Laminar Diffusion Flames Research Articles

Experimental and numerical study of polycyclic aromatic hydrocarbon formation in ethylene laminar co-flow diffusion flames

Recent literature kinetic studies revealed the importance of new mechanisms for polycyclic aromatic hydrocarbon (PAH) and soot inception beyond hydrogen–abstraction–acetylene–addition (HACA) and hydrogen–abstraction–vinylacetylene–addition (HAVA) mechanisms in the combustion of ethylene and other hydrocarbons. Co-flow diffusion flame is a canonical flame used to investigate the interaction between fluid dynamics and PAH chemistry. In this study, supersonic molecular beam sampling technique was utilized for the first time with synchrotron vacuum ultraviolet photoionization mass spectrometry (SVUV-PIMS) to measure laminar co-flow diffusion flame at atmospheric pressure. We report quantitative measurement of precursor radicals as well as critical intermediates and odd carbon number PAH species. A custom-designed computational code, based on OpenFOAM and Cantera, was adopted to simulate laminar co-flow diffusion flames with literature kinetic model. Chemical kinetic analyses show that addition reactions of odd carbon number species provide considerable contribution to PAH formation processes beside HACA and HAVA mechanisms. Reasonable mass growth reactions are postulated for aromatic species with odd carbon numbers, such as ethynyl-indene, fluorene, benzo-indene, which need further investigations. Reactions of resonantly stabilized radicals followed by ring expansion are shown to be critical for both odd and even carbon number aromatics, and are suggested to be included in future PAH models.

Effects of CO/H2/N2 addition on the soot morphology and nanostructure in laminar co-flow ethylene diffusion flame

This study investigated the effect of CO/H2/N2 addition on the morphological evolutions and nanostructures of soot generated from a laminar co-flow ethylene diffusion flame through utilizing the method of thermophoretic sampling and transmission electron microscopy (TEM). The additions were introduced into the inner tube along with the fuel stream, and the volume fractions of the additions are 10%, 20% and 30%. Specifically, the parameters of soot production, primary particle diameter (Dp), fractal dimension (Df), fractal pre-factor (kg) and the nanostructures were determined and analyzed. A reduction in soot production was observed with increasing proportion of the additives, demonstrating the addition of these diluent gases could inhibit the soot formation. Among the three additives, N2 is the most effective in reducing soot yield. The size analysis of soot particles and aggregates indicated the addition of CO or N2 could diminish the primary particle diameter and fractal parameters. The addition of H2 could reduce the primary particle diameter at the bottom flame while yields greater primary particles at 40 mm HAB. This result demonstrated that H2 addition may weaken the oxidation occurring on soot particles. High resolution transmission electron microscopy (HRTEM) analysis was also developed and the results showed that the addition of CO or H2 results in greater fringe length, smaller fringe tortuosity and inter-fringe spacing for soot particles, demonstrating the more orderly and compact lattice structure. In particular, the nanostructure parameters of soot particles in H2 enriched flame vary more significantly than those in CO enriched flame, demonstrating the soot nanostructure is more graphitized in H2 enriched flame. Different from the case of CO or H2 addition, the lattice characteristics for soot in the flames with N2 addition present inverse variation tendency.

An experimental and modeling study on sooting characteristics of laminar counterflow diffusion flames with partial premixing

The effects of partial premixing on soot formation were investigated both experimentally and numerically in counterflow diffusion flames (CDFs) of ethylene, propane and ethylene/propane binary mixtures. The partial premixing ratio was maintained at relatively low levels so as to minimize the hydrodynamic effects; while the chemical effects of partial premixing on soot formation can be highlighted. The experimental results and simulations consistently showed that partial premixing notably increased soot formation in ethylene flames; while it slightly reduced the soot formation in propane flames, suggesting that partial premixing played different chemical roles in the soot formation process of ethylene and propane flames. Kinetic analyses revealed that partial premixing showed opposing effects on propargyl formation of ethylene and propane flames; this in turn led to the opposing effects on the formation of benzene and polycyclic aromatic hydrocarbons (PAH), and finally on the soot inception process. For partial premixing in propane/ethylene mixtures, the observed synergistic effects with propane added to ethylene CDFs disappeared, indicating a chemical cross-linking effect between partial premixing and ethylene/propane mixture. In addition, the present soot model showed satisfactory performance in reproducing the mismatch between PAH and soot formation tendencies in ethylene and propane CDFs. This result implies that great care is needed when using PAH concentration as an indicator of sooting tendency of different fuels. The present modeling results provided new insights into the chemical role of partial premixing in the soot evolution process of ethylene, propane and their blends. Further investigations towards the partial premixing effects on soot formation of larger liquid hydrocarbons such as diesel surrogates are required.

