Flame Lift-off Research Articles

Controlling pollutant emissions in a high-pressure combustor with fuel-diluent blending

This paper investigates the formation of nitric oxide (NO) and carbon monoxide (CO) emissions as a function of nitrogen (N2) or carbon dioxide (CO2) diluent content, added to a premixed reacting jet in crossflow. Reaction characteristics of a rich methane-air-diluent jet injected into a lean vitiated crossflow were analyzed at an elevated pressure level 5 atm, which is critical to obtain data scalable to industry conditions. The jet was pre-heated and enriched with 0%, 15%, 30%, and 50% dilution by mass percentage to quantify the effect on pollutant emissions. Simulated results of the full chemistry Star-CCM + CFD model were verified with data taken in an experimental high-pressure combustion facility, which provides pressure, temperature, and velocity profiles, as well as line-of-sight chemiluminescence and exit emission measurements. The analysis revealed the significant influence of the diluent to increase the axial flame lift-off and delay axial combustion. Increasing the diluent content increased the timescale for the flame to stabilize, which allowed for greater entrainment of crossflow oxidizer into the axial jet stream, and led to decreased pollutant emissions. Hence, crossflow entrainment is a critical driving force at high diluent content. Local nitric oxide (NO) emission formation in the axial stage was predicted numerically, showing a correlation between diluent addition, axial heat release, and the formation of nitric oxide pollution. The high diluent levels delayed axial combustion, while reducing prompt and thermal NOx by mitigating flame hotspots and minimizing the timescale that the products remain at high temperatures.

Insight into flickering/shedding in buoyant droplet-diffusion flame during interaction with vortex

The interaction of incident vortex with buoyant diffusion flame of a pendant fuel droplet (n-dodecane) is investigated in the present study. The flame stretching/shedding events occur depending on vortex strengths, which are characterized using vortex Reynolds number (Rev). Different flame response regimes have been identified for different vortex strengths. The velocity scale associated with the vortices is higher when compared to natural convection velocity leading to forward extinction, followed by flame lift-off. The flame lift-off increases with Rev, and partial extinction occurs at very high Rev. In all the cases, the flame oscillates/sheds during the vortex interaction, which causes a momentary increase in the flame heat release rate. As the vortex moves past the flame, the oscillations dampen gradually, and the flame regains the natural buoyant behavior. The high-velocity scales imposed during vortex interaction make the flow momentum-dominant, resulting in a low Richardson number (Ri). Since, the flickering frequency scaling valid for Ri>>1 is not obeyed, a new scaling has been proposed to incorporate the vortex momentum effect by using vortex circulation instead of gravity. Only a portion of the incident vortex circulation contributes to the shedding as some of it is lost to forward extinction and lift-off events. Thus, the reduced vortex velocity scale is used to estimate the critical circulation. The flame shedding height calculated using new scaling in the critical circulation equation agrees with experimental observations.

Large Eddy Simulations of a Low-Swirl Gaseous Partially Premixed Lifted Flame in Presence of Wall Heat Losses

The use of lifted flames presents some very promising advantages in terms of pollutant emissions and flame stability. The focus here is on a specific low-swirl injection system operated with methane and derived from an air-blast atomizer for aero-engine applications, which is responsible for flame lift-off. The key feature of this concept is the interaction between the swirling jet and the confinement walls, leading to a strong outer recirculation zone and thus to an upstream transport of combustion products from the main reaction region to the flame base. Here, the representation of the physics involved is challenging, since finite-rate effects govern the lift-off occurrence, and only a few numerical studies have been carried out on this test case so far. The aim of the present work is therefore to understand the limits of some state-of-the-art combustion models within the context of LES. Considering this context, two different strategies are adopted: the Flamelet-Generated Manifold (FGM) approach and the Thickened Flame (TF) model. A modified version of the FGM model including stretch and heat loss effects is also applied as an improvement of the standard model. Numerical results are compared with the available experimental data in terms of temperature and chemical species concentration maps, showing that the TF model can better reproduce the lift-off than the FGM approach.

