Sub-grid Scale Combustion Research Articles

Numerical Analysis of High Frequency Transverse Instabilities in a Can-Type Combustor

Abstract Dry low emissions (DLE) systems are well-known to be susceptible to thermoacoustic instabilities. In particular, transverse, spinning modes of high frequency may appear, and lead to severe damage in a matter of seconds. The thermoacoustic response of an engine is usually specific to the combustor geometry, operating conditions and difficult to reproduce at the lab-scale. In this work, details of high frequency dynamics (HFD) observed during the early development phase of a new DLE system are provided, where a multipeaked spectrum was noticed during testing. Beginning with an analysis of the measured pressure spectra from three different concepts, an analytical model of the clockwise and anticlockwise transverse waves was fitted to the experimental data using a nonlinear curve fitting approach to produce a simple yet useful understanding of the phenomena. A flamelet-based large eddy simulation (LES) of the entire combustion system was used to complement this analysis and confirm the mode shapes using dynamic mode decomposition (DMD). Both approaches independently identified a spinning second-order mode as the dominant one in the high frequency regime. The LES indicates the coupling of a distortion of swirl profile with a precessing vortex core as a possible cause for the onset of instability. With regard to modeling sensitivities, it is shown that subgrid scale combustion modeling has a strong impact on predicted amplitudes. Ultimately, a thickened-flame model with a modified efficiency function provided consistent results.

Large eddy simulation of partially premixed flames with inhomogeneous inlets based on the DTF combustion model

Study of partially premixed flames involves a practical importance. In the present work, large eddy simulations (LES) are performed using dynamic thickened flame (DTF) model to study the challenging turbulent partially premixed flames with inhomogeneous and near-homogeneous inlets. In order to apply the DTF model for partially premixed flames, laminar flame thickness and laminar flame speed parameters are introduced as fitting functions of mixture fraction values. Therefore, thickening factor and efficiency function used in the DTF model can be adjusted according to mixture fraction values at the flame locations. The detailed comparisons of calculated mixture fraction and temperature radial statistics with experimental results show the good agreement between simulation and experimental results. Investigations of important combustion characteristics like local extinctions and blow-off limits further validate current sub-grid scale combustion modeling approach. Additionally, quantitative assessment of simulation results is performed by Wasserstein metric. The Wasserstein metric calculations exhibit modeling error close to one in terms of standard deviation for all the considered flame cases, showing the good performance of current simulations.

The partially stirred reactor model for combustion closure in large eddy simulations: Physical principles, sub-models for the cell reacting fraction, and open challenges

For their ability to account for finite-rate chemistry, reactor-based models are well suited Turbulence–Chemistry Interactions (TCI) Sub-Grid Scale (SGS) closures for Large Eddy Simulations (LES). The SGS closure in the Partially Stirred Reactor (PaSR) model relies on the determination of the reacting fraction of each computational cell, whose definition is based on estimates of the characteristic mixing and chemical time scales. Direct Numerical Simulations (DNS) of turbulent combustion can supply key information on TCI for the development, validation, and comparison of combustion models. In particular, a priori testing allows the direct validation of model assumptions. In the present work, an a priori assessment of the PaSR model is conducted. Its ability to reconstruct thermo-chemical quantities of interest is investigated along with model assumptions. Sub-grid quantities are extracted from the DNS to investigate the role of the cell reacting fraction. Various definitions are then proposed to estimate the characteristic chemical timescale in the PaSR model. Modeled chemical source terms and heat release rates are compared against the filtered quantities from DNS data of a two-dimensional, spatially developing, turbulent nonpremixed jet flame with detailed kinetics. The results demonstrate the importance of accounting for the fine structures quantities in the context of reactor-based models. A new formulation of the chemical timescale is proposed and provides improved overall predictions. Several issues are raised in the discussion, representing realistic prospects for further developments of the PaSR model as a SGS combustion closure for LES.