Single-phase instability of intermediate flamelet states in high-pressure combustion

The single-phase instability of high-pressure, steady, laminar counterflow diffusion flame is studied using the Vapor–Liquid Equilibrium (VLE) theory. The a posteriori study focuses on the identification of the potentially unstable regions emerging throughout the full combustion states represented by the high-pressure counterflow diffusion flame solution; a specific emphasis is placed on the middle branch solutions which are characteristic of intermediary combustion states. The a posteriori analysis provides useful bounds on the multicomponent phase separation. We use a mixture-fraction space flamelet formulation with real-fluid thermodynamics; the results compare favorably to both a spatial-based flamelet and a two-dimensional direct numerical simulations (DNS) solution, at high-pressure conditions. The a posteriori, single-phase instability for a high-pressure LO2/GH2 flame is investigated by investigating the effects of operating pressure and inlet temperature on the flame structure; then the analysis is extended to a LO2/GCH4 flame to evaluate the fuel effect. It is found that a single-phase instability can appear at both the fuel and oxidizer inlets, and its location and extent is determined by the water vapor concentration and, more importantly, temperature. For the flamelets in the upper burning branch, the size and location of unstable region is nearly invariant to the scalar dissipation rate, while for the flamelets in the middle branch, the instability region near the fuel inlet expands significantly with the decrease of chemical reactivity.

Effects of adding cyclohexane, n-hexane, ethanol, and 2,5-dimethylfuran to fuel on soot formation in laminar coflow n-heptane/iso-octane diffusion flame

The molecular structure strongly affects the fuel sooting propensity. This study aims to investigate the effects of oxygen-free and oxygen-containing aliphatic and aromatic hydrocarbons on soot formation. Experiments were carried out in laminar coflow diffusion flames of a n -heptane and iso-octane mixture doped sequentially with cyclohexane, n -hexane, ethanol, and 2,5-dimethylfuran by 15%. Soot samples were collected for characterization using field emission scanning electron microscope (FESEM), transmission electron microscope (TEM), X-Ray diffraction (XRD), Raman spectroscopy, and thermal gravimetric analysis. Peculiar structures consisting of ultrafine particles were observed in the deposit collected in all the flames except the one doped with ethanol. According to the surface morphology of soot samples, a model of the peculiar structure of soot deposit was proposed: oxidation free radical attack-deposit and edge defect carbon oxidation model. Based on the results of TEM image, XRD pattern, Raman spectroscopy signal, and thermogravimetric analysis, the reasons for the formation of this peculiar structure were analyzed. The results show that the aged soot particles formed in the flame doped with 15% ethanol (E15) have smaller interlayer spacing, larger polycyclic aromatic hydrocarbon stacking thickness ( L c ), smaller grain width ( L a ), smaller crystal stacking layer ( N ), smaller size, and a higher apparent activation energy. These characteristics were consistent with the fairly flat surface of the E15 soot deposits. Soot produced in other flames, especially doped with cycloalkanes, had a large grain width and crystal stacking layers, a lower apparent activation energy, higher oxidation reactivity, and displayed large gaps on the surface and was easily attacked by oxidizing free radicals. Moreover, the large gaps between the crystal layers of the shell structure of single soot particles, and the existence of edge polycarbonate made the crystal edge easy to be activated to produce defects. These soot had high oxidation activity and were prone to secondary combustion, leading to smaller soot particles and forming the peculiar structure on the deposit surface.

