Simulation of turbulent combustion in DLR Scramjet

Turbulent combustion in a DLR (German Aerospace Center) Scramjet engine was simulated using the newly-proposed Partially Resolved Numerical Simulation (PRNS) procedure. The PRNS procedure uses temporal filtering to define large-scale turbulence, and the model developed to account for unresolved scales is grid independent. No problem about inner commutation error and inconsistencies will arise from the PRNS, while such issues are of concern in traditional Large Eddy Simulation (LES) methods. The mean results have good agreement with the experiment data and the flow structures with small scales are well resolved.

A robust autoregressive long-term spatiotemporal forecasting framework for surrogate-based turbulent combustion modeling via deep learning

This paper systematically develops a high-fidelity turbulent combustion surrogate model using deep learning.We construct a surrogate model to simulate the turbulent combustion process in real time,based on a state-ofthe-art spatiotemporal forecasting neural network.To address the issue of shifted distribution in autoregressive long-term prediction,two training techniques are proposed:unrolled training and injecting noise training.These techniques significantly improve the stability and robustness of the model.Two datasets of turbulent combustion in a combustor with cavity and a vitiated co-flow burner(Cabra burner)have been generated for model validation.The effects of model architecture,unrolled time,noise amplitude,and training dataset size on the long-term predictive performance are explored.The well-trained model can be applicable to new cases by extrapolation and give spatially and temporally consistent results in long-term predictions for turbulent reacting flows that are highly unsteady.

Theoretical analysis of flamelet model for supersonic turbulent combustion

Being physically intuitionistic and computationally efficient, the flamelet model has obtained increasing attention and becomes popular in numerical simulations of supersonic combustion. However, the flamelet model was firstly built for simulations of the low-speed flows. Therefore it is still unclear whether the assumption of the fiamelet model is reasonable in supersonic combustion. This paper tries to identify the existence of the flamelet mode in supersonic combustion. Firstly, the interaction of the turbulent fluctuation and the flame is discussed, based on which the combustion regime is distinguished. Then the charac- teristic length scale of the turbulent flows and the combustion inside the scramjet combustor are calculated and compared, which are used to identify whether the assumption of the flamelet model is reasonable in supersonic flows. The results show that for premixed combustion, due to the low fluctuation velocity in the recirculafion zone and the shear layer, the assumption of the flamelet model is established for all the flight Mach numbers, while for non-premixed combustion the assumption is also established for most of the flight Mach numbers except for very high Mach number under which the slow reaction mode dominates the combustion. In order to quantitatively examine the combustion mode, numerical calculations are performed to simulate the strut-injection supersonic combustion which has been experimentally investigated in German Aerospace Center. The results show that the supersonic combustion occurs in the fully-developed turbulent regions and the assumption of the flamelet model is established in the whole flow field.

A hybrid LES (Large Eddy Simulation)/assumed sub-grid PDF (Probability Density Function) model for supersonic turbulent combustion

A hybrid LES (Large Eddy Simulation)/assumed sub-grid PDF (Probability Density Function) closure model has been devel-oped for supersonic turbulent combustion. Scalar transport equations for all species in a given chemical kinetic mechanism were solved, which are necessary in the supersonic combustion where the non-equilibrium chemistry is essentially involved. The clipped Gaussian PDF of temperature and multivariate ? PDF of composition were used to close the sub-grid chemical sources that appear in the conservation equations. The sub-grid variances of temperature and composition were constructed based on scale similarity approach. A semi-implicit approach based on the PDF model was proposed to tackle the resulting numerical stiffness associated with finite rate chemistry. The model was applied to simulate a supersonic, coaxial H2-air burner, where both the mean and rms (root mean square) results were compared with the experimental data. In general, good agree-ments were achieved, which indicated that the present sub-grid PDF method could work well in simulating supersonic turbu-lent combustion. Moreover, the calculation showed that the sub-grid fluctuations of temperature and major species in the combustion region were of the order of 10%-20% of their rms, while the sub-grid fluctuation of hydroxyl might be as high as 40%-50% of its rms.

Effect of Hydrogen-Rich Fuels on Turbulent Combustion of Advanced Gas Turbine

The effect of high hydrogen fuel on turbulent combustion in advanced gas turbine combustor with our newly designed arrayed-vanes premixer was studied by large eddy simulation(LES).The dynamic Smagorinsky model is used to calculate the subgrid stress.Finite-rate chemistry is included using a four steps mechanisms.A thickened flame model was used to deal with the reaction rate.The transport and thermal properties are obtained by CHEMKIN packages.The results show that with the increase of hydrogen content,the wake recirculation zone and central toroidal recirculation zone separate.The turbulent fluctuations of H_(2)/air flame first decreases and then increases.For the response of turbulent flame,the results show that the flame brush is narrow and short with the increase of hydrogen content.When the hydrogen content is low,the syngas/air flame can also propagate in the high-speed flow.Therefore,for different flames,the position of outer propagating flame is almost the same.The results also show that the fluctuation of flame intensity increases with the increase of hydrogen content.Although the increase of hydrogen content shortens the chemical reaction time and suppresses the perturbation of turbulent eddies,the cellular instability may further enhance the fluctuation of flame intensity.

