Hot Coflow Research Articles

Assessment of On-the-Fly Chemistry Reduction and Tabulation Approaches for the Simulation of Moderate or Intense Low-Oxygen Dilution Combustion.

The current paper focuses on the numerical simulation of the Delft jet in hot co-flow (DJHC) burner, fed with natural gas and biogas, using the eddy dissipation concept (EDC) model with dynamic chemistry reduction and tabulation, i.e., tabulated dynamic adaptive chemistry (TDAC). The central processing unit (CPU) time saving provided by TDAC is evaluated for various EDC model constants and chemical mechanisms of increasing complexity, using a number of chemistry reduction approaches. Results show that the TDAC method provides speed-up factors of 1.4–2.0 and more than 10 when using a skeletal mechanism (DRM19) and a comprehensive kinetic mechanism (POLIMIC1C3HT), respectively. The directed relation graph with error propagation (DRGEP), dynamic adaptive chemistry (DAC), and elementary flux analysis (EFA) reduction models show superior performances when compared to other approaches, such as directed relation graph (DRG) and path flux analysis (PFA). All of the reduction models have been adapted for run-time reduction. Furthermore, the contribution of tabulation is more important with small mechanisms, while reduction plays a major role with large ones.

Detailed investigation of NO mechanism in non-premixed oxy-fuel jet flames with CH4/H2 fuel blends

This study systematically investigates the detailed mechanism of nitrogen oxides (NOx) in CH4 and CH4/H2 jet flames with O2/CO2 hot coflow. After comprehensive validation of the modeling by experiments of Dally et al. [Proc. Combust. Inst. 29 (2002) 1147–1154]; the effects of CO2 replacement of N2, mass fraction of oxygen in the coflow (YO2), and mass fraction of hydrogen in the fuel jet (YH2) on NO formation and destruction are investigated in detail. For methane oxy-fuel combustion, the NNH route is found to control the NO formation at YO2 ≤ 3%, while both NNH and N2O-intermediate routes dominate the NO production at 3% < YO2 < 10%. When YO2 ≥ 10%, NO is obtained mainly from thermal mechanism. Moreover, in the oxy-combustion of methane and hydrogen fuel blends with YO2 = 3%, with hydrogen addition the contribution of the NNH and prompt routes increases, while that of the N2O-intermediate route decreases. Furthermore, the chemical effect of CO2 is significant in reducing NO in both oxy-combustion of methane with YO2 ≤ 3% and combustion of methane and hydrogen fuel blends with YH2 ≤ 10%.

Structural differences of ethanol and DME jet flames in a hot diluted coflow

This study compares the flame structure of ethanol and dimethyl ether (DME) in a hot and diluted oxidiser experimentally and computationally. Experiments were conducted on a Jet in Hot Coflow (JHC) burner, with the fuel jet issuing into a 1250-K coflow at three oxygen levels. Planar measurements using OH-LIF, CH2O-LIF, and Rayleigh scattering images reveal that the overall spatial distribution and evolution of OH, CH2O, and temperature were quite similar for the two fuels. For both the ethanol and the DME flames, a transitional flame structure occurred as the coflow oxygen level increased from 3% to 9%. This indicates that the flames shift away from the MILD combustion regime. Reaction flux analyses of ethanol and DME were performed with the OPPDIF code, and ethane (C2H6) was also included in the analyses for comparison. These analyses reveal that the H2/O2 pathways are very important for both ethanol and DME in the 3% O2 cases. In contrast, the importance of fuel-specific reactions overtakes that of H2/O2 reactions when fuels are burnt in the cold air or in the vitiated oxidant stream with 9% O2. Unsteady laminar flamelet analyses were also performed to investigate the ignition processes and help interpret experimental results. Flamelet equations were solved in time and mixture fraction field, which was provided by non-reactive Large-Eddy Simulation (LES).

What Differences Does Large Eddy Simulation Find among Traditional, High-Temperature, and Moderate or Intense Low Oxygen Dilution Combustion Processes of a CH4/H2 Jet Flame in Hot Oxidizer Coflow?