Sooting characteristics of ethanol-ethylene blends in laminar coflow diffusion flames up to 10 bar

Ethanol, being a renewable product, has been perceived for a long time as a fuel that burns clean and more complete than conventional hydrocarbons as well as being thought as a soot suppressant in blends with gasoline and diesel fuel. Some of the results from a wide-range of studies investigating soot-suppression potential of ethanol at atmospheric pressure, including Smoke Point measurements, have yielded occasionally contradictory conclusions. Further, scaling of soot studies conducted under atmospheric conditions to engine-relevant pressures has always been problematic because of the non-monotonic response of combustion events to pressure. Ethanol’s soot inhibiting characteristics at elevated pressures under tractable conditions have not been previously documented and may not be inferred easily from atmospheric pressure measurements. Information from benchmark investigations with ethanol-hydrocarbon blends at elevated pressures could help in understanding the soot aerosol emissions from practical engines fueled with ethanol blends. Soot production in ethanol/ethylene laminar diffusion flames at elevated pressures was investigated on a 3 mm diameter coflow burner mounted in a high-pressure combustion vessel. Ethanol was added to ethylene, which is diluted with nitrogen, in various quantities from 0% to 40% in terms of total carbon of the fuel stream keeping the carbon flow rate constant at 0.41 mg/s. Experiments were conducted at pressures from 3 to 10 bar, and in addition to keeping the mass flow of carbon constant, the ratio of mass flow of nitrogen to mass flow of carbon was also kept fixed at 6 at the fuel nozzle. The soot spectral emission spectroscopy technique was used to obtain line-of-sight intensities of soot radiation which were then converted to radial soot temperature and volume fraction distributions at a given axial height within the flame. Soot production increased by a small but measurable amount with the fuel mixture containing 10% carbon from ethanol at 5 and 8 bar pressure flames indicating synergistic interactions; while at 3 and 10 bar, no measurable change was observed. When the ethanol’s carbon contribution was increased to 30% and 40%, soot formation was suppressed and the soot yields decreased below the soot level of neat nitrogen-diluted ethylene at 3 to 10 bar pressures. Sensitivity of soot production to pressure in ethylene/ethanol flames is found to be similar to that of ethylene up to 8 bar.

Coupling of the Chemical Effects of CO and CO2 on Soot Formation in C2H4 Diffusion Flame in O2-CO2 Atmosphere

Previous studies found that the chemical effect of CO (CE-CO) enhances the formation of soot in laminar coflow C2H4/air diffusion flames, while the chemical effect of CO2 (CE-CO2) inhibits soot formation. The coupled CE-CO and CE-CO2 on soot formation characteristics is investigated numerically in O2-CO2 atmosphere through blending fuel flow rate with 10%, 30% and 50% CO and virtual CO (FCO) in a laminar C2H4 diffusion flame. The results show that CE-CO increases the flame temperature, while CE-CO2 is opposite; the CE-CO promotes soot formation attributed to soot oxidation process which decreases the mole fraction of OH by OH+CO↔H+CO2, while the CE-CO2 inhibits the formation of soot ascribed to the increase of OH and O2 concentration promoting soot oxidation rate and the decrease of flame temperature, mole fractions of C2H2 and H inhibiting surface growth rate. Moreover, the chemical promotion effect of CO decreases in O2-CO2 atmosphere, and the coupled CE-CO and CE-CO2 suppresses soot formation.

Investigation of PAH and soot formation in a dimethyl ether (DME) laminar coflow diffusion flame

Dimethyl ether (DME) is an alternative fuel with significantly reduced particulate (i.e., soot) emissions. The polycyclic aromatic hydrocarbons (PAHs) and soot formation in pure DME flames have not been fully investigated and there is a need to develop a detailed understanding of soot formation from pure DME without the confounding effect of other more sooting hydrocarbons. Therefore, PAH and soot formation in a pure DME /air laminar coflow diffusion flame were explored in this study for two different initial temperatures of the fuel. Gas chromatography /mass spectroscopy (GC/MS), and time-resolved laser-induced incandescence (TiRe-LII) were used to quantitatively measure PAH concentrations, soot volume fractions and primary particle diameters. The experimental results show that, before the appearance of the soot particles, the gas phase reactions are dominated by PAHs smaller than three rings (A3: phenanthrene and anthracene), which decrease along the centerline as the soot volume fractions increase. The primary particle diameters remain relatively constant along the radial profiles at different flame heights, which suggests a moderate surface growth. The numerical predictions of soot and PAHs (smaller than A3) formation are consistent with the measurements (within a factor of 2.4), while the overprediction of the mole fraction of A3 is near an order of magnitude. The model can also capture the effect of increasing the initial fuel temperatures with reasonable accuracy. The experimental results suggest that PAHs are mainly formed by Hydrogen Abstraction Carbon (Acetylene) Addition (HACA), C1 addition is efficient for monoaromatic while the addition of C3 and species containing cyclopentadienyl moiety are not dominant for PAH growth. A database of the PAHs and soot formation has been created for further investigation of this DME flame.