Spray Combustion Characteristics of Single/Multicomponent Surrogate Fuel for Aviation Kerosene

Abstract In this research, n-dodecane and world-average jet fuel (JW) are selected as single and multicomponent surrogate fuel of aviation kerosene to study the Jet-A spray combustion characteristics. The spray combustion phenomena are simulated using large eddy simulation coupled with detailed chemical reaction mechanism. Proper orthogonal decomposition method is applied to analyze the flow field characteristics, and the instantaneous velocity field is decomposed into four parts, namely, the mean part, coherent part, transition part, and turbulent part, respectively. The four parts have their own characteristics. In terms of different fuels, JW has a higher intensity of coherent structures and local vortices than n-dodecane, which promotes the fuel–air mixing and improves the combustion characteristics, and the soot formation is significantly reduced. In addition, with the increase of initial temperature, the combustion is more intense, the ignition delay time is advanced, the flame liftoff length is reduced, and soot formation is increased accordingly.

On the Flow Structure and Dynamics of Methane and Syngas Lean Flames in a Model Gas-Turbine Combustor

The present paper compares the flow structure and flame dynamics during combustion of methane and syngas in a model gas-turbine swirl burner. The burner is based on a design by Turbomeca. The fuel is supplied through injection holes between the swirler blades to provide well-premixed combustion, or fed as a central jet from the swirler’s centerbody to increase flame stability via a pilot flame. The measurements of flow structure and flame front are performed by using the stereo particle image velocimetry and OH planar laser-induced fluorescence methods. The measurements are performed for the atmospheric pressure without preheating and for 2 atm with the air preheated up to 500 K. The flow Reynolds numbers for the non-reacting flows at these two conditions are 1.5 × 103 and 1.0 × 103, respectively. The flame dynamics are analyzed based on a high-speed OH* chemiluminescence imaging. It is found that the flame dynamics at elevated conditions are related with frequent events of flame lift-off and global extinction, followed by re-ignition. The analysis of flow structure via the proper orthogonal decomposition reveals the presence of two different types of coherent flow fluctuations, namely, longitudinal and transverse instability modes. The same procedure is applied to the chemiluminescence images for visualization of bulk movement of the flame front and similar spatial structures are observed. Thus, the longitudinal and transverse instability modes are found in all cases, but for the syngas at the elevated pressure and temperature the longitudinal mode is related to strong thermoacoustic fluctuations. Therefore, the present study demonstrates that a lean syngas flame can become unstable at elevated pressure and temperature conditions due to a greater flame propagation speed, which results in periodic events of flame flash-back, extinction and re-ignition. The reported data is also useful for the validation of numerical simulation codes for syngas flames.

Numerical Steady and Transient Evaluation of a Confined Swirl Stabilized Burner

Lean premixed combustion technology became state of the art in recent heavy-duty gas turbines and aeroengines. In combustion chambers operating under fuel-lean conditions, unsteady heat release can augment pressure amplitudes, resulting in component engine damages. In order to achieve deeper knowledge concerning combustion instabilities, it is necessary to analyze in detail combustion processes. The current study supports this by conducting a numerical investigation of combustion in a premixed swirl-stabilized methane burner with operating conditions taken from experimental data that were recently published. It is a follow-up of a previous paper from Farisco et al., 2019 where a different combustion configuration was studied. The commercial code ANSYS Fluent has been used with the aim to perform steady and transient calculations via Large Eddy Simulation (LES) of the current confined methane combustor. A validation of the numerical data has been performed against the available experiments. In this study, the numerical temperature profiles have been compared with the measurements. The heat release parameter has been experimentally and numerically estimated in order to point out the position of the main reaction zone. Several turbulence and combustion models have been investigated with the aim to come into accord with the experiments. The outcome showed that the combustion model Flamelet Generated Manifold (FGM) with the k-ω turbulence model was able to correctly simulate flame lift-off.