NUMERICAL SIMULATION OF METHANE-AIR TURBULENT JET FLAME USING A NEW SECOND-ORDER MOMENT MODEL

A new second-order moment model for turbulent combustion is applied in the simulation of methane-air turbulent jet flame. The predicted results are compared with the experimental results and with those predicted using the well-known EBU-Arrhenius model and the original second-order moment model. The comparison shows the advantage of the new model that it requires almost the same computational storage and time as that of the original second-order moment model, but its modeling results are in better agreement with experiments than those using other models. Hence, the new second-order moment model is promising in modeling turbulent combustion with NOx formation with finite reaction rate for engineering application.

An implicit relation between temperature and reaction rate in the SLFM

The Arrhenius law implies that reaction rate is a continuous function of temperature. However,the steady laminar flamelet model(SLFM) does not explicitly give this functional relationship. The present study addresses this relation in the SLFM.It is found that reaction rate is not continuous in the mixture-fraction space.As a result,the SLFM is unable to predict local extinction and reignition.Furthermore,we use the unstable branch of the'S-curve'to fill the gap between steady burning branch and extinction one,and find that this modification leads to a continuous dependent of reaction rate on temperature.Thus the modified SLFM can describe the local extinction and reignition.

CFD study of non-premixed swirling burners: Effect of turbulence models

This research investigates a numerical simulation of swirling turbulent non-premixed combustion.The effects on the combustion characteristics are examined with three turbulence models:namely as the Reynolds stress model,spectral turbulence analysis and Re-Normalization Group.In addition,the P-1 and discrete ordinate(DO)models are used to simulate the radiative heat transfer in this model.The governing equations associated with the required boundary conditions are solved using the numerical model.The accuracy of this model is validated with the published experimental data and the comparison elucidates that there is a reasonable agreement between the obtained values from this model and the corresponding experimental quantities.Among different models proposed in this research,the Reynolds stress model with the Probability Density Function(PDF)approach is more accurate(nearly up to 50%)than other turbulent models for a swirling flow field.Regarding the effect of radiative heat transfer model,it is observed that the discrete ordinate model is more precise than the P-1 model in anticipating the experimental behavior.This model is able to simulate the subcritical nature of the isothermal flow as well as the size and shape of the internal recirculation induced by the swirl due to combustion.

Visual analytics of combustion on time-varying turbulent-flow

Visualization is crucial for analyzing the turbulent combustion simulation.Time-varying data allows us to investigate the evolution process of the turbulent flow field.To study the combustion effects,we calculated the enstrophy of the flow field since high enstrophy region can display valuable features,and extract components based on these features.We isolated large components to track their behaviors and characterized them using volume and spatial locations,which helps scientists to explore the dynamics and temporal changes of intense events individually.We analyzed the components’structures and visualized them in contouring and statistical charts.

A review of engine application and fundamental study on turbulent premixed combustion of hydrogen enriched natural gas

The application and fundamental study on turbulent premixed combustion of hydrogen enriched natural gas is reviewed in this paper.Discussions include the combustion characteristics of direct injection engine fueled with hydrogen enriched natural gas,visualization study of direct injection combustion of hydrogen enriched natural gas using a constant volume vessel,and the fundamental study of turbulent premixed combustion of hydrogen enriched natural gas.The effect of additional hydrogen on the combustion process of natural gas engine is investigated from the fundamental view of the interaction between combustion reaction and turbulent flow.

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 hydrogen/air scramjet combustion using tabulated thermo-chemistry approach

Large eddy simulations （LES） have been performed to investigate the flow and combustion fields in the scramjet of the German Aerospace Center （DLR）. Turbulent combustion is mod- eled by the tabulated thermo-chemistry approach in combination with the presumed probability density function （PDF）. A/3-function is used to model the distribution of the mixture fraction, while two different PDFs, g-function （Model I） and //-function （Model II）, are applied to model the reaction progress. Temperature is obtained by solving filtered energy transport equation and the reaction rate of the progress variable is rescaled by pressure to consider the effects of compressibil- ity. The adaptive mesh refinement （AMR） technique is used to properly capture shock waves, boundary layers, shear layers and flame structures. Statistical results of temperature and velocity predicted by Model II show better accuracy than that predicted by Model I. The results of scatter points and mixture fraction-conditional variables indicate the significant differences between Model I and Model II. It is concluded that second moment information in the presumed PDF of the reaction progress is very important in the simulation of supersonic combustion. It is also found that an unstable flame with extinction and ignition develops in the shear layers of bluff body and a fuel- rich partially premixed flame stabilizes in the central recirculation bubble.