Correction of laser-induced incandescence signal trapping in soot measurement in a microgravity boundary layer laminar diffusion flame

This work presents a novel method to auto-correct the trapping effect of laser-induced incandescence (LII) signals inside the flame to obtain soot volume fraction distributions in laminar axisymmetric (2D) and three-dimensional (3D) flames using planar LII. The development of the proposed model is described with an estimation of the propagated uncertainties. The proposed model is firstly numerically validated in a canonical 2D coflow Santoro flame using synthetic LII signals and soot properties predicted by the CoFlame code. Secondly, experimental measurements of LII were carried out in the same 2D coflow Santoro flame and the LII signals were converted to soot volume fraction using the proposed model. Simultaneous LII and line-of-sight attenuation measurements were conducted to obtain the calibration factor. The proposed model automatically took into account the signal trapping effect in the conversion of the detected LII signal to soot volume fraction. Finally, the validated methodology is applied to a 3D laminar boundary layer diffusion flame established in microgravity. The significant differences in the measured soot volume fraction distributions with and without considering signal trapping, particularly in the 3D zone of the flame, demonstrate the importance of considering signal trapping to LII measurements of soot volume fraction in this flame. The model developed in this work can be readily applied in planar LII measurements of soot in any flame configuration as long as it is steady or statistically steady, to allow measurements to be performed at different positions inside a flame, such as for pool-fires or pulsating flames.

Retrieving soot volume fraction fields for laminar axisymmetric diffusion flames using convolutional neural networks

Typical procedures for estimating soot volume fraction distribution in laboratory flames require solving ill-posed inverse problems to recover the fields from convoluted signals that integrate light extinction from soot particles along the line-of-sight of a photo-detector. Classical deconvolution methods are highly sensitive to noise and the choice of tunable regularization parameters, which prevents obtaining consistent estimations even for the same reference flame settings.This paper presents a novel approach based on Convolutional Neural Networks (CNNs) for estimating the soot volume fraction fields from 2D images of line-of-sight attenuation (LOSA) measurements in coflow laminar axisymmetric diffusion flames. Using a set of reference synthetic soot volume fraction fields of canonical flames and their corresponding projected LOSA images, we trained a CNN for reconstructing soot fields from images representing the data captured by a camera. Experimental results show that the proposed CNN approach outperforms classical deconvolution methods when reconstructing the flame spatial soot distribution from noisy images of LOSA.

On the relative contributions of soot to radiative heat transfer at different oxygen indices in ethylene – O2/CO2 laminar diffusion flames

Soot concentration measurements in ethylene-O2/CO2 diffusion flames (fuel Re 120) encompassing oxidizer oxygen mole fractions (oxygen index, OI) of 35% to 90% were employed to validate a methodology for modeling soot production. A skeletal 10-step reaction mechanism and modeling the soot nucleation parameter (Cα) in the Moss-Brookes soot model as a function of OI provided adequate agreement with measurements. Cross-validation of the OI based Cα model was ascertained through ethylene-O2/CO2 diffusion flame simulations in a different experimental configuration (fuel Re 43, 35%-50% O2) by ensuring that peak soot concentrations were in good agreement with previously published results. A mesh resolution of 0.0625 mm within the flame was deemed necessary to obtain grid independent results. Despite increases in peak soot volume fractions and flame temperatures with OI, the radiant fraction (χR) (in-flame) decreases due to decreasing flame lengths. Although in-flame radiation is dominated by gases, the spectrally-continuous soot radiation plays an important role on χR (when it is assessed across the entire domain) due to the attenuation of flame radiation by the cold, radiatively-participating, co-flow. Consequently, in-flame radiation show a strong sensitivity to gas-radiative property models whereas χR when assessed across the entire domain is impacted by the choice of soot radiative property model.