Large Eddy Simulation of Lean Premixed Swirling Flames via Dynamically Thickened Flame Model Coupling with the REDIM Chemistry Table

In this paper, the so-called REDIM–DTF sub-grid scale combustion model is proposed to improve the well-known dynamically thickened flame (DTF) combustion model by combining the DTF modeling strategy with the reaction-diffusion manifold (REDIM) chemistry table, which is generated from a detailed reaction mechanism. The new model is then used to calculate two lean premixed swirling flames in the PRECCINSTA combustor via large eddy simulation. The radial profiles of velocity, temperature, and species concentration, as well as the flame surface area and vortex structure are analyzed. The results are in good overall agreement with the corresponding experiment, and the performance of the REDIM–DTF model is very similar to that of the DTF model. The predicted CO mass fraction profiles, however, show a relatively larger discrepancy between the original DTF model and the proposed REDIM–DTF model, which is thought to be due to different reaction mechanisms used. As a smaller number of species transport equations are solved in the REDIM–DTF model, its computational efficiency is about 15% higher than that of the original DTF model.

REDIM-PFDF modelling of turbulent partially-premixed flame with inhomogeneous inlets using top-hat function for multi-stream mixing problem

Present study focuses on numerical investigation of two important partially-premixed inhomogeneous types of flames using REDIM-PFDF sub-grid scale combustion model. REDIM-PFDF model is based on the reaction-diffusion manifold (REDIM) technique and presumed filtered density function (PFDF) method. The detailed chemical kinetics is reduced into a two-dimensional chemistry table which is using the mass fractions of CO2 (YCO2) and N2 (YN2) as the reduced coordinates. The shapes of the Filtered Density Function (FDF) for YCO2 and YN2 are presumed to be clipped Gaussian and top-hat, respectively. We find that the top-hat function, in context of LES, is a suitable option to deal with multi-stream mixing problem. The comparison between LES and experimental data shows that the mixture fraction and mass fractions of several species are in a good agreement with the experimental data. The characteristics of two flames are also analyzed by comparing their OH mass fractions and CO2 production rates.

REDIM-PFDF Sub-grid Scale Combustion Modeling for Turbulent Partially-premixed Flame: Assessment of Combustion Modes

ABSTRACT The focus of the present study is to evaluate the capability of reaction-diffusion manifold (REDIM) technique combined with presumed filtered density function (PFDF) method to calculate the important characteristics of turbulent partially-premixed flame with near-homogeneous inlets and inhomogeneous inlets. The REDIM approach reduces detailed chemical kinetics into two-dimensional chemistry table, where mass fractions of CO2 and N2 are used as reduced coordinates. The fluctuation of the scalars within the LES filter volume is modeled by the PFDF method. Higher shear generated turbulent intensity at the mixing layer of the fuel and air streams provides the near-homogeneous inlet conditions. The radial statistics analysis of the mixture fraction, temperature and species mass faction, and their comparison with the experimental data show the good performance of the REDIM-PFDF model. Scatter plots show the better agreement for the inhomogeneous case. The near-homogeneous case exhibits little discrepancy with the measured distribution because of the limited ability of the present model to capture local extinction effects. Flame index parameter identifies the multiple combustion modes present at different locations. This parameter shows the presence of premixed and diffusion dominated types of combustion near the main jet exit plane in the inhomogeneous and near-homogeneous cases, respectively.

Review on Large Eddy Simulation of Turbulent Premixed Combustion in Tubes

This paper reviews the existing knowledge on the large eddy simulation (LES) of turbulent premixed combustion in empty tubes and obstructed tubes. From the view of model development in LES, this review comprehensively analyzes the development history and applicability of the important Sub-Grid Scale (SGS) viscosity models and SGS combustion models. LES is also used to combine flow and combustion models to reproduce industrial explosion including deflagration and detonation and the transition from deflagration to detonation (DDT). The discussion about models and applications presented here, leads readers to understand the progress of LES in the explosion of tube and reveals the deficiencies in this area.