Spray Combustion of Rocket-Grade-Kerosene-Based Nanofluids with Oxygen

The spray and spray combustion characteristics of rocket-grade kerosene (Isrosene) loaded with graphene nanoplatelets and aluminum nanoparticles are investigated in this study. A shear coaxial injector with recess was employed to generate the nanofluid spray flowing through the center orifice and 99.98% oxygen through the annular gap. The mass flow rate was maintained at for both the fuel and oxidizer streams, and the flame was nonconfined. Effect of nanoparticles on droplet size of the sprays, flame liftoff height, flame temperature, and CO emissions were measured. Droplet size measurement was conducted using a laser-based nonintrusive technique (MALVERN particle sizer). It was observed that for the same mass flow rates of the fuel and oxygen, the droplet size of the spray increased by 50% due to the addition of nanoparticles. The direct videography method was used to visualize the flame shape and size, and it was observed that the flame liftoff height decreased by 50 and 10% due to the addition of graphene and aluminum nanoparticles, respectively. The presence of nanoparticles enhanced the reaction rate, which was concluded based on the measurement of flame temperature and CO concentrations.

Impact of Hydrogen Addition on the Thermoacoustic Instability and Precessing Vortex Core Dynamics in a CH4/H2/Air Technically Premixed Combustor

Abstract We study the impact of H2 enrichment on the unsteady flow dynamics and thermoacoustic instability in the prediction and control of instabilities in industrial turbines (PRECCINSTA) swirl combustor. The experiments were performed at atmospheric conditions with H2/CH4 fuel mixtures at a global equivalence ratio of 0.65 and a constant thermal power of 20 kW. We analyze data with three fuel compositions: 0%, 20%, and 50% H2 in two operating modes, premixed (PM) and technically premixed (TPM). A new multiresolution modal decomposition method, using a combination of wavelet transforms and proper orthogonal decomposition (WPOD) is performed on time resolved flow velocity and OH planar laser induced fluorescence (OH planar laser induced fluorescence (OH-PLIF)) measurements. Thermoacoustic oscillations are observed in the TPM operating mode alone, indicating that the primary heat release driving mechanism is due to fuel-air ratio oscillations. WPOD results for the 0% H2 TPM case reveal intermittent helical precessing vortex core (PVC) oscillations along with axisymmetric hydrodynamic flow oscillations due to the thermoacoustic oscillations. These oscillations cause local flame extinction near the nozzle centerbody resulting in liftoff. A PVC then develops in the flow and enables intermittent flame reattachment. In the 0% H2 premixed case, the flame remains lifted off the centerbody despite the presence of PVC oscillations. H2 enrichment results in the suppression of flame liftoff and the PVC in both operating modes. We show from flow strain rate statistics and extinction strain rate calculations that the increase of the latter with H2 addition, allows the flame to stabilize in the region near the centerbody where the pure CH4 cases show lift off.

Preheating and premixing effects on NOx emissions in a high-pressure axially staged combustor

NOx emissions remain a primary concern for modern and future gas turbines, particularly as we progress towards a net-zero carbon future. It is critical to identify novel strategies to mitigate environmental pollutants for both modern turbines operated on natural gas and future turbines projected to use carbon-free fuels (such as hydrogen). In the current study, NOx emissions of a reacting methane-air jet in a vitiated crossflow are experimentally investigated in a model axially staged combustor at 5 atm. The combustion products from the main stage combustor flow into the axial stage at temperatures ranging from 1580 to 1650 °C. The axial stage contains a reacting jet in crossflow, which provides an overall temperature rise ranging from 50 to 220 °C to generate exit temperatures similar to gas turbine engines. The focus of this paper is to explore the effects of flame liftoff and ignition timescales on the NOx contribution of this secondary jet. This is done by varying the premixed level and preheat temperature of the axial jet. The effects are explored at different momentum flux ratios (J), jet equivalence ratios (φjet), exit temperatures (Texit), and main stage (headend) temperatures (THE). High-speed CH* chemiluminescence imaging is employed for each test case to determine the flame stabilization point. For a given temperature rise and combustor exit temperature, an increase in ignition delay was observed with a decrease in jet temperature. This provided a NOx benefit, particularly for jets that are injected at or close to an ignitable mixture fraction. Similarly, the coaxial injector caused a significant ignition delay compared to the fully premixed injector; this also led to a NOx benefit. Overall, the results show that a thermally driven delay in the axial stage provides a greater NOx reduction than a mixing delay.