Transient analysis of the nonmonotonic response of counterflow laminar diffusion flames in alternating current and step electric fields using a one-dimensional ionic transport model.

The electrohydrodynamic response of a counterflow laminar diffusion flame to applied alternating current (ac) electric fields is investigated experimentally and numerically. The flame positions are observed to show typical response to applied ac electric fields with high and moderate frequencies. The flame position does not respond above a threshold frequency corresponding to a certain collision response time, below which it oscillates in phase with the applied electric field. At a very low frequency (less than approximately 1 Hz), however, the flame position is observed to vary nonmonotonically as a function of time. To elucidate the nonmonotonic behaviors, a one-dimensional ionic transport model was employed by applying time-dependent electric fields. The responses of flame positions for ionized layers substituting for counterflow diffusion flames were systematically investigated with respect to one-way ionic wind (OIW) and two-way ionic wind (TIW) models. Consequently, it is demonstrated that the ionic models can produce not only harmonic flame oscillations for relatively low ac frequencies, but also free flame oscillations for stepwise voltages, which originated from the interaction between electrostatic force and ionic wind-induced force in the ionic system for both the OIW and TIW models.

The unsteady response of radiating laminar diffusion flames exposed to decreasing mixing rate conditions representative of fires

The objective of this study is to bring fundamental information on the coupling between combustion and thermal radiation occurring at flame scale. The study considers a simplified configuration corresponding to one-dimensional counterflow planar laminar diffusion flames subjected to time-evolving moderate-to-slow mixing conditions that are representative of fires. The flame simulations are performed assuming methane–air combustion and using two formulations: a simplified formulation featuring single-step chemistry and an optically thin flame radiation model; and a high-fidelity formulation featuring detailed chemistry and different radiation models based on solving the Radiative Transfer Equation combined with grey or non-grey descriptions of gas radiation properties – the non-grey description is based on a Weighted-Sum-of-Gray-Gases model. The simulations describe the gradual decrease in heat release rate resulting from the evolution towards slower mixing conditions, as well as the corresponding increase in the flame optical depth and decrease in peak flame temperature. Radiation extinction is observed at critically low values of the rate of mixing. The analysis demonstrates that for conditions before the extinction limit is reached, the flame belongs to the semi-unsteady regime in which mixing processes occurring in the outer diffusive layers of the flame are unsteady whereas heat release processes occurring in the inner reactive layer remain quasi-steady. In the semi-unsteady regime, the flame structure can be parametrised by an effective strain rate (or scalar dissipation rate) and remains close to the structure of radiating flames considered at steady state and subjected to the effective value of the strain rate.

Soot aggregate morphology deduced from thermophoretic sampling in coflow laminar methane diffusion flames at pressures up to 30 bar

A high pressure combustion chamber, housing a co-flow burner and a novel thermophoretic soot sampling system, was used to collect soot samples from methane-air diffusion flames at pressures from 5 bar to 30 bar. To have tractable diffusion flames with the same residence times, constant methane flow rate of 0.55 mg/s was maintained throughout the experiments. Soot aggregates were collected on 3 mm diameter transmission electron microscope (TEM) grids and the images of the soot aggregates were captured by a TEM device. Sampling was done at three heights above the burner rim at 3, 6, and 8 mm. An automated image processing method was used to analyze the TEM images of soot aggregates to infer the primary soot particle diameters and the aggregate morphology. The mean primary soot particle diameters increased from 5 to 15 bar at all three sampling locations and seemed to reach a peak around 15–20 bar at 6 and 8 mm height and start decreasing with further increase in pressure. On the other hand, mean primary soot diameter at 3 mm height increased slightly beyond 15 bar. In view of the measured soot concentrations and temperatures at the sampling locations at 3, 6, and 8 mm above the burner, the soot number density is argued to be increasing with increasing pressure. Throughout the all experimental conditions, fractal dimension of the aggregates in these methane diffusion flames did not exceed 1.4 which is much smaller than the accepted universal value of about 1.85. Morphological characteristics inferred from TEM images indicate that the aggregation is less prominent in methane flames in comparison to ethylene diffusion flames under similar conditions.

