Non-adiabatic Flame Research Articles

Application of dense neural networks for manifold-based modeling of flame-wall interactions

Artifical neural networks (ANNs) are universal approximators capable of learning any correlation between arbitrary input data with corresponding outputs, which can also be exploited to represent a low-dimensional chemistry manifold in the field of combustion. In this work, a procedure is developed to simulate a premixed methane-air flame undergoing side-wall quenching utilizing an ANN chemistry manifold. In the investigated case, the flame characteristics are governed by two canonical problems: the adiabatic flame propagation in the core flow and the non-adiabatic flame- wall interaction governed by enthalpy losses to the wall. Similar to the tabulation of a Quenching Flamelet-Generated Manifold (QFM), the neural network is trained on a 1D head-on quenching flame database to learn the intrinsic chemistry manifold. The control parameters (i.e. the inputs) of the ANN are identified from thermo-chemical state variables by a sparse principal component analysis (PCA) without using prior knowledge about the flame physics. These input quantities are then transported in the coupled CFD solver and used for manifold access during simulation runtime. The chemical source terms are corrected at the manifold boundaries to ensure bounded- ness of the thermo-chemical state at all times. Finally, the ANN model is assessed by comparison to simulation results of the 2D side-wall quenching (SWQ) configuration with detailed chemistry and with a flamelet-based manifold (QFM).

Flame stability under flow-induced anisotropic diffusion and heat loss

A two-dimensional model for premixed flames accounting for flow-enhanced diffusion or Taylor dispersion and heat loss is investigated. This is the first analytical study addressing the effect of Taylor dispersion on the thermo-diffusive instabilities of non-adiabatic flames. It is also the first numerical study coupling flame instability with Taylor dispersion. A linear stability analysis is carried out in the limit of infinite Zeldovich number β. This leads to a dispersion relation, generalising classical relations in the literature, and involving three parameters, l (the reduced Lewis number), p (the Taylor-dispersion coefficient which is proportional to the Peclet number), and κ (the heat loss coefficient). Stability diagrams are determined and their implications on the cellular and oscillatory instabilities are discussed. A Kuramoto–Sivashinsky type equation incorporating the parameters l, p and κ and characterising the flame dynamics in the weakly non-linear regime near the onset of the cellular instability is derived. The theoretical results demonstrate the ability of Taylor dispersion and heat loss to significantly affect the flame stability. In particular, the oscillatory instability is found to be promoted by an increase in κ and hampered by an increase in p. On the other hand, both p and κ have a destabilising effect in connection with the cellular instability. Also, the theory provides a formula predicting the typical size of cells first emerging from the cellular instability which is found to be a decreasing function of p.Numerical simulations are carried out illustrating and significantly extending the analytical findings. Particular attention is devoted to the influence of β. In particular, we reconcile apparent quantitative and sometimes qualitative discrepancies between the numerical and theoretical predictions, which are found to be more pronounced for larger values of p. Luckily, the asymptotic theory is found to be robust in the sense that its predictions are recovered numerically if β is taken large enough, although such predictions may be questionable for realistic values of β. In general, the effect of β on flame stability is found to be opposite to that of p: an increase in p or a decrease in β have a destabilising effect in connection with the cellular instability, and a stabilizing effect in connection with the oscillatory instability; both instabilities are promoted by an increase in κ.

Optimization of Filtered Rayleigh Scattering for the Measurement of Pressure and Temperature

ABSTRACT The stand-off measurements of temperature and pressure are important for a variety of physical and engineering processes. Filtered Rayleigh scattering (FRS) configurations that enhance and suppress sensitivity to temperature are modeled and demonstrated using premixed, atmospheric methane-air Hencken burner generating temperatures up to 2100 K. For a characteristic flame gas mixture, modeling of the FRS signal using a molecular iodine filter shows that conditions exist for which the signal has a greater dynamic range or does not change significantly with respect to high gas temperatures. Laser Rayleigh scattering (LRS) measurements coupled with nonadiabatic flame simulations validate the temperatures, while FRS measurements are compared with the model predictions for different observation angles and iodine cell conditions. Results are discussed in the context of making direct measurements of gas pressure at high temperatures and within turbulent flames.