Large eddy simulation of a partially pre-vaporized ethanol reacting spray using the multiphase DTF/flamelet model

Spray reactive flow finds application in various technical devices, and due to the complex nature, their optimization is very challenging, requiring proper modeling of turbulence/chemistry interactions as well as of the contribution from spray evaporation. This work presents a study of sub-grid scale combustion models, where relevant assumptions on multiphase coupling and their effects are analyzed in detail. For this purpose, two different flamelet approaches, i.e. progress variable spray flamelet and multi-regime gas flamelet are examined in an implementation coupled with the dynamic thickened flame model, along with which the impact of inlet inhomogeneities condition and droplet evaporation taking into account internal temperature gradient is also investigated. The numerical evaluation is carried out in large eddy simulations of a benchmark ethanol spray flame with partial pre-vaporization, where an Eulerian-Lagrangian numerical framework is adopted. The analysis demonstrated that the flame dynamics under consideration is governed by a close coupling between spray evaporation, turbulent dispersion and unsteady flame propagation at upstream shear layers. Results show that the spray flamelets built from counterflow partially-premixed spray flames achieved a better agreement with experiments, capturing the flame structure in terms of gas-phase temperature, OH mass fraction as well as spray statistics.

Pocket formation and behavior in turbulent premixed flames

Pocket formation is an important characteristic of turbulent premixed flames and understanding pocket behavior is key to developing high-fidelity numerical combustion models. In this study, a dual-burner experiment is used to study pockets in single- and dual-flame configurations and synchronized high-speed OH-planar laser-induced fluorescence and stereoscopic-particle image velocimetry imaging techniques are implemented to track flame pockets and the surrounding flow field. Statistical analysis of pocket origin and fate is performed using a novel tracking algorithm incorporating non-rigid image registration. Results show that pocket formation rates increase as a function of increasing inlet turbulence level; reactant pocket formation increases as a function of downstream distance, whereas product pocket formation decreases. Tracking reactant pocket lifetime shows that a majority of these pockets burn out and displacement speeds are characterized. Product pockets usually merge with the main flame surface, which could have an impact on local flame structure and propagation. Results presented in this study show that pocket behavior in turbulent flames can change local flame dynamics and it is important to capture these effects in sub-grid scale combustion models to accurately predict flame behavior.

Large Eddy Simulation of a Bluff Body Stabilised Premixed Flame Using Flamelets

Large Eddy Simulations of an unconfined turbulent lean premixed flame, which is stabilised behind a bluff body, are conducted using unstrained flamelets as the sub-grid scale combustion closure. The statistics from the simulations are compared with the corresponding data obtained from the experiment and it is demonstrated that the experimental observations are well captured. The relative positioning of the shear layers and the flame brush are analysed to understand the radial variations of the turbulent kinetic energy at various streamwise locations. These results are also compared to confined bluff body stabilised flames, to shed light on the relative role of incoming and shear driven turbulence on the behaviour of the flame brush and the turbulent kinetic energy variation across it.

Large eddy simulation of turbulent premixed piloted flame using artificial thickened flame model coupled with tabulated chemistry

A sub-grid scale (SGS) combustion model, which combines the artificial thickened flame (ATF) model with the flamelet generated manifold (FGM) tabulation method, is proposed. Based on the analysis of laminar flame structures, two self-contained flame sensors are used to track the diffusion and reaction processes with different spatial scales in the flame front, respectively. The dynamic formulation for the proposed SGS combustion model is also performed. Large eddy simulations (LESs) of Bunsen flame F3 are used to evaluate the different SGS combustion models. The results show that the proposed SGS model has the ability in predicting the distributions of temperature and velocity reasonably, while the predictions for the distributions of some species need further improvement. The snapshots of instantaneous normalized progress variables reveal that the flame is more remarkably and severely wrinkled at the flame tip for flame F3. More satisfactory results obtained by the dynamic model indicate that it can preserve the premixed flame propagation characteristics better.