Pressure effects on the soot production and radiative heat transfer of non-buoyant laminar diffusion flames spreading in opposed flow over insulated wires

Abstract This paper investigates experimentally and numerically pressure effects on soot production and radiative heat transfer in non-buoyant opposed-flow flames spreading over wires coated by Low Density PolyEthylene (LDPE). Experiments, conducted in parabolic flights, consider pressure levels ranging from 50.7 kPa to 121.6 kPa and an oxidizer flowing parallel to the wire's axis at a velocity of 150 mm/s and composed of 20% O2/80% N2 in volume. The numerical model includes a detailed chemistry, a two-equation smoke-point based soot production model, a radiation model coupling the Full-Spectrum correlated-k method with the finite volume method and a simple degradation model for LDPE. An analysis of the experimental data shows that the spread rate, the pyrolysis mass flow rate, and the residence time for soot formation are independent of pressure whereas the soot formation rate is third-order in pressure. The model reproduces quantitatively the effects of pressure on soot production and captures the transition from non-smoking to smoking flames. The radiant fraction increases with pressure because of an enhancement in soot radiation whereas the contribution of radiating gases remains approximately constant over the range of pressures considered. In addition, gas radiation dominates at pressure lower than 75 kPa whereas soot radiation prevails at higher-pressure levels. Consistently with the data obtained at normal gravity, the smoke-point transition is found to occur for a radiant fraction of about 0.3 and the soot oxidation freezing temperature is estimated in the range 1350–1450 K. Eventually, whatever the pressure considered, the surface re-radiation from the wire is higher than the incident radiative flux from the flame to the surface along the entire wire. This shows that radiative heat transfer contributes negatively to the heating of the unburnt LDPE and to the heat balance along the pyrolysing surface.

Combustion of Iron Particles Suspension in Laminar Premixed and Diffusion Flames

ABSTRACT In the work, the combustion regimes of iron particles (d10 = 4.7 µm) in a laminar pre-mixed flame (LPF) and in an axisymmetric laminar diffusion flame (LDF) are studied. For (LPF), we studied the effect of variations in gas suspension of the concentrations of fuel Cf = 0.4 ÷ 0.6 kg/m3 and oxygen CO2 = 21 ÷ 40%, and for (LDF), pure oxygen blowing and variation in the iron concentration Cf = 0.4 ÷ 1.5 kg/m3. The results of particle temperature measurements in the combustion zone of LPF and LDF, the dispersed and phase composition of the combustion products (depending on the concentration of fuel and oxidizer) are presented. It is shown that, under certain conditions, iron particles in LPF switch to a “high-temperature” diffusion mode of combustion. This mode is accompanied by intensification of gas-phase reactions and leads to a two-mode particle size distribution of combustion products. In this mode of combustion of a gas suspension (at temperatures much lower than the boiling point of iron), the mass yield of a nanoscale fraction of combustion products reaches tens of percent and increases with increasing oxygen concentration in LPF. This mode of iron particles combustion was not observed in LDF. The features of the thermal structure of the diffusion dust flame are analyzed. It is shown that in LDF the burning time of metal particles in the diffusion mode is proportional to the particle radius (in the kinetic mode), and should depend on the concentration of fuel and the size of the burner.

Experimental and Numerical Investigation into the Effects of Unsaturated Carbon Bonds of Hydrocarbon Fuels on Soot Formation in Laminar Diffusion Flames

ABSTRACT Fuel structure is the main factor controlling soot behavior. Compared to the carbon number, the unsaturated carbon bonds have a greater effect on soot formation. This work analyzed the mechanism of C = C and C ≡ C on soot by adding ethylene and acetylene to the fuel stream of methane laminar diffusion flame experimentally and numerically. The results showed that the blending of C2H4 results in an increase of flame height without causing smoking, while the C2H2 addition up to 40% results in soot wings. The effect of C ≡ C bond on the flame structure and natural luminosity is much higher than that of C = C bond. The peak soot volume fraction (SVF) increases with the increasing blending ratio of C2H4 and C2H2, but the latter is one order of magnitude higher. The promotion effect of C ≡ C on soot growth attributes to the synergistic increase of HACA rate and PAHs condensation rate, while that of C = C is mainly due to the increased HACA rate. With the increase of unsaturated carbon bond concentration, the primary particle number density increases, the soot inception process is advanced. This trend is further enhanced with the increase of carbon bond unsaturation due to the shortening of the precursor formation pathway and the acceleration of reaction rate. With the increase of C2H4 blending ratio, the peak O2 oxidation rate increases linearly, while with the increase of C2H2 ratio, peak O2 oxidation rate increase first and then decrease, peaks at 40% ratio. Compared with O2 oxidation, the OH oxidation rate is not sensitive to carbon bond unsaturation and unsaturated carbon bond concentration.