Improved prediction model for H2/CO combustion risk using a calculated non-adiabatic flame temperature model

During severe nuclear power plant (NPP) accidents, a H2/CO mixture can be generated in the reactor pressure vessel by core degradation and in the containment as well by molten corium-concrete interaction. In spite of its importance, a state-of-the-art methodology predicting H2/CO combustion risk relies predominantly on empirical correlations. It is therefore necessary to develop a proper methodology for flammability evaluation of H2/CO mixtures at ex-vessel phases characterized by three factors: CO concentration, high temperature, and diluents. The developed methodology adopted Le Chatelier’s law and a calculated non-adiabatic flame temperature model. The methodology allows the consideration of the individual effect of the heat transfer characteristics of hydrogen and carbon monoxide on low flammability limit prediction. The accuracy of the developed model was verified using experimental data relevant to ex-vessel phase conditions. With the developed model, the prediction accuracy was improved substantially such that the maximum relative prediction error was approximately 25% while the existing methodology showed a 76% error. The developed methodology is expected to be applicable for flammability evaluation in chemical as well as NPP industries.

A projection procedure to obtain adiabatic flames from non-adiabatic flames using heat flux method

Laminar burning velocity SL at elevated temperature Tu and its temperature dependence coefficient α in SLSL0=(TuTu0)α are important parameters for industrial applications. However, experimental systems with high unburned gas temperatures may encounter pre-dissociation, leading to significant data scattering in the measurements. To negate this, the present work proposes a projection procedure to obtain adiabatic flame parameters at various unburned gas temperatures using non-adiabatic flames on a heat flux burner, by which the preheating can be achieved within much shorter time scale than, e.g., in conventional spherical flame methods, and the advantage of good data consistency in the heat flux method is kept. Burning velocity experiments were carried out with CH4+air atmospheric flames covering Tu = 298–473 K, and the results show good agreement with the proposed projection equations. OH* spontaneous emission profiles were measured, indicating that the projection may extend to other flame characteristics. Uncertainty of the projection process was evaluated and comparisons were made with six popular kinetic mechanisms: GRI-Mech, FFCM-1 mech, Konnov mechanism, Glarborg mechanism, San Diego mechanism and Aramco mechanism. It is found that the simulated coefficients α are higher than experimental data especially at rich conditions; this is also found for literature values of high unburned gas temperature experiments. Possible reasons for this divergence are discussed.

Properties of nonadiabatic combustion waves in competitive exothermic reactions

We study the properties of travelling combustion waves in a diffusional thermal model with a two-step competitive exothermic reaction mechanism considering the system in one spatial dimension under nonadiabatic conditions. Based on the notion of the crossover temperature, the model is first examined analytically to predict the behaviour of travelling combustion waves in the limit of large activation energies. It is then studied numerically over a wide range of parameter values, such as those describing the ratios of the enthalpies, pre-exponential factors and activation energies. It is demonstrated that the nonadiabatic flame speed as a function of these parameters is a single-valued monotonic function with two flame regimes identified, each of which represents the region of the parameter values when one reaction dominates the other, while the flame speed as a function of the activation energy of the reaction R2 (the exothermicity parameter) is either a c- or m-shaped. The extinction conditions for these two flame regimes are investigated analytically.

A flammability limit model for hydrogen-air-diluent mixtures based on heat transfer characteristics in flame propagation

Abstract Predicting lower flammability limits (LFL) of hydrogen has become an ever-important task for safety of nuclear industry. While numerous experimental studies have been conducted, LFL results applicable for the harsh environment are still lack of information. Our aim is to develop a calculated non-adiabatic flame temperature (CNAFT) model to better predict LFL of hydrogen mixtures in nuclear power plant. The developed model is unique for incorporating radiative heat loss during flame propagation using the CNAFT coefficient derived through previous studies of flame propagation. Our new model is more consistent with the experimental results for various mixtures compared to the previous model, which relied on calculated adiabatic flame temperature (CAFT) to predict the LFL without any consideration of heat loss. Limitation of the previous model could be explained clearly based on the CNAFT coefficient magnitude. The prediction accuracy for hydrogen mixtures at elevated initial temperatures and high helium content was improved substantially. The model reliability was confirmed for H2-air mixtures up to 300 °C and H2-air-He mixtures up to 50 vol % helium concentration. Therefore, the CNAFT model developed based on radiation heat loss is expected as the practical method for predicting LFL in hydrogen risk analysis.

