lengths of Lifted Laminar Flames Under Vertical Magnetic Field Gradient

The effect of a vertical magnetic field gradient on stoichiometric lengths of methane-air diffusion lifted flames was numerically investigated. The study was developed for understanding the evolution of these lengths when coaxial jet air velocity was varied keeping constant fuel velocity. As the air jet velocity is increased, gradient in mixture fraction and liftoff height increase and flame length decreases. Calculated reacting flow results of liftoff and flame length are presented for two cases, with and without magnetic field. For the same initial flow conditions, the application of the negative magnetic field gradient results in an increase of the stoichiometric flame lengths and a reduction of liftoff heights. For the lifted flame under magnetic field gradient, the Froude number was modified to take into account the magnetic buoyancy effect and it is shown that the correlation proposed by Altenkirch et al. (1976) can still be applied to represent lengths of lifted flames.

Stabilization and liftoff length of a non-premixed methane/air jet flame discharging into a high-temperature environment: An accelerated transported PDF method

A particle-based transported probability density function (PDF) method with a novel chemistry acceleration technique is developed in this work. The technique is based on the chemistry coordinate mapping (CCM) approach that was proposed in our previous works for accelerating direct numerical simulations (DNS) of partially premixed combustion. The method is first validated using Sandia flames D and F. It is shown that PDF-CCM results converge toward those obtained without CCM as phase-space resolution increases. PDF-CCM is then applied to simulate methane/air lifted jet flames in vitiated coflow reported in experiments by Cabra et al. (2005). It is shown that combustion is initiated in the form of auto-ignition in very fuel-lean gases where the gas velocity is low (the residence time is long) and the gas temperature is high (the ignition delay time is short). The ignition delay of the mixture below the liftoff position scales well with the liftoff height at different coflow temperature conditions. The combustion process above the liftoff height can develop into different modes depending on the coflow temperature. For high-temperature coflow, a premixed-burned combustion is formed above the liftoff height, which involves fuel-lean to fuel-rich burning modes; a triple-flame structure eventually is formed a few nozzle diameters above the liftoff position. For low-temperature coflow, the ignition delay and the liftoff height are sufficiently large to allow premixing between fuel and oxidizer before the onset of high-temperature combustion; in this case, a lean-to-stoichiometric premixed burn combustion is established downstream of the liftoff height, and no obvious triple-flame structure is formed at this condition.

A simplified CMC approach based on tabulated reaction rates applied to a lifted methane–air jet flame

This paper explores the possibility to close the conditional reaction rate in the Conditional Moment Closure (CMC) turbulent combustion model using tabulated chemistry. This combination, called TCMC (Tabulated CMC), is based on the resolution of transport equations for the conditional mean and variance of the progress variable mass fraction Yc=YCO+YCO2 in which chemical source terms are taken from a look-up table built with homogeneous reactor calculations. First and second order TCMC closures (respectively TCMCI and TCMCII) are proposed and are compared on the lifted methane–air flame experiment of Cabra et al. (2005) [2]. In this configuration, it is shown that the first order model is unable to predict a lifted flame, whereas TCMCII gives correct results in terms of mean composition and temperature. These results are found comparable to those obtained by Michel et al. (2009) [7] with the ADF-PCM (Approximated Diffusion Flame Presumed Conditional Moment) flamelet model based on the same progress variable and tabulated kinetics. On the contrary, conditional flame profiles are different in the lift-off zone, and lift-off height predictions are improved with TCMCII. Finally, the CPU time of TCMCII is found less than ten times that of ADF-PCM, using a unique mesh for all transport equations, which makes TCMC a realistic alternative to tabulated flamelet models.