Search

International audience; This work presents an experimental study of combustion characteristics of micron-sized aluminum particles in the transition regime under constant volume combustion experiments. Burning velocities were estimated from the measured pressure traces using both a simplified model for aerosol combustion on closed spherical bombs and a semi-empirical correlation, and compared to previous literature. Flame temperatures were measured by bi-color pyrometry and indicate that for particles smaller than 12 m, the flame moves closer to the particle's surface, since flame temperatures were close to aluminum boiling point. For 17:9 µm particles, flame temperatures were close to predicted adiabatic flame temperature and alumina vaporization-dissociation temperature, indicating a classical vapor phase flame under a diffusion-controlled mechanism. However, spectroscopy measurements did not detect significant reductions on molecular AlO emissions for ner particles. This indicates a still very significant presence of vapor phase reaction for powders with a Sauter mean diameter at least as large as 7 µm, which is further supported by the presence of nanometric spheres in the combustion residues, since alumina formed under a vapor phase reaction is expected to condensate into nanometric droplets.

International audience; Due to its high lubricity, nitromethane is a fuel regularly used in model engine or more generally in race engine. The objective of this study is to improve our knowledge and understanding of the combustion of nitromethane for better evaluating its potential as fuel for automotive spark-ignition engines. To achieve this goal, unstretched laminar burning velocities of nitromethane-air mixtures were measured using spherical propagation methodology at 423 K over a pressure range from 0.5 to 3 bar and equivalence ratios from 0.5 to 1.3. The data indicated a typical adverse effect of pressure on laminar burning velocities. Based on the work done by Zhang et al., Proc. Combust. Inst., 33 (2011) 407-414, a modified detailed kinetic model including 88 species and 701 reactions was proposed. Comparisons between experimental and simulated un-stretched laminar flame speed were made and showed good agreement. The new kinetic mechanism was also used to successfully simulate published experiments and rationalize the unusual occurrence of maximum flame speed in the fuel-lean region.

The burning rate in the vicinity of a piston is estimated from a fractal analysis. The fractal parameters are determined from laser sheet tomography flame images for methane–air mixtures with three equivalence ratios (1, 0.9, 0.8) in a transparent spark-ignition engine. Two imaging configurations were used: five horizontal planes placed at different distances from the piston (0, 1, 2, 3, and 5 mm) and a vertical one passed through the center of the combustion chamber. The methodology proposed by Foucher et al. [F. Foucher, S. Burnel, C. Mounaim-Rousselle, Proc. Combust. Inst. 29 (2002) 751–757] allows the effect of cyclic variations to be avoided. The fractal formulation is modified to take into account the flame–piston distance and flame quenching. Far from the piston, evolution of the fractal dimension versus q ′ / S L 0 is found to be in good agreement with literature results. Near the piston, the fractal dimension evolves significantly when the distance is about twice the integral length scale and tends toward 2, the fractal dimension of a laminar flame front. The quenching ratio parameter Q R is introduced to consider the quenching of the flame by the piston. Finally, the burning rate is determined as a function of the distance between the wall and the mean flame contour and compared to a flame density approach, and similar results are found.

International audience; In Spark Ignition engines, the heat release rate is not only piloted by the mixture reactivity but also by its sensitivity to stretch effects. Only few results can be found in the literature about flame stretch effect in SI engine configurations.For this study, three different fuels (Methane, Propane, Iso-octane) were studied, but at different air-fuel lean mixture conditions, to present almost equivalent laminar flame speeds and thermo-dynamical properties at ignition timing condition. Besides those mixtures present different Lewis numbers which are relevant parameters to describe flame-stretch interactions. Mie-scattering tomography was then performed in an optical Spark Ignition (S.I.) engine. Using a high speed camera, flame propagation images were acquired through the piston. Thermodynamic analyses based on in-cylinder pressure traces were performed to estimate in-cylinder temperature and burnt mass fraction during the engine cycle.From the determination of flame areas, the global flame stretch and an equivalent propagation speed have been then defined and estimated. Results prove that mixtures have the same ranking in terms of flame stretch sensitivities as in the laminar regime. Probability density functions of flame curvature are centered on 0.05 mm−1 for the three mixtures whereas differences are observed on the global wrinkling. Finally the impact of the Lewis number on the Burned Mass Fraction curved was studied thus showing a linear relationship between crank angle corresponding to 5 and 10 % of burned mass fraction and the Lewis number.