Coal particle volatile combustion and flame interaction. Part II: Effects of particle Reynolds number and turbulence

Direct numerical simulation (DNS) is performed to characterize volatile combustion of isolated coal particles and closely spaced particle ensembles in laminar and turbulent flow. In part I of the paper the transient evolution of devolatilization and group flame effects were studied, Tufano et al. (2018). This analysis was limited to laminar flows and relatively low particle Reynolds numbers Rep. Here, we investigate the effect of large Rep and considerable levels of turbulence on the devolatilization and burning behavior of the particles. The complex physico-chemical interactions during PCC are characterized for laminar flow conditions first, before arrays of infinite particle layers are subjected to turbulence. Increasing Rep in laminar flow leads to delayed ignition of single particles with local extinction due to high upstream scalar dissipation rates and the formation of wake flame structures downstream of the particle. An attempt is made to recover the single particle results with a standard steady laminar flamelet approach, which is shown to work well at low Rep, but fails at high Reynolds numbers, where multi-dimensional effects occur and must be incorporated into flamelet modeling. It is found that applying standard film theory to model the effect of convection on devolatilization rates can lead to qualitatively wrong trends and up to 66% error in the peak devolatilization rate compared to the DNS results at Rep=8. The analysis of particle arrays in laminar flow shows a strong dependence of flame interactions on the values of Rep due to the different extent of the particle wakes. The occurrence of significant levels of turbulence introduces a wide range of additional chemical states due to the randomness of the turbulent fluctuations that can either act to increase or decrease the local strain, in turn weakening or enhancing particle interactions. For the studied conditions turbulence slightly promotes the mass release from the most upstream particle set, but considerably delays the volatile release from the downstream particles, which is explained by the different extents and degrees of interaction of the up- and downstream volatile flames. The present results are considered useful for the development of LES sub-grid scale combustion models for pulverized coal flames, such as flamelets and others.

Large-Eddy Simulation of Reacting Flows in Industrial Gas Turbine Combustor

The turbulent reacting flow in an industrial gas turbine combustor operating at 3 bar is computed using the large-eddy simulation paradigm. The subgrid-scale combustion is modeled using a collection of unstrained premixed flamelets including mixture stratification. The nonpremixed combustion mode is also included using a simple closure involving the scalar dissipation rate of the mixture fraction. Close attention is paid to maintain physical consistencies among subclosure models for combustion, and these consistencies are discussed on a physical basis. The importance of the nonpremixed mode and subgrid-scale mixture fraction fluctuations are investigated systematically. The results show that the subgrid-scale mixture fraction variance plays an important role and comparisons to measurements improve when contributions from the premixed and nonpremixed modes are included. These numerical results and observations are discussed on a physical basis along with potential avenues for further improvements.

Large eddy simulation of turbulent stratified combustion using dynamic thickened flame coupled with tabulated detailed chemistry

• A SGS model of DTF coupled with FGM is developed for investigating stratified flame. • LES transport equation for covariance in the DTF-FGM model is deduced. • An in-depth analysis for the flame structures of Cambridge stratified flames is achieved. A sub-grid scale (SGS) combustion model of the dynamic thickened flame (DTF) coupled with flamelet generated manifolds (FGM) approach (i.e. DTF-FGM) is developed to investigate Cambridge stratified flames. Two issues in developing the DTF-FGM model are addressed. Firstly, the copula method is extended to construct a multivariate joint probability density function (PDF) of the tabulated scalars. A modified laminar flamelet PDF is incorporated to describe the probable asymmetric multimodal distribution for the reaction progress variable. Secondly, large eddy simulation (LES) transport equations for the tabulated scalars, relevant sub-grid variances and covariance in the DTF-FGM model are deduced. LES of Cambridge stratified flames is performed with the DTF-FGM model and the widely used SGS combustion model which combines a presumed PDF with FGM (i.e. PPDF-FGM). Detailed comparisons between simulations and experiments are carried out to evaluate the SGS combustion models and mesh resolution effects. Meanwhile, an in-depth analysis for the flame structures of Cambridge stratified flames is also achieved. The results indicate that the developed SGS combustion model considering the effect of correlations between the mixture fraction and reaction progress variable has the ability to solve the stratified flame structures under LES mesh resolution exactly at a reasonable computational cost.