Modeling of liftoff heights of non-premixed turbulent flames in co-flows having various temperatures and O2 concentrations

Lifted flames in combustion furnaces are diluted with burned gas entrained into the fuel jet. The reduced concentrations of reactants resulting from this dilution increase the liftoff height, while the associated temperature increase decreases the height. The aim of the present study was to develop a premixed model capable of predicting the variation in the liftoff height resulting from entrainment. A triple concentric burner incorporating fuel gas, oxidizer and co-flow gas nozzles was employed to simulate a combustion furnace. Prior to combustion tests, the fraction of the fuel gas in the non-reactive jets forming on the burner was determined, to allow an evaluation of parameters affecting the entrainment rate of the co-flow gas. The flame liftoff height above the burner was found to increase with decreases in the O2 concentration in the co-flow gas and was decreased with increases in temperature. Three premixed models were examined: a conventional premixed model, a DP1 model including only the effect of decreasing reactant concentrations and a DP2 model including the effects of both decreasing concentrations and temperature increases. Validations of these models demonstrated that the conventional model failed to predict variations in the liftoff height at a variety of co-flow gas O2 concentrations and temperatures. The DP1 model also provided insufficient correlations between the bulk velocity and liftoff height, such that the correlation line at a high co-flow gas temperature separated from that at room temperature. In contrast, the DP2 model provided excellent correlations in conjunction with different virtual origin positions.

A combined oscillation cycle involving self-excited thermo-acoustic and hydrodynamic instability mechanisms

The paper examines the combined effects of several interacting thermo-acoustic and hydrodynamic instability mechanisms that are known to influence self-excited combustion instabilities often encountered in the late design stages of modern low-emission gas turbine combustors. A compressible large eddy simulation approach is presented, comprising the flame burning regime independent, modeled probability density function evolution equation/stochastic fields solution method. The approach is subsequently applied to the PRECCINSTA (PREDiction and Control of Combustion INSTAbilities) model combustor and successfully captures a fully self-excited limit-cycle oscillation without external forcing. The predicted frequency and amplitude of the dominant thermo-acoustic mode and its first harmonic are shown to be in excellent agreement with available experimental data. Analysis of the phase-resolved and phase-averaged fields leads to a detailed description of the superimposed mass flow rate and equivalence ratio fluctuations underlying the governing feedback loop. The prevailing thermo-acoustic cycle features regular flame liftoff and flashback events in combination with a flame angle oscillation, as well as multiple hydrodynamic phenomena, i.e., toroidal vortex shedding and a precessing vortex core. The periodic excitation and suppression of these hydrodynamic phenomena is confirmed via spectral proper orthogonal decomposition and found to be controlled by an oscillation of the instantaneous swirl number. Their local impact on the heat release rate, which is predominantly modulated by flame-vortex roll-up and enhanced mixing of fuel and oxidizer, is further described and investigated. Finally, the temporal relationship between the flame “surface area,” flame-averaged mixture fraction, and global heat release rate is shown to be directly correlated.

Understanding the influence of burner structure on the stability and chemiluminescence of inverse diffusion flame

Inverse diffusion lift-off flame was widely used in industrial fields such as non-catalytic partial oxidation of methane. In order to investigate the stability and chemiluminescence characteristics of the inverse diffusion lift-off flame, the OH∗ and CH∗ radiation characteristics, lift-off height, and transition (attachment, lift-off and blow-out) of flames under different burner structure were discussed. The results showed that burner rim thickness and incidence angles would affect the stability of the inverse diffusion flame. When the thickness of the burner rim exceeded 0.5 mm, the flame would directly change from the attachment state to blow-out state as oxygen velocity increased. Additionally, the values of the blowout limit of the nozzle were inversely proportional to the rim thickness of the burner. Different incident angles would result in various shear angles, which would affect the flame structure. As the incidence angle decreased, the tangential velocity of flame increased and the flame tended to be more stable. When the lift-off flame generated, OH∗ intensity and distribution showed a sudden change, and the OH∗/CH∗ peak intensity ratio of the flame appeared abrupt changes.