An Artificial Neural Network for the Low-Cost Prediction of Soot Emissions

Soot formation in combustion systems is a growing concern due to its adverse environmental and health effects. It is considered to be a tremendously complicated phenomenon which includes multiphase flow, thermodynamics, heat transfer, chemical kinetics, and particle dynamics. Although various numerical approaches have been developed for the detailed modeling of soot evolution, most industrial device simulations neglect or rudimentarily approximate soot formation due to its high computational cost. Developing accurate, easy to use, and computationally inexpensive numerical techniques to predict or estimate soot concentrations is a major objective of the combustion industry. In the present study, a supervised Artificial Neural Network (ANN) technique is applied to predict the soot concentration fields in ethylene/air laminar diffusion flames accurately with a low computational cost. To gather validated data, eight different flames with various equivalence ratios, inlet velocities, and burner geometries are modeled using the CoFlame code (a computational fluid dynamics (CFD) parallel combustion and soot model) and the Lagrangian histories of soot-containing fluid parcels are computed and stored. Then, an ANN model is developed and optimized using the Levenberg-Marquardt approach. Two different scenarios are introduced to validate the network performance; testing the prediction capabilities of the network for the same eight flames that are used to train the network, and for two new flames that are not within the training data set. It is shown that for both of these cases the ANN is able to predict the overall soot concentration field very well with a relatively low integrated error.

Soot Formation With Light Extinction and Grayscale Extraction Methods Applied to Ethanol-Gasoline Blends Laminar Flame

AbstractIn this paper, the two-dimensional parallel light extinction method was carried out to study the soot formation in laminar diffusion flames of four different ethanol-gasoline blends, of which ethanol volume fractions ranging from 0% up to 100% (E0, E20, E80, and E100). The flame images were processed synthetically via matlab to accurately calculate the flame height. In addition, the flame structure was redefined as three zones to observe the soot formation. The results indicate that the flame height changes with the variation of gas volume flowrate and fuel mass flowrate during the experiment. In terms of soot formation, as the volume fraction of ethanol increases, the proportion of soot forming zone decreases, while the area of blue flame zone grows. Simultaneously, the transition zone accounts for about 21% of the total flame area, which has no significant change with the increase of ethanol volume fraction.

A decomposition method for the simultaneous reconstruction of temperature and soot volume fraction distributions in axisymmetric flames

This paper presents a method for the simultaneous reconstruction of temperature and soot volume fraction distributions in axisymmetric flames. In the proposed method, the radiation temperature and emissivity are determined using a multi-wavelength method. Then, based on the decomposition of the emissivity, the soot volume fraction distribution is reconstructed. Accordingly, the temperature distribution is calculated using the reconstructed local radiation source terms under a single wavelength. Based on a correction formula proposed in this paper, a further correction process is conducted, and the soot volume fraction and temperature distributions are successively obtained. Simulations are conducted with the temperature and soot volume fraction distributions at two different heights from the flame burner of an axisymmetric ethylene laminar diffusion flame. In the simulations, normal random errors are superimposed, and the radiation temperature and emissivity are calculated based on the intensities at three wavelengths corresponding to the R, G, and B channels of color cameras. Good agreement with the assumed data is obtained, which verifies the performance of the proposed method. Additionally, comparison with the results obtained using the widely-used two-color reconstruction method indicates that the proposed method exhibits a more stable performance with smaller errors. Moreover, the proposed method can take advantage of measurement at multiple wavelengths. Based on the multi-wavelength data captured using a hyperspectral imaging device, the measurement results of an axisymmetric ethylene laminar diffusion flame using the proposed method are also presented.