Ethanol production is generally considered as the main option for converting lignocellulose to biofuels. However another route can produce valeric esters from lignocellulose through levulinic acid. While the process has been well characterized, only few experimental results describe their combustion behavior. Using two engine test benches, compression ignition (CI) and spark ignition (SI), we have investigated the engine performances and emissions of methyl, ethyl, butyl and pentyl valerate. According to their physicochemical properties, pure and blends of 20%vol methyl or ethyl ester with PRF95 were tested in the SI engine, while blends of 20%vol butyl or pentyl esters with gasoil were tested in the CI engine. In the SI experiments, we observed that the methyl and ethyl valerate have a higher flame speed than PRF95 which required a slight change of ignition timing to optimize the work output. However, both the performances and the emissions are not significantly modified. In the CI experiments, the blends including butyl or pentyl valerate performed equally well as the diesel. The lower cetane number of the esters slightly increase the ignition delay of the blends without affecting the performances. According to the results of this study, valeric biofuels represent very good alternative biofuels for both SI and CI engines.

International audience; The aim of this article is to determine the laminar burning speeds and Markstein lengths of p-cymene. This fuel is emitted by a typical vegetal species of the Mediterranean region often involved in forest fires. Experiments are performed in a spherical vessel for different equivalence ratios ranging from 0.7 to 1.4 at 180 °C and at atmospheric pressure. The effect of temperature (85 to 180 °C) at the stoichiometry is also studied. Three models (one linear and two nonlinear) are used in the extraction process. Depending on the equivalence ratio, the more accurate models are adopted to determine the laminar burning speeds and Markstein lengths of p-cymene. Results are compared favorably to experimental values of a similar fuel (α-pinene) and to numerical data of n-decane computed using the in-house code A-SURF.

The aviation fuel, jet propellant-8, was applied to premixed charge ignition combustion as well as conventional combustion in a single-cylinder diesel engine and was compared with diesel fuel. The engine performance and emissions were tested with and without exhaust gas recirculation under two operating conditions. The liquid- and vapor-phase penetration of diesel fuel and jet propellant-8 were also compared by Mie-scattering and Schlieren method under evaporating conditions using a constant-volume chamber. It was observed that jet propellant-8 exhibited slightly longer vapor-phase penetration and evidently shorter liquid-phase penetration compared with diesel fuel due to the lower distillation temperature and density of jet propellant-8. However, despite the faster evaporation rate of jet propellant-8, it was found that the ignition delay with jet propellant-8 was 2–3 crank angle degrees longer than that with diesel fuel due to its lower cetane number. This result was consistent for all operating conditions and combustion regimes. Jet propellant-8 also showed lower nitrogen oxides (NOx) and smoke emissions for both combustion regimes under the low-load condition because of locally leaner air–fuel mixture caused by longer ignition delay, higher volatility, and lower aromatic contents. As the engine load increased, jet propellant-8 emitted more NOx under conventional combustion regime due to the more vigorous premixed burn phase compared with diesel fuel. With exhaust gas recirculation, jet propellant-8 showed an improved trade-off relationship between NOx and smoke emissions for both combustion regimes due to its better evaporation characteristics and lower aromatic contents. Premixed charge ignition combustion with jet propellant-8 emitted lower smoke than conventional combustion with jet propellant-8 under near-zero NOx level.

A theoretical explanation is given of a technique based on Mie scattering interferometry (MSI), obtained by defocusing of the collecting optics, to size droplets. The originality of this study is the development of a droplet sizing method by planar laser light scattering for the case of a scattering angle range close to 90°. The feasibility of this method and its limitations are fully described. The dependence on intensity levels and refractive index variations can be neglected. After discussion of some practical details about particle size, imaging and camera constraints, the results obtained in the combustion chamber of a spark ignition (SI) engine, near the spark plug, prior to ignition and for different injection timings are described and discussed. It can be concluded that the implementation of the MSI method in this experimental set-up has been realized successfully to provide droplet distributions in an SI engine. To allow the easier use of the technique, image processing software will be developed in the Matlab environment.