Large eddy simulation (LES) of a CH4/H2 diffusion jet flame in hot coflow (JHC) is undertaken to find distinct behaviors of moderate or intense low oxygen dilution (MILD) combustion (MILDC), high temperature combustion (HTC), and traditional combustion (TC). These three JHC flames are realized by using different coflow temperatures and oxygen mass fractions: (1) 1300 K and 9% for MILDC; (2) 1300 K and 30% for HTC; (3) 600 K and 30% for TC. The modeling of LES combining the eddy dissipation concept (EDC) with a global four-step reaction mechanism is validated by the JHC measurements of Dally et al. (Proc. Combust. Inst. 2002, 29, 1147–1154). The instantaneous and time-averaged velocities, temperatures, and species concentrations such as carbon monoxide (CO) are presented and compared for the three cases. It is demonstrated that the JHC flames of MILDC and HTC both develop from autoignition nearly immediately downstream of the nozzle exit, while the lift-off JHC flame of TC evolves from an induced-ignition ...

Combustion Characteristics of a Methane Jet Flame in Hot Oxidant Coflow Diluted by H2O versus the Case by N2

This numerical study investigates combustion characteristics of a CH4 jet flame in hot oxidant coflow (JHC) diluted by H2O versus the case by N2. Both MILD oxy–fuel combustion and traditional combustion are generated by varying the coflow oxygen level from 3% to 85% in volume. To distinguish physical and chemical effects of changing the diluent from N2 to H2O, the JHC flame diluted by XH2O (artificial H2O without reactivity) is also simulated. It is demonstrated that both the temperature and the dimension of the JHC flame reduce substantially when changing the diluent from N2 to H2O. Such reductions derive primarily from different fluid properties of N2 and H2O, i.e., H2O has a lower fluid density and a higher thermal capacity. The change from N2 to H2O is also found to have an impact more chemically on formation of the main intermediate species (i.e., H, O, CO, OH) and simultaneously to increase the flame lift-off distance due to longer ignition delay. Hence, the MILD combustion is expected to occur more...

Large eddy simulation of turbulent lifted flame in a hot vitiated coflow using tabulated detailed chemistry

Large eddy simulation (LES) of Cabra flame is carried out by a sub-grid scale (SGS) model of tabulated detailed chemistry combined with presumed probability density function (PDF). The chemistry table is constructed by utilizing the coupling of auto-ignition model and premixed flame propagation model. The copula method is applied to build multivariate joint PDF of mixture fraction and reaction progress variable. LES transport equations for characteristic scalars, relevant sub-grid variances and covariance are deduced. The LES results show that the developed SGS model can give an accurate prediction for lift-off height and flame structures. The stabilization mechanism and combustion regime of Cabra flame are discussed.

Comprehensive numerical study of the Adelaide Jet in Hot-Coflow burner by means of RANS and detailed chemistry

The present paper shows an in-depth numerical characterisation of the Jet in Hot Co-flow (JHC) configuration using the Reynolds Averaged Navier-Stokes (RANS) modelling with detailed chemistry. The JHC burner emulates the MILD combustion by means of a hot and diluted co-flow and high speed injection. The current investigation focuses on the effect of turbulent combustion models, turbulence model parameters, boundary conditions, multi-component molecular diffusion and kinetic mechanisms on the results. Results show that the approaches used to model the reaction fine structures, namely as Perfectly Stirred Reactors (PSR) or Plug Flow Reactors (PFR), do not have a major impact on the results. Similarly, increasing the complexity of the kinetic mechanism does not lead to major improvements on the numerical predictions. On the other hand, the inclusion of multi-component molecular diffusion helps increasing the prediction accuracy. Three different Eddy Dissipation Concept (EDC) model formulations are compared, showing their interaction with the choice of the C1ε constant in the k−ε turbulence model. Finally, two approaches are benchmarked for turbulence-chemistry interactions, the EDC model and the Partially Stirred Reactor (PaSR) model.