Experimental assessment of the impact of variation in jet fuel properties on spray flame liftoff height

Ensuring efficient and clean combustion performance of liquid-fueled engines requires comprehension of the influence of fuel composition and properties on flame behavior, such as flame liftoff height (LOH) and lean blowout limit (LBO). Spray flame stability is strongly affected by both the fuel reactivity and physical properties. Herein, the flame stability mechanism represented by LOH is investigated for seven jet turbine fuels, including surrogate, alternative, and conventional jet fuels, using a laboratory spray burner. Based on the experimental observations, the current work introduces a new analysis, which provides insight into the competing/complementing processes that occur in a multi-phase reacting system and highlights the key properties important in spray flame dynamics, accounting for both the fuel spray/vaporization as well as the chemical reactivity, to explain the relative differences in LOH of complex multicomponent fuels. Results show that spray flame stabilization occurs when there is a balance between the local spray burning velocity and the incoming jet velocity, which is strongly associated with laminar flame speed and the relative amount of liquid and gaseous fuel crossing the flame preheat region. Using a multicomponent droplet evaporation model, it was observed that preferential vaporization of the lighter and more reactive species of the simple surrogate fuels contribute to a shorter lifted flame as compared to fuels consisting of heavier and/or less reactive components. The LOH of real jet fuels showed strong sensitivity to droplet vaporization and mixing, which is controlled by the fuels’ volatility (i.e., boiling temperature at 50% distillation volume) and atomized droplet size, and to a lesser extent the reactivity represented by laminar flame speed. The enhanced linearity in the correlation of 50% vapor distilled compared to the other distillation cuts suggests that preferential vaporization could play an important role in defining the LOH stabilization mechanism even for more complex fuels.

Computational characterization of hydrogen direct injection and nonpremixed combustion in a compression-ignition engine

With the revived interest in hydrogen (H2) as a direct combustion fuel for engine applications, a computational study is conducted to assess the characteristics of H2 direct-injection (DI) compression-ignition (CI) non-premixed combustion concept. Development of a CFD modeling using CONVERGE CFD solver focuses on hydrogen's unique characteristics by utilizing a suitable numerical method to reproduce the direct H2 injection phenomena. A grid sensitivity study is performed to ensure the fidelity of results with optimal cost, and the models are validated against constant-volume optical chamber and diesel engine experimental data. The present study aims to contribute to the future development of DICI H2 combustion engines, providing detailed characterization of the combustion cycle, and highlighting several distinct aspects of CI nonpremixed H2 versus diesel combustion. First, unlike the common description of diesel sprays, hydrogen jets do not exhibit significant flame lift-off and air entrainment near injector nozzle, and the fuel-air interface is drastically more stratified with no sign of premixing. It is also found that the DICI H2 combustion concept is governed first by a free turbulent jet mixing phase, then by an in-cylinder global mixing phase. The former is drastically more dominant with the DICI H2 engine compared to conventional diesel engines. The free-jet mixing is also found to be more effective that the global mixing, which indicates the need to completely rethink the optimization strategies for CI engines when using H2 as fuel.