Impacts of plate slits on flame acceleration of premixed methane/air in a closed tube

The paper aims at revealing the interaction of various numbers of premixed methane/air jet flames in a closed duct. In the experiment, a high-speed video camera and pressure transducers are used to study the flame structure and pressure dynamics. In the numerical simulations, large eddy simulation (LES) with Power-Law combustion model is employed to investigate the interaction between the moving flame and vortices induced by the thin plate. The results demonstrate that the flame propagation for all plate configurations can be divided into four typical stages, i.e. hemispherical flame, finger-shaped flame, jet flame and bidirectional propagation flame. For three plate configurations, the jet flames merge together under the effect of the vortices, and the more slits with the same blockage ratio (BR) do not mean the stronger deflagration. It is observed that the peaks of flame tip speed and pressure growth rate decrease with the increase of the number of slits. The sub-grid scale combustion model, Power-Law model, coupled with sub-grid scale viscosity model, dynamic Smagorinsky-Lilly eddy viscosity model can well reproduce the flame propagation. By analyzing the numeric flow structure, the flame propagation mechanism of premixed methane/air flame propagation in a tube with various slits can be explained in the view of pure hydrodynamics.

LES of flame acceleration and DDT in small-scale channels

Large eddy simulation (LES) has been performed to investigate the process of flame acceleration (FA) and deflagration-to-detonation transition (DDT) in a small-scale channel filled with methane-oxygen mixture. A transport model for sub-grid kinetic energy was used to enclose the sub-grid fluxes while the sub-grid scale combustion was modeled by the artificially thickened flame (ATF) model. A fifth-order Weighted Essential Non-oscillatory (WENO) finite difference scheme was used to discretize the advection term of the governing equations. The results showed that the interaction of the flame with the flow and the pressure wave ahead led to the flame acceleration. The overdriven detonation led by the local explosion formed and the flame propagated upstream which then coupled with the precursor shock wave. A retonation wave propagating downstream was further observed. The experimental results substantiate the validation of the calculation and further identified the DDT mechanism.

Large Eddy Simulation of Turbulent Premixed Swirling Flames Using Dynamic Thickened Flame with Tabulated Detailed Chemistry

A sub-grid scale (SGS) combustion model by combining dynamic thickened flame (DTF) with flamelet generated manifolds (FGM) tabulation approach (i.e. DTF-FGM) is developed for investigating turbulent premixed combustion. In contrast to the thickened flame model, the dynamic thickening factor of the DTF model is determined from the flame sensor, which is obtained from the normalized gradient of the reaction progress variable from the one-dimensional freely propagating premixed flame simulations. Therewith the DTF model can ensure that the thickening of the flame is limited to the regions where it is numerically necessary. To describe the thermo-chemistry states, large eddy simulation (LES) transport equations for two characteristic scalars (the mixture fraction and the reaction progress variable) and relevant sub-grid variances in the DTF-FGM model are presented. As to the evaluation of different SGS combustion models, another model by utilizing the combination of presumed probability density function (PPDF) and FGM (i.e. PPDF-FGM) is also described. LES of two cases with or without swirl in premixed regime of the Cambridge swirl burner flames are performed to evaluate the developed SGS combustion model. The predicted results are compared with the experimental data in terms of the influence of different LES grids, model sensitivities to the thickening factor, the wrinkling factor, and the PPDF of characteristic scalars, the evaluation of different modelling approaches for the sub-grid variances of characteristic scalars, and the predictive capability of different SGS combustion models. It is shown that the LES results with the DTF-FGM model are in reasonable agreement with the experimental data, and better than the results with the PPDF-FGM approach due to its ability to predict better in regions where flame is not resolved.