Measurement of mixing-induced thermal non-equilibrium in a supersonic shear layer using spontaneous Raman scattering

Mixing-induced vibrational non-equilibrium of N2 is studied in the turbulent shear layer between a supersonic cold jet and a surrounding hot air coflow. The jet fluid is either air, N2, or Ar, and the heated coflow is air at a maximum temperature of 850 K. The rotational and vibrational temperatures of N2 are determined by fitting the measured time-averaged spontaneous Raman spectra to an analytical model that allows for different equilibrium distributions for the vibrational and rotational states. The mixing of the jet fluid with the coflow gases occurs over time scales of the order of 5 μs, which is found to be sufficiently fast to induce vibrational non-equilibrium in the mixture of hot and cold gases. Results show that the non-equilibrium can be measured, but not on the cold side of the shear layer where the vibrational population in the first hot band is negligible. The effect of fluctuating temperatures within the time-averaged Raman measurement was quantified with the use of Rayleigh thermometry and found to not significantly alter the Raman scattering results. It was also found that the non-equilibrium increases in the shear layer when N2 is removed from the jet fluid, indicating that the observed non-equilibrium is an averaged result of two competing processes that occur simultaneously at a molecular scale, i.e., vibrationally hot N2 is being cooled by a fast jet fluid and vibrationally cold jet fluid is being heated by a hot coflow fluid. An interesting inference of this view is that the averaging effect is always present, regardless of the measurement resolution.

Characteristics of turbulent n-heptane jet flames in a hot and diluted coflow

Distinctive behaviour of turbulent n-heptane jet flames is revealed by conventional photography and laser-induced fluorescence of the hydroxyl radical (OH-LIF) performed in a Jet in Hot Coflow (JHC) burner. Prevaporised n-heptane, carried by air, issues into a hot vitiated coflow at two temperatures (1250 K and 1315 K) with three coflow oxygen levels (XO2 = 3%, 6%%, and 9%), mimicking moderate or intense low oxygen dilution (MILD) combustion conditions. Results for n-heptane flames are compared to those for ethylene, ethanol, and natural gas flames. Two main discrepancies are revealed: firstly, the apparent liftoff height of n-heptane flames decreases gradually with the increasing coflow oxygen level, while the apparent liftoff height of other fuels exhibit a non-monotonic trend; secondly, a transitional flame structure of n-heptane occurs in a coflow with XO2 = 3%, while this structure occurs in a coflow with XO2≥ 9% for other fuels. Calculations using a closed homogeneous reactor model support the interpretation of the experimental data, namely that the temporal profiles of OH in the n-heptane flames are similar for the 3% and 9% O2 cases, differing from the behaviour of other fuels. A comparative analysis of n-heptane and ethanol chemistry, focused on the fuel pyrolysis and net negative heat release regions, suggests that it is more difficult for n-heptane than ethanol to meet one of the criteria of MILD combustion.

Multi-species PLIF study of the structures of turbulent premixed methane/air jet flames in the flamelet and thin-reaction zones regimes

Simultaneously planar laser-induced fluorescence (PLIF) measurements of OH, CH, CH2O and toluene are carried out to investigate the structures of turbulent premixed methane/air jet flames in the flamelet regime and the thin-reaction zones regime. A premixed flame jet burner of an inner diameter of 1.5mm is employed. Stoichiometric methane/air mixtures introduced as a jet are ignited and stabilized in a hot co-flow generated by a coaxial porous plug pilot flame surrounding the jet. The Reynolds number for the studied jet ranges from 960 to 11,500 with the characteristic Karlovitz number ranging from 1 to 60. The focus of this study is on the characterization of the structures and turbulent burning velocity of premixed flames in the flamelet and the thin-reaction zones regimes. The preheat zone is analyzed using the CH2O and toluene PLIF fields, whereas the reaction zone is analyzed using the CH and OH PLIF fields. Laser Doppler Anemometer (LDA) measurements are performed to characterize the turbulence field and it is noted that when the Reynolds/Karlovitz number increases a successive thickening of the preheat zone is observed, whereas the reaction zone, characterized by the CH layer maintains nearly the same thickness. The heat release zone, characterized by the combination of the OH and CH2O PLIF fields, is shown to nearly maintain the same thickness under the present experimental conditions. The flame surface wrinkle ratio is shown to be Reynolds number and Karlovitz number independent when the Reynolds number is high enough such that the smallest wrinkle scales reach to the length scales of the thin reaction layers. The global fuel consumption speed of the jet flame is analyzed using the toluene PLIF field and the OH PLIF field. A discrepancy in the two consumption velocities is found as the Karlovitz number increases. This is found to be a result of the broadening of the oxidation zone. These findings provide experimental support to the flamelet and thin-reaction zone regime hypotheses of turbulent premixed combustion.