Oxygen-enriched air combustion has been extensively studied, but not for internal-combustion engines. This paper presents how oxygen-enriched intake air influences early flame growth and the combustion process, and looks at how oxygen (O2) enrichment translates into resistance to carbon dioxide (CO2) dilution. The experiments were conducted with an optical single-cylinder engine instrumented with an in-cylinder pressure transducer and a high-speed camera to monitor early flame growth. During the experiments, O2 concentrations were varied from 18% to 29%, while CO2 was varied from 0% to 35%. The main findings are that O2 addition speeds up the combustion process and renders the flame more resistant to CO2 dilution. It was observed that the amount of CO2 can be significantly increased to reduce the combustion process to a level similar to that with air. Finally, laminar flame speed correlated well with early flame growth and fully developed turbulent combustion periods.

As combustion takes place in smaller combustion chambers at higher pressure, the development of boosted Spark-Ignition (SI) engines can induce non classical flame development. Moreover to limit abnormal combustion and thermal NOx production, a high dilution rate by exhaust gases is used. To improve the combustion processes occurring in SI engines and to validate new concepts, turbulent premixed combustion propagation is now estimated by using 3D CFD calculations, usually based on flamelet theory. However, this theory may be limited when the boosted engine is running in high diluted conditions. Therefore, in this paper, the flame characteristics of a stoichiometric iso-octane/air mixture were determined from in-cylinder measurements for a relatively high intake pressure and different dilution rates. By using an evaluation of both laminar and turbulent characteristics at real engine conditions, the trajectories of combustion processes occurring inside the cylinder were plotted in a Peters-Borghi diagram. Dilution induces a shift from the middle of the corrugated flamelet zone to the beginning of the thin reaction zones. Moreover, thanks to high speed PIV-flame tomography measurements, it was shown that dilution enhances flame-turbulence interactions and that the corrugation generated through dilution occurs at smaller scales than the integral length scales.

International audience

International audience; A potential implication of Volatile Organic Compounds (VOCs) emitted by vegetal species has been introduced in the literature to explain accelerating forest fires. These fires are characterized by the sudden increase of the rate of spread and of the energy released by the fire front. The main purpose of this paper is to determine the combustion characteristics of a major VOC emitted by Thymus vulgaris needles, namely p-cymene. The emission of this compound is studied for the temperature range 343-453 K with an emission peak observed at 443 K. Laminar burning speeds, Markstein lengths and flame thicknesses are determined using outwardly propagating spherical flames in a combustion vessel at atmospheric pressure. The effects of equivalence ratio (0.8 to 1.4) and unburned gas temperature (358 to 453 K) are investigated. Results are compared to experimental data of α-pinene and to computed data of two fuels, JP-10 and n-decane obtained with the PREMIX code of the CHEMKIN package.

International audience

Small concentrations of very reactive chemical species such as nitric oxide (NO) have a very important impact on the onset of combustion. In engine applications, this may lead to irregular combustion phenomena in spark ignition engines or contribute to the control of the homogeneous charge compression ignition (HCCI) engines. To numerically analyze and predict the effect of these species, detailed chemical mechanisms are required. However, including these mechanisms in computational fluid dynamics (CFD) simulations often results in prohibitive computational cost. Using a single-cylinder HCCI engine test bench, we have analysed the effect of initial NO concentrations ranging from 0 to 500 ppm on the ignition of iso-octane. We have also investigated whether the tabulation of dynamic adaptive chemistry (TDAC) method, that significantly reduces the CPU time associated with using detailed chemistry in CFD simulations, could capture this effect. This paper first presents the experimental setup and the validation data. Then it compares these experimental data with the CFD simulation results using a detailed mechanism involving more than 1000 species and 4500 reactions. The significant effect of NO on iso-octane ignition is efficiently captured over the whole range of NO concentration with a speed-up factor of up to 1500 (compared to simulations without TDAC). This work further demonstrates that TDAC represents a very efficient tool to include detailed kinetic mechanisms able to describe complex phenomena, such as the kinetic effect of small concentrations of reactive species, on ignition in a HCCI engine.

International audience

Nowadays, the main objectives in the automobile engine field are to reduce fuel consumption and pollutant emissions. HCCI engines can be a good solution to meet pollutant emission requirements and to achieve high combustion efficiency. However, before an HCCI engine is used as a conventional engine, several problems must be overcome, in particular control of the progression of combustion. Many studies have been conducted into possible control methods. A new strategy consists in using oxidizing chemical species such as ozone to seed the intake of a HCCI engine. As increasingly smaller ozonizers are now being designed, this kind of device could be integrated on a vehicle and on a HCCI engine, in order to control combustion phasing and promote the future use of this engine as a conventional engine. In the present study, experiments on a HCCI engine fuelled with iso-octane were carried out with ozone seeding in the intake. Results showed that when assisted by the addition of ozone, combustion can be enhanced and moved forward. Consequently, the use of oxidizing chemical species can be a means to control HCCI combustion. Depending on the inlet temperature, the control of combustion phasing may be more or less easy due to sensitivity to the ozone concentration. The present results also show the existence of a cool flame in the case of iso-octane combustion, indicating that ozone seeding can also be used in order to study iso-octane cool flame in a HCCI engine.