Experimental and LES investigation of premixed methane/air flame propagating in a tube with a thin obstacle

In this paper, an experimental and numerical investigation of premixed methane/air flame dynamics in a closed combustion vessel with a thin obstacle is described. In the experiment, high-speed video photography and a pressure transducer are used to study the flame shape changes and pressure dynamics. In the numerical simulation, four sub-grid scale viscosity models and three sub-grid scale combustion models are evaluated for their individual prediction compared with the experimental data. High-speed photographs show that the flame propagation process can be divided into five stages: spherical flame, finger-shaped flame, jet flame, mushroom-shaped flame and bidirectional propagation flame. Compared with the other sub-grid scale viscosity models and sub-grid scale combustion models, the dynamic Smagorinsky–Lilly model and the power-law flame wrinkling model are better able to predict the flame behaviour, respectively. Thus, coupling the dynamic Smagorinsky–Lilly model and the power-law flame wrinkling model, the numerical results demonstrate that flame shape change is a purely hydrodynamic phenomenon, and the mushroom-shaped flame and bidirectional propagation flame are the result of flame–vortex interaction. In addition, the transition from “corrugated flamelets” to “thin reaction zones” is observed in the simulation.

Unstructured LES-CMC modelling of turbulent premixed bluff body flames close to blow-off

A finite volume Large Eddy Simulation-Conditional Moment Closure (LES-CMC) formulation is applied to turbulent premixed bluff body methane-air flames at conditions far from (A1) and close to (A4) blow-off. The unstructured topology of the CMC grid allows refinement in regions where turbulence inhomogeneity is expected, providing an improved description of the turbulence-chemistry interaction phenomenon, non-negligible at conditions investigated in this study. Subgrid scale (SGS) progress variable variance and scalar dissipation rate are closed with models associated with the SGS combustion, turbulence and molecular diffusion processes of premixed flames using detailed kinetics (GRI-3). The simulations effectively reproduce the general trends, especially considering the challenging conditions at very lean mixtures (down to φ=0.64): the characteristic ‘M’-shaped morphology of flame A4 and the significant increase in flame brush thickness compared to flame A1 is accurately replicated, although overall flame heights are under-predicted. Formaldehyde and OH distributions are compared with PLIF measurements and excellent agreement is reported supporting previous experimental observations: for flame A1, OH is seen throughout the recirculation zone while CH2O is present only close to the shear layer along which heat release occurs in a thin, unbroken region of CH2O/OH overlap, disturbed only occasionally by vortex-like structures. For flame A4 close to blow-off, the calculation generally shows appreciable quantities of CH2O throughout the recirculation zone with isolated pockets of OH, surrounded by high heat release, in agreement with experimental findings. The simulation further evidences that in regions experimentally void of both CH2O and OH, large quantities of partially reacted fluid are present which entered the recirculation zone from the top. Quantitative comparisons of flame surface density and local flame curvature statistics are also calculated which agree well with experimental data, suggesting a comprehensive description of the physico-chemical processes of the proposed modelling strategy.

Large eddy simulation of turbulent premixed combustion using tabulated detailed chemistry and presumed probability density function

ABSTRACTA method of chemistry tabulation combined with presumed probability density function (PDF) is applied to simulate piloted premixed jet burner flames with high Karlovitz number using large eddy simulation. Thermo-chemistry states are tabulated by the combination of auto-ignition and extended auto-ignition model. To evaluate the predictive capability of the proposed tabulation method to represent the thermo-chemistry states under the condition of different fresh gases temperature, a-priori study is conducted by performing idealised transient one-dimensional premixed flame simulations. Presumed PDF is used to involve the interaction of turbulence and flame with beta PDF to model the reaction progress variable distribution. Two presumed PDF models, Dirichlet distribution and independent beta distribution, respectively, are applied for representing the interaction between two mixture fractions that are associated with three inlet streams. Comparisons of statistical results show that two presumed PDF models for the two mixture fractions are both capable of predicting temperature and major species profiles, however, they are shown to have a significant effect on the predictions for intermediate species. An analysis of the thermo-chemical state-space representation of the sub-grid scale (SGS) combustion model is performed by comparing correlations between the carbon monoxide mass fraction and temperature. The SGS combustion model based on the proposed chemistry tabulation can reasonably capture the peak value and change trend of intermediate species. Aspects regarding model extensions to adequately predict the peak location of intermediate species are discussed.