Flame Balls in a non-uniform reactive mixture: preferential diffusion, heat-loss and stability

The paper examines the existence and stability of axisymmetric flame balls in a non-uniform reactive mixture corresponding to a mixing layer taking into account preferential diffusion and volumetric heat-loss. The mixture's non-uniformity is measured with a non-dimensional parameter ε which is inversely proportional to the square root of the Damköhler's number. The investigation is carried out analytically in the limit of large activation energy and small values of ε, and numerically in the general case. New simple formulas accounting for preferential diffusion are derived which determine in particular the thermal energy and location of the flame ball; these may be argued to represent the minimum ignition energy and optimum spark location for a successful forced ignition of the diffusion flame in the mixing layer. A new free boundary problem (FBP) with two dependent variables is derived which describes non-adiabatic flame balls subject to volumetric heat-loss from the burnt gas. For small ε, the analytical solution to the FBP shows that the main effect of weak non-uniformity can be understood in a simple way if the volume of the distorted flame ball is characterised by an equivalent radius which is plotted versus a heat-loss parameter κ. Specifically, the curve is the same inverse-C shaped curve found in the literature in the uniform case () but shifted to the left by an amount, proportional to , which explicitly accounts for all parameters. The numerical investigation addresses the existence of the axisymmetric flame balls and their stability within two models familiar in studies on flame balls in uniform mixtures, namely the ‘far-field losses model’ where heat-losses from the burnt and unburnt gas are accounted for, and the ‘near-field losses model’ adopted in our analytical investigation, where heat-loss from the unburnt gas is neglected. Typically four regions are determined in the κ-ε plane for fixed Lewis numbers which identify conditions for the existence of either the flame ball, the diffusion flame or of both. This subdivision is argued to provide useful insight regarding the possible modes of burning in the mixing layer. A particularly interesting type of solutions identified for moderate values of ε corresponds to ring-shaped flame balls, termed ‘flame rings’, in regions where the diffusion flame cannot exist. As for the stability of the flame balls, we have found these to be typically unstable, as expected for their spherical counterparts. However, we have also determined these to be stable in special circumstances requiring low Lewis numbers and the presence of heat-losses and depending on the non-uniformity parameter ε. Furthermore, an increase in ε was found to play a stabilising effect, at least for the cases considered.

Chemiluminescence of Burner-Stabilized Premixed Laminar Flames

ABSTRACTThe OH*, CH* and CO chemiluminescence signals of methane/air premixed laminar flames stabilized over a nonadiabatic porous plug burner are compared to the signals measured from a nearly adiabatic conical flame in a series of experiments. The impact of reactant stream temperature is also characterized. A numerical study based on 1-D flame models then follows to support the experimental results. It is found both in experiments and in simulations that the linear relationship between the mixture flowrate and the chemiluminescence intensities is no longer valid when flames are closely attached to the burner surface due to the heat transfer between the flame and the burner. The transition between the linear and the nonlinear regimes is identified as the gas flow velocity drops below the adiabatic laminar burning velocity calculated at the bulk temperature of the flow leaving the burner. When the mass flowrate is kept constant, preheating of the reactant stream increases the chemiluminescence intensity for a freely propagating flame, but has almost no impact for a burner-stabilized flame. It is finally found that the OH* and CH* chemiluminescence intensities correlate with the burnt gas temperature for the adiabatic but also the nonadiabatic flames. The underlying physical mechanisms are discussed. Finally, the evolution of the CH*/OH* ratio with the inlet gas velocity is discussed.

Analysis and flamelet modelling for laminar pulverised coal combustion considering the wall effect

In this work, the wall effect on the characteristics of pulverised coal combustion is examined. To this end, laminar pulverised coal flames stabilised near a planar non-slip wall with isothermal and adiabatic boundary conditions are simulated with detailed chemistry. The phenomenon of side-wall quenching is observed in the non-adiabatic case due to significant wall heat losses. The temperature difference between the adiabatic flame and non-adiabatic flame at the planar non-slip wall can be as high as 1200 K. It is found that the distributions of different species show various sensitivities to the wall effect. Particularly, the species involved in the volatile matter and char off-gases are significantly influenced by the wall heat losses. In addition, the conventional flamelet model for pulverised coal combustion is extended to consider the heat transfers between the flame and wall. The performance of the extended flamelet model on prediction of the thermo-chemical quantities in both adiabatic and non-adiabatic flames is evaluated through a priori tests. The results show that the thermo-chemical quantities in both adiabatic and non-adiabatic flames can be overall reproduced by the extended flamelet model at both upstream and downstream locations. Interestingly, it is found that the gaseous temperature and CH mass fraction in the non-adiabatic flame can be better predicted by the extended flamelet model than those in the adiabatic flame at the downstream location.