To improve our understanding of the combustion characteristics of ethyl pentanoate, a possible second generation biofuel, new experimental data were acquired for its oxidation kinetics in two complementary experiments. In a JSR (jet stirred reactor), concentration profiles of stable species were measured at 10 atm over a range of conditions (equivalence ratios of 0.6, 1, 2, fuel initial concentration of 1000 ppm, and temperatures between 560 to 1160 K). In a spherical combustion chamber, unstretched laminar burning velocities of ethyl pentanoate–air mixtures were measured at different pressures and temperatures and for equivalence ratios in the range 0.7–1.4. The oxidation of ethyl pentanoate was modeled using a new detailed kinetic reaction scheme (2719 reactions, 522 species). The chemical kinetic mechanism proposed here yielded good agreement with the present data. To interpret the results, reactions flux and sensitivity analyses were carried out.

Several researchers have reported that under certain conditions, forest fires with normal behavior suddenly start to propagate at unusual and very high rate of spread. Over the last decades, these accelerating forest fires were responsible for many fatalities in Europe. A thermochemical approach, based on the ignition of a volatile organic compounds (VOCs) cloud, has been proposed previously to explain this phenomenon. Indeed, some vegetal species emit volatile substances when they are heated. A typical Mediterranean plant, Rosmarinus officinalis, emits 14 components, mainly α-pinene and limonene. The acceleration of the rate of spread can be the consequence of the ignition of these emitted gases. The determination of α-pinene/air and limonene/air premixed flame speeds is essential to take into account this approach in forest fire modeling. It is the main purpose of this article. The spherical expanding flames methodology coupled with a nonlinear model was used to determine the unstretched premixed flame ...

International audience; Several researches have reported that under certain conditions forest fires with normal behavior suddenly start to propagate at unusual and very fast rate of spread. A thermochemical approach, based on the ignition of a Volatile Organic Compounds (VOCs) cloud, has been proposed previously to explain these accelerating forest fires. Indeed, some vegetal species when heated emit volatile substances. We have shown using a flash pyrolysis apparatus that a typical Mediterranean plant, Rosmarinus officinalis, emits eighteen components, mainly α-pinene. Laminar burning speeds and Markstein lengths as well as flame thicknesses of α-pinene/air premixed flames are determined using the spherical expanding flames method. Experiments are carried out in a spherical vessel at atmospheric pressure. The effects of equivalence ratio (0.7-1.4) and unburned gas temperature (353-453 K) are studied. Combustion characteristics are obtained using a nonlinear methodology. A correlation is developed to calculate the laminar burning speeds as a function of equivalence ratio and temperature. The experimental results are compared to the computed ones of JP-10 and n-decane as well as to those found in the literature for these compounds.

International audience; This work presents experimental tools for the determination of Minimum Ignition Energy (MIE) and results concerning the influence of an initial oxidation state on ignition threshold energies and flame velocity. These studies are carried out with micrometric aluminium particles which are oxidised using an anodising process. The first part of this work concerns the description of the experimental devices (Hartmann tube, for MIE measurements, and constant volume combustion chamber for flame velocity measurement with using high speed recording shadowgraphy). In the second part, a review of some results obtained for the sensitivity (MIE) of aluminium particle evolution versus particle diameter, air-fuel equivalence ratio and oxide content is presented. The effect of the oxide content is demonstrated: the MIE increases with the initial oxide content. The sensitivity of oxidised dust remains relatively high for high oxide contents (17.1wt%). The flame velocity is also modified and decreases as the oxide content increases. The most important result seems to be the role of the water content contained in the oxide shell which increases the reactivity of the oxidised aluminium dust.