Ignition Process and Flame Lift-Off Characteristics of dimethyl ether (DME) Reacting Spray

Advanced combustion systems that utilize different combustion modes and alternative fuels have significantly improved combustion performance and emissions compared to conventional diesel or spark-ignited combustions. As an alternative fuel, dimethyl ether (DME) has been receiving much attention as it runs effectively under low-temperature combustion (LTC) modes such as homogeneous charge compression ignition (HCCI) and reactivity control combustion ignition (RCCI). Under compression-ignition (CI), DME can be injected as liquid fuel into a hot chamber, resulting in a diesel-like spray/combustion characteristic. With its high fuel reactivity and unique chemical formula, DME ignites easily but produces almost smokeless combustion. In the current study, DME spray combustion under several different conditions of ambient temperature (Tamb= 750–1100 K), ambient density (ρamb= 14.8–30 kg/m3), oxygen concentration (O2= 15–21%), and injection pressure (Pinj= 75–150 MPa) were studied. The results from both experiments (constant-volume combustion vessel) and numerical simulations were used to develop empirical correlations for ignition and lift-off length. Compared to diesel, the established correlation of DME shows a similar Arrhenius-type expression. Sensitivity studies show that Tamband Pinjhave a stronger effect on DME's ignition and combustion than other parameters. Finally, this study provides a simplified conceptual mechanism of DME reacting spray under high reactivity ambient (high Tamb, high O2) and LTC conditions. Finally, this paper discusses engine operating strategies using a non-conventional fuel such as DME with different reactivity and chemical properties.

An Experimental Study on Kerosene Spray Combustion Under Conventional and Hot-Diluted Conditions

ABSTRACT The combustion of kerosene spray under hot-diluted conditions and conventional conditions was experimentally investigated. By examining flame photographs, chemiluminescence images, and in-field temperature measurements, the separate effect of different variables including oxygen concentration, temperature and velocity of the co-flowing air, fuel flow rate and injection pressure, and eventually the type of spray nozzle on multiple parameters such as flame stability, structure, luminosity, temperature field, and qualitative CH radical distribution, as well as HCO and NO2 with lower precision, in the reaction region, have been studied. It was observed that an increment in injection pressure and co-flow temperature enhances the spray flame stability, while dilution exacerbates it. Also, a solid cone spray pattern with a lower spray angle has better stability than hollow cone ones with a higher spray angle. Moreover, it was noted that liquid fuels, compared to gaseous fuels, require higher preheating temperatures, for the same dilution level, to engender a stable flame. For combustion of spray in conventional conditions, a double-flame structure was observed consisting of a bluish section at the leading edge emerging into a yellowish sooting trail. An increase in co-flow velocity, as well as injection pressure, strengthens the inner flame front, whereas raising the co-flow temperature or diluting the oxidant, deteriorates the inner flame front. In the case of highly preheated air (without dilution), the flame liftoff height is reduced to as close to the atomizer as a few millimeters, forming a single flame structure similar to gaseous flames. In this case, the peak temperature was considerably higher than the conventional combustion, yet the gain was much lower than the preheating level. Combined effects of preheating and dilution alter the spray flame structure in a way that the flame volume is reduced, the temperature field has become more homogeneous, the peak temperature is limited to less than 1500 K, and temperature fluctuations have significantly decreased, seemingly approaching MILD combustion regime conditions.

Optimization of the flame characteristics of H2–O2 coaxial injection applied to hydrogen-fueled argon cycle engines

The argon power cycle is one of the most promising technologies for high efficiency and low emission hydrogen-fueled internal combustion engines. The application of coaxial injection technology in the hydrogen-fueled argon engine can improve the mixing process and the combustion performance of the H2/O2 mixture. In this study, an innovative H2–O2 coaxial injection combustion system was designed to investigate the jet flame characteristics of oxygen coaxially wrapped by hydrogen in a controllable argon thermal atmosphere. The findings of this study could provide a new perspective for designing hydrogen-fueled argon engines in the future. The influences of co-flow temperature, jet injection pressure, and excess oxygen coefficient were all determined. Observations of the flame showed a bright blue flame with a reddish glow in the far-burner region. Experimental results show that the flame length, cross-sectional area, and area/perimeter ratio first decrease with increasing jet injection pressure and subsequently increase, reaching maximum values at 0.4–0.6 MPa. When increasing the co-flow temperature from 1023 K to 1223 K, the cross-sectional area of the flame increases significantly by 61.1% at an excess oxygen coefficient of 0.4. Furthermore, the liftoff flame height shrinks when the co-flow temperature and the excess oxygen coefficient increase, while it rises along with an increasing jet injection pressure.