Effect of turbulence on NO formation in swirling combustion

Turbulence affects both combustion and NO formation. Fluctuation correlations are ideally used for quantitative analysis. From the instantaneous chemical reaction rate expression,ignoring the third-order correlation terms, the averaged reaction rate will have four terms, including the term of averaged-variable product, a concentration fluctuation correlation term, and temperature-concentration fluctuation correlation term. If the reaction-rate coefficient is denoted as K, the temperature fluctuation would be included in the K fluctuation. In order to quantitatively study the effect of turbulence on NO formation in methane-air swirling combustion, various turbulencechemistry models are tested. The magnitudes of various correlations and their effects on the time-averaged reaction rate are calculated and analyzed, and the simulation results are compared with the experimental measurement data. The results show that among various correlation moments, the correlation between the reaction-rate coefficient K fluctuation with the concentration fluctuation is most important and is a strong nonlinear term.

Numerical study of high-pressure GO_2/GH_2 combustion of a single-element injector

A single-element shear-coaxial combustor using gaseous hydrogen(GH2) and oxygen(GO2) was designed and hot-tested.The wall temperature was measured.The combustion flowfield of this GH2 /GO2 single-element combustor was modeled by RANS(Reynolds Averaged Navier-Stokes) and LES(Large Eddy Simulation) methods respectively.The impact of using various turbulence and turbulent combustion models was investigated to obtain the model combination which best represented the experimental data in the RANS modeling.The flamelet model was used in the LES modeling and the validity of its application to the GH2 /GO2 combustion in the combustor was carefully examined.The combustor wall heat flux distributions of both RANS and LES results show good agreement with the experimental data.The experimental wall temperature distribution can be clearly explained through analyzing the inner flowfield structure.The results indicate that both RANS and LES used in this paper can give good predictions of the development of the whole flowfield and the combustion completion length.LES can resolve large-scale flow motions in the combustor and accurately predict the influence of the wall heat loss on the combustion efficiency.

Explore artificial neural networks for solving complex hydrocarbon chemistry in turbulent reactive flows

Global warming caused by the use of fossil fuels is a common concern of the world today.It is of practical importance to conduct in-depth fundamental research and optimal design for modern engine combustors through However,complex hydrocarbon chemistry,an indispensable component for predictive modeling,is computahigh-fidelity computational fluid dynamics(CFD),so as to achieve energy conservation and emission reduction.tionally demanding,Its application in simulation-based design optimization,although desirable,is quite limited.To address this challenge,we propose a methodology for representing complex chemistry with artificial neural networks(ANNs),which are trained with a comprehensive sample dataset generated by the Latin hypercube sampling(LHS)method.With a given chemical kinetic mechanism,the thermochemical sample data is able to cover the whole accessible pressure/temperature/species space in various turbulent flames.The ANN-based model consists of two different layers:the self-organizing map(SOM)and the back-propagation neural network(BPNN).The methodology is demonstrated to represent a 30-species methane chemical mechanism.The obtained ANN model is applied to simulate both a non-premixed turbulent flame(DLR_A)and a partially premixed turbulent flame(Flame D)to validate its applicability for different flames.Results show that the ANN-based chemical kinetics can reduce the computational cost by about two orders of magnitude without loss of accuracy,The proposed methodology can successfully construct an ANN-based chemical mechanism with significant ffciency gain and a broad scope of applicability,and thus holds a great potential for complex hydrocarbon fuels.

Numerical simulations of turbulent flows in aeroramp injector/gas-pilot flame scramjet

To uncover the internal flow characteristics in an ethylene-fueled aeroramp injector/gaspilot（ARI/G-P）flame scramjet,a Reynolds-averaged Navier-Stokes（RANS）solver is constructed under a hybrid polyhedral cell finite volume frame.The shear stress transport（SST）k-x model is used to predict the turbulence,while the Overmann’s compressibility corrected laminar flamelet model is adopted to simulate the turbulent combustion.Nonreactive computations for Case 1（G-P jet on）,Case 2（ARI jets on）,and Case 3（both ARI and G-P jets on）were conducted to analyze the mixing mechanism,while reactive Cases 4–7 at equivalent ratios of 0.380,0.278,0.199 and0.167 respectively were calculated to investigate the flame structure and combustion modes.The numerical results are compared well to those of the experiments.It is shown that the G-P jet plays significant role in both the fuel/air mixing and flame holding processes;the combustion for the four reactive cases takes place intensively in the regions downstream of the ARI/G-P unit;Cases 4 and 5are under subsonic combustion mode,whereas Cases 6 and 7 are mode transition critical and supersonic combustion cases,respectively;the mode transition equivalent ratio is approximately 0.20.