International audience; This paper presents an experimental investigation of the influence of fuel density, viscosity, and surface tension on the diesel injection process (spray development and droplet size distribution) in non-vaporizing conditions. The spray development was captured in a pressurized vessel with a high-speed camera in order to obtain the penetration length and the spray angle. The droplet size distributions were measured by laser diffraction. A fuel matrix of eight fuels was designed including independent variations of fuel density, viscosity, and surface tension. Experimental conditions, such as injection pressure and air density, were varied according to current diesel engine conditions. The analysis focuses mainly on the quasi-stationary phase of the injection event. Results indicate that viscosity and to a lesser extent fuel density affect the spray characteristics. Increasing fuel density and viscosity by, respectively, 100 kg/m3 and 6.5 mm2/s induces longer spray tip penetration (up to 7%) with a narrower spray angle (drop of up to 3 deg) in quasi-stationary conditions. An increase in the Sauter mean diameter (up to 23% or 3 µm) was also observed. An increase in surface tension from 18 to 30 mN/m did not induce any significant modifications in the droplet size or in the spray development. Correlations, some of which are based on existing ones, are proposed to estimate the spray cone angle, the spray tip penetration, the SMD (Sauter mean diameter) and the drop size distribution considering the experimental conditions and the fuel properties.

To better understand the combustion characteristics of ethyl esters that could become second-generation biofuels, new experimental data were obtained for the combustion of a series of small ethyl esters. Unstretched laminar burning velocities of C4–C7 ethyl ester–air premixed laminar flames were measured in a spherical combustion chamber over a range of pressures (1–10 bar), initial temperatures (323–473 K), and equivalence ratios (0.7–1.5). These experiments were simulated as well as low-pressure flame structures, jet-stirred reactor species profiles, and ignition delays in shock tubes and rapid compression machines. A new detailed chemical kinetic reaction mechanism (1845 reactions versus 232 species) was developed and used in this work. The present model shows good agreement with the present experimental data and those taken from the literature. To interpret the results, sensitivity analyses were performed.

This paper presents an experimental investigation of the influence of fuel density and fuel viscosity on the flow characteristics generated from a high pressure Diesel injector, equipped with conical orifices. For this purpose, mass flow rate measurements were performed with nine different fuels over operating conditions from 30 to 180 MPa of injection pressure and from 1 to 9 MPa of back pressure. Fuel viscosity was varied from 0.6 to 7 mm2/s and fuel density from 683 to 876 kg/m3 at the operating temperature. Fuel viscosity induces a decrease of up to 10% in the discharge coefficient at low pressure difference but for higher pressure difference, fuel density remains the only fuel property driving the mass flow rate. The discharge coefficient is also a function of the pressure difference at any given Reynolds number under non-cavitating conditions. A correlation is suggested to estimate the discharge coefficient taking into account the fuel properties and the operating conditions.

Dynamic and scalar structures of turbulent premixed flames stabilized in the opposed jet configuration are investigated. The opposed jet burner facility allows control of many parameters of turbulent premixed combustion, including the bulk and turbulent strain rates. The main mixture streams are enveloped in coflowing air to reduce flame bouncing, to render the turbulence structure more isotropic and homogeneous at the stagnation plane, and, therefore, to better match the major modelling assumptions. Detailed investigation of the velocity field confirms the strong variation of the initial turbulence parameters as the flow stagnates, and the similarity between nonreacting and reacting flow fields. It is clearly demonstrated that turbulent premixed flames stabilized in the opposed jet configuration are ideal for turbulent combustion model validation, as modification of the turbulence field by the flame is minimized. Two different extinction regimes of opposed jet turbulent premixed flames are identified. For small andmoderate burner separations, the total strain rate causes extinction. For large separation distaces, mixing and dilution by the external air are the causes of flame extinction. For the first regime, it is confirmed that bulk and turbulent strain rates should be taken into account simultaneously to establish the robustness of turbulent premixed flames. For the first time, the scalar structure of opposed jet turbulent premixed flames is explored by laserinducedrayleigh light-scattering technique. All the relevant parameters of instantaneous flame fronts are determined, and estimates of the flame surface density are proposed. It is shown that, for constant mixture composition and constant initial turbulence structure, the flame surface density increases with the bulk strain rate before its extinction value is closely approached. This result is discussed by using various ingredients of flamelet models.