Investigation on Stability and Chemiluminescence Characterization for Liftoff Inverse Diffusion Flames

ABSTRACT In this study, the stability and chemiluminescence characteristics of CH4/O2 liftoff inverse diffusion flame were investigated. The OH* and CH* radiation characteristics, liftoff height and transition (including attachment, liftoff and blow-out) of flames under different conditions were discussed. The OH* chemiluminescence is generated at the root of the flame and increases sharply along the propagation direction. With the increase of oxygen gas velocities, the flame became more instability leading to the radial size of OH* decreases and the reaction zone is generally close to the central axis. Compared to the OH* distribution, the CH* generation area is smaller and the radial dimension is less. At the flame root and OH* peak intensity existed position, OH* is generated by H + O + M = OH*+M reaction and CH+O2 = OH*+CO reaction. While, at the top of flame, CO2 facilitates the OH+CO2 = OH*+CO2 reaction. The flame changes from attachment to blow-out with the increase of velocity. Moreover, O2 velocity of liftoff decreases but the blow-out limit increases when methane velocity increases. There is an obvious linear relationship between the O2 liftoff velocity and the methane velocity. After the methane velocity reached a critical value, the liftoff flame could not be formed even when the oxygen raises and the flame blew out directly. As the methane velocity increases, both the flame height and the methane liftoff height rise. The jet flow entrains the surrounding O2 to form a recirculation zone between the flame root and the nozzle tip, which maintains the liftoff stability.

Active Subspace Variation and Modeling Uncertainty in a Supersonic Flame Simulation

The active subspace method is applied to Burrows–Kurkov supersonic wall-jet flame to quantify modeling uncertainties associated with chemical kinetics, turbulence combustion models, and boundary conditions, as well as to investigate the evolution of key physiochemical process throughout the supersonic flame. One-dimensional active subspaces of the uncertain parameter space are identified for both flame liftoff length and flame–wall distance. Through the analyses of active subspaces, it shows that chemical kinetics plays a dominant role for flame initialization and early development, whereas turbulent transport becomes more important as the flame develops. The univariate response surfaces constructed in the active subspaces are then used to perform uncertainty quantification. For the flame liftoff length, the uncertainty introduced by chemical kinetics is slightly larger than that introduced by turbulence combustion models and significantly larger than boundary induced uncertainty; whereas the uncertainty in the predicted flame–wall distance is dominated by turbulence combustion modeling, and the contribution from boundary and chemical kinetics is trivial. Further comparison between two turbulence models with structural compressibility correction shows that the active subspace is insensitive to turbulence models at the same level of closure. The corrected shear stress tensor and corrected standard yield the same key physiochemical process and similar uncertainties in the predictions.

Monitoring combustion instabilities of stratified swirl flames by feature extractions of time-averaged flame images using deep learning method

The present article investigates the application of deep learning methods to monitor combustion instabilities based on time-averaged flame images. Experiments on BASIS (Beihang Axial Swirler Independently-Stratified) burner are carried out in a wide range of operating conditions to cover various flame shapes. We design a CNN (Convolutional Neural Networks) named BIM (BASIS Image Monitor) to extract features from flame shapes and determine the thermoacoustic states of the flames ahead of time. Experiments of 112 operating conditions are conducted, and their images are fed into BIM for training and testing purposes. BIM achieves around 99% prediction accuracy after a short training process lasting less than 20 minutes. Features extracted by BIM and visualized by Class Activation Maps (CAM) method reflect the underlying statistical links between flame images and their stabilities. Several phenomena (lift-off flame, flame-wall interaction, outer shear layer (OSL) flame and fuel mixing process) and their contributions to stabilities can be captured and quantified by CAM. After that, we test BIM with eight unknown operating conditions, and BIM can still perform well, identifying their stabilities and unstable features accurately. This paper explores the monitoring accuracy and feature extraction abilities of CNN in combustion. The BIM model shows potential to be a precursor in monitoring and active control systems of combustion instabilities.