Flame Propagation Model and Combustion Phenomena: Observations, Characteristics, Investigations, Technical Indicators, and Mechanisms

Critical conditions are usually obtained for ignition in a self-heating solid system consisting of two components generating heat independently, one component being inexhaustible and the other exhaustible by either simple first order or autocatalytic reaction. Ignition depends upon whether the exhaustible component can cause a temperature rise in excess of the upper stationary, but unstable, value possible for the inexhaustible component reacting alone. The system provides a theoretical model for some commonly occurring examples of self-heating and ignition in porous solids containing oxidisable oils. It is shown that: (a) the ignition criterion of the model, which involves a nonarbitrary critical temperature increase, has a high degree of physical reality; (b) the model is, in principle, capable of predicting ignition from primary kinetic and thermal data; (c) it is likely to be possible often to make a reliable prediction of critical size for self-ignition in a two-component system at ordinary atmospheric temperatures by a simple extrapolation from small-scale ignition data, obtained at higher temperatures, in the same way as for ignition due to a single reaction. Examination of both adiabatic and non-adiabatic flame theories showed that a 'steady state' exists only under the special condition that a heat sink exists at the initial temperature. For the general case of freely propagating, non-adiabatic flames only a quasi-steady state can be achieved.

Propagation and quenching of dual-front flames in binary-fuel mixtures

ABSTRACTAn analytical model was developed for the study of nonadiabatic flame propagation and quenching in binary mixtures of independently reacting fuels. Despite its simple formulation, in which step-wise reactions start at an externally imposed ignition temperature and proceed at a constant rate, the model reproduces the well-known regimes of dual-front propagation: front separation, control, and merging. The results indicate that, depending on the relative concentration of the fuels, the two fronts of the flame may either quench separately – each front at a different level of heat losses in a flame with physically detached fronts – or in unison in a flame with merged fronts. The results of the model are in qualitative agreement with previous experimental tests performed on aluminum–methane–oxidizer mixtures.

Nonadiabatic Flamelet Formulation for Predicting Wall Heat Transfer in Rocket Engines

A flamelet-based combustion model is proposed for the prediction of wall heat transfer in rocket engines and confined combustion systems. To account for convective heat loss due to the interaction of the flame with the wall, a permeable thermal boundary condition is introduced in the counterflow diffusion flame configuration. The solution of the resulting nonadiabatic flame structure forms a three-dimensional manifold, which is parameterized in terms of mixture fraction, progress variable, and temperature. The performance of the model is first evaluated through a direct numerical simulation analysis of an diffusion flame that is stabilized at an inert isothermal wall. The developed nonadiabatic flamelet model is shown to accurately predict the temperature, chemical composition, and wall heat transfer. Combined with a presumed probability density function closure, the model is then applied to large-eddy simulation of a single-injector rocket combustor to examine effects of heat transfer on the turbulent flame structure in rocket engines.

On the route to extinction in non-adiabatic flames from competitive exothermic reactions

We consider non-adiabatic combustion waves arising from two-step competitive exothermic reaction schemes. A numerical method is employed to study the behaviour of this system and we show that the inclusion of heat loss can lead to a period-doubling route to the termination of the propagating flame front. The nature of oscillations becomes more complex with increasing loss of heat until the system can no longer sustain a propagating front. In other words, beyond some critical value of heat loss, extinction of the combustion reaction would occur. For the non-adiabatic case, particularly close to the extinction threshold, large excursions in temperature and wave speed above those observed for the adiabatic case can occur. Such behaviour close to extinction may have implications for safety or industrial processes.

On the Stability of Planar Premixed Flames Under Nonadiabatic Conditions and Preferential Diffusion

In this paper, we provide a numerical study of the stability analysis of a planar premixed flame. The interaction of preferential diffusion and heat loss for a planar premixed flame is investigated using a thermodiffusive (constant density) model. The flame is studied as a function of three nondimensional parameters, namely, Damköhler number (ratio of diffusion time to chemical time), Lewis number (ratio of thermal to species diffusivity), and heat loss. A maximum of four solutions are identified in some cases, two of which are stable. The behavior of the eigenvalues of the linearized system of stabilty is also discussed. For low Lewis number, the heat loss plays a major role in stabilizing the flame for some moderately high values of Damköhler number. The results show the effect of increasing or decreasing Lewis number on adiabatic and nonadiabatic flames temperature and reaction rate as well as the range of heat loss at which flames can survive.

Effect of high Lewis number on the non-adiabatic premixed flame edges

In this paper, we investigate numerically the influence of heat loss and Damkohler number, which is inversely proportional to the square of strain rate, on the edges of premixed flames in a two dimensional counterflow for Lewis numbers greater than unity. The propagation speed of flame edges is plotted against Damkohler numbers for different values of the heat loss parameter showing a multiple flame edge solutions in the form of S shaped curves.