A dual timescale model for micro-mixing and its application in LES-TPDF simulations of turbulent nonpremixed flames

The numerical simulation of modern aero-engine combustion chamber needs accurate description of the interaction between turbulence and chemical reaction mechanism. The Large Eddy Simulation(LES) method with the Transported Probability Density Function(TPDF) turbulence combustion model is promising in engineering applications. In flame region, the impact of chemical reaction should be considered in TPDF molecular mixing model. Based on pioneer research, three new TPDF turbulence-chemistry dual time scale molecular mixing models were proposed tentatively by adding the chemistry time scale in molecular mixing model for nonpremixed flame. The Aero-Engine Combustor Simulation Code(AECSC) which is based on LES-TPDF method was combined with the three new models. Then the Sandia laboratory's methane-air jet flames: Flame D and Flame E were simulated. Transient simulation results show that all the three new models can predict the instantaneous combustion flow pattern of the jet flames. Furthermore,the average scalar statistical results were compared with the experimental data. The simulation result of the new TPDF arithmetic mean modification model is the closest to the experimental data:the average error in Flame D is 7.6% and 6.6% in Flame E. The extinction and re-ignition phenomena of the jet flames especially Flame E were captured. The turbulence time scale and the chemistry time scale are in different order in the whole flow field. The dual time scale TPDF combustion model has ability to deal with both the turbulence effect and the chemistry reaction effect, as well as their interaction more accurately for nonpremixed flames.

A decoupled procedure for convection-radiation simulation in scramjets

Following an order analysis of key parameters, a decoupled procedure for simulation of convection-radiation heat transfer problems in supersonic combustion ramjet(scramjet) engine was developed. The radiation module of the procedure consisted of Perry 5GG weighted sum gray gases model for spectral property calculation and discrete ordinates method S4 scheme for radiative transfer computation, while the flow field was computed using the Favrè average conservative Navier-Stokes(N-S) equations, in conjunction with Menter's k-ω SST two-equation model. A series of 2D supersonic nonreactive turbulent channel flows of radiative participants with selective parameters were simulated for validation purpose. Radiative characteristics in DLR hydrogen fueled and NASA SCHOLAR ethylene fueled scramjets were numerically studied using the developed procedure. The results indicated that the variations of spatial distributions of the radiative source and total absorption coefficient are highly consistent with those of the temperature and radiative participants, while the spatial distribution of the incident radiation spreads wider. It also demonstrated that the convective heating is significantly affected by the complexity of the flow field, such as the shock wave/boundary layer interactions, while the radiative heating is simply an integral effect of the whole flow field. Although the radiative heating in the combustion chambers reaches a certain level, an order of magnitude of 10 k W/m2, it still contributes little to the total heat transfer(<7%).

A numerical study of counterflow diffusion flames of methane/air at various pressures

A numerical study of the counterflow diffusion flames of methane/air at both subcritical and supercritical pressures,which have very important applications in the air-breathing rocket and advanced gas turbine engines,is conducted to obtain fundamental understanding of the flame characteristics.The analysis is based on a general mathematical formulation and accommodates a unified treatment of general fluids thermodynamics and accurate calculations of thermophysical properties.Results reveal that the maximum flame temperature occurs on the fuel-rich side for low-pressure conditions and shifts toward the stoichiometric position when the pressure increases.The maximum flame temperature increases with an increasing pressure,but decreases with an increasing strain rate.The flame width is inversely proportional to the square root of the product of the pressure and strain rate as■■1 p·a2/1.The total heat release rate varies with the pressure and strain rate in a relationship of Q release ■(p·a)0.518.An increased pressure leads to a slightly more complete combustion process near the stoichiometric position,but its effect on NO production is minor.Under the test conditions,variations of the strain rate have significant impacts on the formation of major pollutants.An increased strain rate leads to the decreased mole fraction of CO in the fuel-rich region and significantly reduced NO near the stoichiometric position.

Experimental and Numerical Analysis of Natural Bio and Syngas Swirl Flames in a Model Gas Turbine Combustor

Turbulent reacting flows in a generic swirl gas turbine combustor model are investigated both numerically and experimentally.In the investigation,an emphasis is placed upon the external flue gas recirculation,which is a promising technology for increasing the efficiency of the carbon capture and storage process,which,however,can change the combustion behaviour significantly.A further emphasis is placed upon the investigation of alternative fuels such as biogas and syngas in comparison to the conventional natural gas.Flames are also investigated numerically using the open source CFD software OpenFOAM.In the numerical simulations,a laminar flamelet model based on mixture fraction and reaction progress variable is adopted.As turbulence model,the SST model is used within a URANS concept.Computational results are compared with the experimental data,where a fair agreement is observed.