An interband cascade laser-based in situ absorption sensor for nitric oxide in combustion exhaust gases

A direct absorption nitric oxide sensor for combustion exhaust gas measurements, based on an interband cascade laser operating at 5.2 µm, is presented. The sensor was applied to the hot air co-flow of an auto-ignition test rig (800–1300 K), which contains nitric oxide mole fractions of the order of 1 mol%, due to prior microwave plasma heating. The effect of non-uniform temperature along the beam path, on both absorption line strength and gas density, was included in mole fraction measurements at various co-flow temperatures and velocities. At an absorption length of only 82 mm, a noise-limited detection limit of 30 ppm with a 10 ms observation time was achieved at 800 K. The results were compared in detail to previously measured mole fractions, using a sampling gas analyzer.

Characteristic Chemical Time Scale Analysis of a Partial Oxidation Flame in Hot Syngas Coflow

Characteristic chemical time scale analysis plays a key role in the understanding of turbulence–chemistry interaction in turbulent combustion research and is also an important basis for the selection or development of combustion models in turbulent combustion modeling. A new method named main direction identification (MDID) was developed based on the modification of the CTS-ID (chemical time scale identification) method to achieve the function of identifying the characteristic time scale. Direction weight factor combined with mole fraction limit were used as a criterion in MDID to determine the characteristic time scale. MDID was applied to study characteristic chemical time scales of a CH4–O2 inverse diffusion flame in hot syngas coflow, which is a model flame developed before to study the combustion process in partial oxidation reformers. Results show that the chemical time scale given by the MDID method is about 10–5 s in the combustion area and 10–2 s in the reforming area. The main reaction pathway w...

Effects of oxidant stream composition on non-premixed laminar flames with heated and diluted coflows

The moderate or intense low oxygen dilution (MILD) combustion regime offers reductions in pollutant emissions and improvements in efficiency. The implementation of MILD combustion in non-premixed systems, however, still requires a significantly improved understanding of the effects of the oxidant stream composition in both the MILD combustion regime, and in the transition from MILD combustion to conventional spontaneous-ignition. This paper presents a numerical study of two-dimensional flames near the transition between MILD combustion and conventional spontaneous-ignition. The numerical study is performed with a detailed kinetic mechanism, using the laminarSMOKE code to provide insight into the chemical structure of flames with a hot and diluted coflow. This study is complemented by new experimental observations of laminar flames in similar coflows, which demonstrate a non-monotonic change in lift-off height with changing oxidant O2 concentration. This change occurs in conjunction with a transition from spatially gradual ignition to well-defined flame bases, with increasing coflow O2 level. The simulated flames are compared to a previous definition of MILD combustion in a hot coflow as an edge flame without a tribrachial flame structure. The different structures of simulated CH4 flames are consistent with the observed experimental behaviour under similar conditions, however comparisons between experimental observations and simulations of C2H4 flames highlight the importance of the flow-field, even in a simple streaming flow. The simulations show that equilibrium levels of the OH radical (< 10 ppm) in the oxidant stream significantly intensifies a MILD CH4 reaction zone, by increasing CH3 oxidation, however such levels have little effect on tribrachial, spontaneously-igniting flames. Conversely, increasing the ratio of CO2 to H2O in the coflow reduces the intensity of a MILD CH4 reaction zone. Neither the inclusion of equilibrium concentrations of OH, nor the change in CO2 to H2O ratio, in the oxidant results in a transition between MILD reaction zones to tribrachial spontaneously-igniting flames, despite significantly affecting the temperature of reaction zones in the MILD combustion regime. The results show that the intensity of MILD combustion is strongly dependent on the different chemical species in the oxidant stream. Tribrachial spontaneously-igniting flames are, in contrast, relatively resilient against changes other than temperature and O2 level.