International audience

Oxygen-enriched combustion is of great interest for industrial applications, since membrane separation technology can be used. The objective of this work is to provide unique data on laminar burning velocity, a key parameter in real combustion development, for the oxygen-enriched combustion of an iso-octane/air mixture for various dilution (by air or CO 2 ) cases. Experiments were carried out in a stainless steel combustion chamber at atmospheric pressure and 373 K. The iso-octane was mixed with a mixture of O 2 , CO 2, and N 2 . The volume fraction of O 2 was varied from 21% to 29% and CO 2 was varied from 0% to 28%. The classical shadowgraphy technique was used to detect the reaction zone in order to deduce the un-stretched burning velocity, using a nonlinear methodology. All the experimental data were compared with the numerical results obtained with chemical kinetic schemes available in the literature. For further experimental investigations, a correlation is proposed to predict laminar burning velocity as a function of the quantity of O 2 and CO 2 in the gas mixture. Finally, analytical and experimental data concerning Markstein length are discussed.

International audience; A mixture of ethyl acetate (EtAc), ethyl propionate (EtPr), and ethyl butyrate (EtBu) can be obtained from low-value biomass wastes in a simple biochemical process that includes acidogenic fermentation. Their proportions in the mixture vary according to fermentation conditions and biomass feedstocks. To help direct the fermentation process, they have been previously analyzed separately in a homogeneous charge compression ignition (HCCI) engine, but the characteristics of the mixtures are unknown. Using mixture design, we investigated how the tricomponent fuel blends of these esters impact the combustion characteristics in HCCI. This paper reports the ignition timing for 12 different blends. It characterizes the direct effect and the types of interaction using a blending model. The ignition timing is mainly determined by the proportion of EtAc and EtBu, with EtBu having a smaller ignition delay than EtAc. EtPr has no appreciable direct effect on this timing, but it has an antagonistic effect on EtAc and EtBu. The faster ignition of EtBu could help decrease the inlet temperature and, therefore, be more compatible with real engines. Moreover, dilution with EtPr could also reduce the effect of mixture variability. The successful implementation of these esters in HCCI may, however, require blending them with other fuels to further decrease the inlet temperature. This will remain the focus of future studies.

The homogeneous charge compression ignition (HCCI) engine can be run on a large range of fuels if the appropriate operating conditions are chosen. This can improve the efficiency of biofuel production from low-value biomass by suppressing the need for the transformation process to obtain products that are compatible with spark ignition or compression ignition engines. A simple biochemical process that includes acidogenic fermentation and produces a mixture of various esters can take advantage of this flexibility. However, the behavior of this mixture under HCCI conditions needs to be characterized. It can also have a great impact on the HCCI operating limits and its successful implementation. Using an HCCI engine, we investigated how the operating limits are modified by the combustion characteristics of three of these esters: ethyl acetate, ethyl propionate, and ethyl butanoate. This paper reports the experimental results for each of these products and for ethanol taken as the reference fuel. It also analyzes their effects on the ignition timing and the combustion rate. For the selected operating conditions, stable HCCI operations on a large range of equivalence ratios were obtained for every fuel The difference in specific heats of the air/fuel mixtures and in the ignition kinetics both contributed to the ignition characteristics. Ethanol ignites earlier, which leads to a low upper limit, whereas the late ignition of ethyl acetate shifts the operating zone upward due to smoothed high loads but unstable low loads. As a consequence, these low-grade products can be used in an HCCI engine. Fuel blends of these products may take advantage of the different combustion characteristics to extend the HCCI zone. Still, the range of this extension is difficult to estimate and the research of the optimal fuel blend composition will, therefore, remain the focus of future work.

International audience; 1-Pentanol is among the possible alcohols usable as an alternative to conventional gasoline and diesel fuels. In order to better understand the combustion characteristics of 1-pentanol, this study presents new experimental data for 1-pentanol in two complementary experimental configurations. Stable species concentration profiles were measured in a jet-stirred reactor (JSR) at 10 atm over a range of equivalence ratios and temperatures. Laminar flame speeds of 1-pentanol-air premixed laminar flame were measured at 1 atm and 423 K for equivalence ratios of 0.7–1.4. The oxidation of 1-pentanol in these experimental configurations was modeled using an extended detailed chemical kinetic reaction mechanism (2099 reactions involving 261 species) derived from a previously proposed scheme for the oxidation of methanol-, ethanol-, and 1-butanol-based fuels. The proposed mechanism shows good agreement with the present experimental data. Reaction path analyses were conducted to interpret the JSR results.