Cellular instabilities of non-adiabatic laminar flat methane/hydrogen oxy-fuel flames highly diluted with CO2

Cellular instability of non-adiabatic laminar premixed flat methane/hydrogen oxy-fuel flames highly diluted with CO2 was studied systematically. The CH4/H2/CO2/O2 flames were stabilized on a McKenna burner at atmospheric pressure. The CH4/H2/CO2/O2 flames with varied equivalence ratios and hydrogen mole fractions were investigated. Quantitative structure parameters of the cellular flame fronts were obtained with OH-PLIF technique. On the basis of experimental results, theoretical analysis was made to clarify the effect of mixture properties and heat loss on flame cellular instability. Results showed that cellular flame structure appeared near the lean limit combustion condition. Cell number of flame and its hydrodynamic factor, ωDL, present a positive correlation. The average stand-off distance, L, is well identified to characterize the cellular flames. Average cell amplitude, A, and wavelength, λ, increased with the increase of L. A good indicator of instability level, A·λ, was defined. The RS model was used in a linear stability analysis, and the onset of cellular flames was found to be governed by both hydrodynamic and thermal-diffusive instability mechanisms, while the heat loss made the flame more stable. The RS model is not an appropriate model to predict critical wavenumber for the non-adiabatic flame.

Global characteristics of non-premixed jet flames of hydrogen–hydrocarbon blended fuels

Blending hydrogen into hydrocarbon fuels can reduce the carbon-intensity of the fuel and extend the lean flammability limit. However, limited information is available of the global performance of attached, non-piloted hydrogen–hydrocarbon jet flames under well-defined boundary conditions. Three groups of blended fuels were investigated in the current study: Natural gas +H2 (with H2 volume fraction varying from 18.6% to 100%), C2H4+H2 (with H2 volume fraction varying from 0% to 100%), and 40% C2H4+41% H2+19% N2. Measurements were performed of flame dimensions, radiant fraction, and emission indices of NOx and CO. For flames with constant exit strain rate, the increase of hydrogen volume fraction was found to decrease the radiant fraction, decrease the global residence time and increase the NOx emission index. For flames of the same fuel composition, a higher strain rate results in a lower radiant fraction. The NOx production rate scales with the reciprocal of non-adiabatic flame temperature, consistent with the thermal NOx mechanism. The CO/CO2 ratio is determined by the competing influences of flame residence time, carbon input rate and mixing rate of the fuel and air.

Differential diffusion effects inclusion with flamelet generated manifold for the modeling of stratified premixed cooled flames

CFD predictions of flame position, stability and emissions are essential in order to obtain optimized combustor designs in a cost efficient way. However, the numerical modeling of practical combustion systems is a very challenging task. As a matter of fact, the use of detailed reaction mechanisms is necessary for reliable predictions, especially for highly diffusive fuels. Unfortunately, the modeling of the full detail of practical combustion equipment is currently prohibited by the limitations in computing power, given the large number of species and reactions involved. The Flamelet Generated Manifold (FGM) method reduces these computational costs by several orders of magnitude without loosing too much accuracy. Hereby, FGM enables the application of reliable chemistry mechanisms in CFD simulations of combustion processes. In the FGM technique the progress of the flame is generally described by a few control variables. For each control variable a transport equation is solved during run-time. The flamelet system is computed in a pre-processing stage, and a manifold with all the information about combustion is stored in a tabulated form. In the present paper, the FGM model is implemented for the analysis of partially premixed non-adiabatic flames, including the effects of differential diffusion. Subsequently, a computational analysis of partially premixed non-adiabatic flames is presented. In this scope, a series of test simulations is performed using FGM for a two dimensional geometry, characterized by a distinctive stratified methane/air inlet, and compared with detailed chemistry simulations. The results indicate that detailed simulations are well reproduced with the FGM technique.

Considering of the Internal Heat Losses and Continues Reaction Rate Effect in Aluminum Dust

In this study, a mathematical model for investigation the effects of internal heat losses on the flame speed and temperature profile in Aluminum Particle Clouds combustion with continues reaction rate have been studied. The present study extended previous results by bridging the theories of the non-adiabatic stationary dust flame and the propagation of premixed flames in one-dimensional channels accounting for heat-losses to particles and environment. The results showed that the effects of heat losses played an important role in flame regimes and flame transition. Furthermore, it was found that convective heat losses significantly decreased the velocity of flame propagation and temperature in post-flame zone. Comparisons between the analytical solutions and the experiment results showed a good agreement.