Temporal evolution of auto-ignition of ethylene and methane jets propagating into a turbulent hot air co-flow vitiated with NOx

In this paper, the temporal evolution of auto-ignition (AI) of C2H4 and CH4 jets propagating into a NOx vitiated hot co-flow at high velocity and turbulence was studied. Simultaneous temporally-resolved planar laser-induced fluorescence (PLIF) experiments of OH and CH2O were carried out in a recently developed test rig for auto-ignition studies of turbulent non-premixed flows. Flame stabilization mechanisms were analyzed for both fuels at several operating conditions. The reaction progress of AI kernels as well as their apparent growth rate were evaluated. Results revealed that the stabilization mechanism (i.e. lifted flame or isolated kernel) strongly depends on the turbulent mixing, co-flow temperature and fuel composition. A further statistical analysis of the heat release rate (HRR) zones, calculated as the product of the CH2O- and OH-PLIF signals, delivered an indication on the different AI characteristics of C2H4 and CH4. An evaluation of the temporal evolution of the HRR for C2H4 jets provided a deeper insight about the reaction progress at different conditions of the co-flow, when AI was initiated by isolated kernels. Finally, estimations of the apparent growth rate of AI kernels indicated a faster propagation of C2H4 kernels when compared with those of CH4.

Methane combustion in MILD oxyfuel regime: Influences of dilution atmosphere in co-flow configuration

MILD (moderate or intense low oxygen dilution) oxyfuel combustion is a recently proposed clean combustion mode which can remedy the shortcomings of the standard oxyfuel combustion technology. Nowadays most available studies on MILD oxyfuel combustion focus on how to realize this new combustion regime in O2/CO2 atmosphere. The open research on methane MILD oxyfuel combustion in O2/H2O atmosphere is quite sparse. In the present work, we carry out a comprehensive comparison study on methane MILD oxyfuel combustion in different dilution atmosphere for the first time. The JHC (jet in hot co-flow) burner is adopted as a research prototype. The investigation is based on numerical simulation, so firstly the adopted numerical approach is validated by some experimental data in open literature. The numerical comparison is conducted by varying the mass fraction of oxygen in the co-flow and the temperature of the hot co-flow, two key parameters affecting fine reaction structures in JHC. Through the present investigation, a number of findings are reported for the first time and some conclusions presented in previous publications are checked with analyses, especially on some conflicted claims between the previous publications. In addition, several new questions are raised, which may inspire further research activities in future.

Stabilisation of turbulent auto-igniting dimethyl ether jet flames issuing into a hot vitiated coflow

Experimental results focusing on understanding the role of autoignition in the stabilisation of turbulent lifted jet flames of dimethyl ether issuing into a hot vitiated coflow are explored in this paper. A stability map of the flame lift-off height as a function of coflow temperature and jet partially premixing ratios is reported, indicating a strong sensitivity of the mean and rms lift-off heights to the coflow temperature. A new experimental setup that utilises high repetition rate (10kHz) joint OH and CH2O PLIF combined with a third camera to track out of plane motion via chemiluminescence emission is employed. This allows the temporal dynamics of the flame base and autoignition kernels to be tracked at high spatial resolution over a spatial window of 20mm (height) × 43mm (width). To capture the temporal statistics of the flame over the full lift-off height range, a complementary, but not simultaneous, large field of view high-speed chemiluminescence imaging experiment was employed. The use of high-speed diagnostics is shown to be invaluable to understand the temporal dynamics of the flame base region, such as identifying and tracking autoignition kernels ahead of, and merging with, the main flame base. For low coflow temperatures, autoignition kernels are found to be fundamentally important to the stabilisation of the flame: small autoignition initiated kernels rapidly grow to form large flame kernels that eventually merge with the main flame base and in doing so rapidly establish a new lift-off height significantly upstream of the original flame base. For high coflow temperatures, autoignition kernels whilst still present, are significantly smaller in mean size and play a reduced role in the overall flame stabilisation. This leads to the conclusion that partially premixed flame propagation plays an increased role in overall flame stabilisation for high coflow temperature flames.

Autoignition and Stabilization of Diesel–Propane Lifted Flames Issuing into a Hot Vitiated Co-flow

Turbulent lifted flames of diesel–propane blend fuels issuing into a vitiated co-flow are recorded with a high-speed camera. The spray characteristics, flame structures, ignition delays, and lift-off heights of jet flames of diesel–propane blend fuels are analyzed in this research. As an additive to diesel fuel, propane has little influence on the autoignition process of diesel from the perspective of chemical kinetics but it improves the atomization, evaporation, and turbulence of the fuel spray. The addition of propane is beneficial for studying the interaction of the chemical kinetics and fluid dynamics in turbulent lifted flames. The experimental results show that the propane fraction has different influences on the ignition delay in two temperature ranges. The ignition delay increases with the increase of the fraction of propane when the co-flow temperature is lower than 1080 K and decreases when the co-flow temperature is higher than 1117 K. Thus, a related mechanism controlling the ignition delay o...

Laminar Flame Calculations for Analyzing Trends in Autoignitive Jet Flames in a Hot and Vitiated Coflow

Experiments of autoignitive jet flames in a hot and vitiated coflow have previously shown various flame behaviors, spanning lifted flames to moderate or intense low oxygen dilution (MILD) combustion. For better understanding the behavior of flames in this configuration, regime diagrams and ignition delay results are presented from well-stirred reactor calculations across a wide range of operating conditions for methane and ethylene fuels. In conjunction with two-dimensional calculations, the importance of flame precursors and oxygen penetration across the reaction zone is revealed. It is found that widely accepted definitions and regime diagrams are inadequate to classify and reconcile the different flame behaviors that are observed experimentally. For accurate prediction of the ignition process, it is necessary to incorporate boundary conditions that capture minor species in the oxidizer. The role of fuel type also has a major impact on the ignition process and flame appearance.

Experimental and numerical study on ethanol and dimethyl ether lifted flames in a hot vitiated co-flow

Lifted flames of gaseous ethanol and dimethyl ether (DME) issuing into a hot co-flow are recorded with a high-speed camera. The flames of the two fuels at 433K are studied in a co-flow temperature range from 966K to 1149K and at an ambient pressure of 1atm. The experimental results show that the co-flow temperature has a substantial influence on the jet flame characteristics. It is concluded that the lift-off height is controlled by the ignition delay, whereas the fluctuation is mainly controlled by the sensitivity of the ignition delay to the co-flow temperature. A numerical study coupling the Reynolds averaged Navier-Stokes (RANS) equation with the eddy-dissipation-concept model (EDC) employing detailed reaction mechanisms is established and is consistent with the previous flame structure and lift-off height obtained by the experimental results. The radial profiles and axial distribution of important parameters reveal the acceleration of the mixing process caused by exothermic reactions and indicate the differences in the mixture fraction at the stabilization points between simple fuel (H2 and CH4) and the present fuels. The influences of the chemical reaction on the stabilization mechanism at low and high co-flow temperatures are also discussed.

Large eddy simulation of methane/air lifted flame with hot co-flow

ABSTRACTThe lifted flame with hot co-flow can represent typical combustion features in the practical systems with recirculation of the combustion product. In this study, the methane/air lifted flame was simulated by large eddy simulation. Regarding the special stabilization mechanism in the lifted flame, two different one-step methane oxidation mechanisms were used: (i) the conventional mechanism widely used in engineering simulations and (ii) a modified mechanism considering the effects of the equivalence ratio. By comparing the simulation results with the experimental data, both mechanisms could predict the liftoff phenomenon; however, the simulation using the modified mechanism provided more reasonable results.