Experimental investigations on the effects of hydrogen addition to spark-ignition (SI) engines running under lean and diluted conditions are presented in this paper. Experiments were carried out in a mono-cylinder SI engine with different rates of nitrogen dilution (0 to 20 per cent by volume in the total mixture) and hydrogen/iso-octane blends (from 0 to 80 per cent by volume in the fuel). The study of the impact of hydrogen addition on combustion characteristics and emissions was performed for two different engine speeds and loads. The equivalence ratio, the rate of dilution, and the intake pressure were varied either separately or simultaneously in order to maintain a constant engine load at a fixed hydrogen fraction in the fuel. The lean and dilution operating limits were also determined for all the iso-octane/hydrogen/air/nitrogen mixtures investigated, and results show that these limits are extended only when the hydrogen percentage in the fuel is higher than 40 per cent by volume. At a fixed engine load, hydrocarbon (HC) and CO emissions decrease with an increase in the hydrogen fraction in the intake mixture, while NO x emissions are mainly affected by the equivalence ratio and by the amount of dilution. Pumping losses, combustion efficiency and indicated efficiency are also improved with the addition of hydrogen. High values of indicated engine efficiency with low values of HC, NO x, and CO emissions can be achieved by combining hydrogen addition with lean and/or diluted conditions.

International audience; This paper assesses different butanol–gasoline blends used in a port fuel-injection, spark-ignition engine to quantify the influence of butanol addition on the emission of unburned hydrocarbons, carbon monoxide, and nitrogen oxide. Furthermore, in-cylinder pressure was measured to quantify combustion stability and to compare the ignition delay and fully developed turbulent combustion phases as given by 0%–10% and 10%–90% Mass Fraction Burned (MFB). The main findings are: 1) a 40% butanol/60% gasoline blend by volume (B40) minimizes HC emissions; 2) no significant change in NOx emissions were observed, with the exception of the 80% butanol/20% gasoline blend; 3) the addition of butanolimproves combustion stability as measured by the COV of IMEP; 4) butanol added to gasoline reduces ignition delay (0%–10% MFB); and 5) the specific fuel consumption of B40 blend is within 10% of that of pure gasoline for stoichiometric mixture.

International audience; In all experimental configurations, the flames are affected by stretch (curvature and/or strain rate). To obtain the unstretched flame speed, independent of the experimental configuration, the measured flame speed needs to be corrected. Usually, a linear relationship linking the flame speed to stretch is used. However, this linear relation is the result of several assumptions, which may be incorrected. The present study aims at evaluating the error in the laminar burning speed evaluation induced by using the traditional linear methodology. Experiments were performed in a closed vessel at atmospheric pressure for two different mixtures: methane/air and iso-octane/air. The initial temperatures were respectively 300 K and 400 K for methane and iso-octane. Both methodologies (linear and nonlinear) are applied and results in terms of laminar speed and burned gas Markstein length are compared. Methane and iso-octane were chosen because they present opposite evolutions in their Markstein length when the equivalence ratio is increased. The error induced by the linear methodology is evaluated, taking the nonlinear methodology as the reference. It is observed that the use of the linear methodology starts to induce substantial errors after an equivalence ratio of 1.1 for methane/air mixtures and before an equivalence ratio of 1 for iso-octane/air mixtures. One solution to increase the accuracy of the linear methodology for these critical cases consists in reducing the number of points used in the linear methodology by increasing the initial flame radius

International audience; The impact of dilution on laminar burning speed of two different fuels (methane and isooctane) is studied. In the present study, three different diluents are used--nitrogen, carbon dioxide, and a mixture representative of exhaust gases issued from a stoichiometric combustion of methane. Experimental results and PREMIX computations of the CHEMKIN package, using two different kinetic schemes, are presented and compared with literature results, when available. Initial pressure and temperature conditions are respectively 0.1 MPa and 300 K. For both fuels, a larger decrease of the laminar burning speed is obtained for carbon dioxide dilution than for nitrogen dilution. This observation is directly linked to the increase in heat capacity of the dilution gas but also to the carbon dioxide dissociation, even if the heat capacity effect seems to be predominant.

International audience

International audience

International audience; An original application of the Laser-Induced-Incandescence (LII) technique was set up to quantify soot particles inside the combustion chamber of an optically accessible Direct-Injection Diesel engine. Planar soot concentration and local particle diameter were measured for several Exhaust Gas Recirculation (EGR) rates. The impact of the injection timing on the soot evolution for the highest EGR rate was also studied. Based on the analysis of LII images it is shown that the planar distribution of soot becomes more and more uniform across the combustion chamber and globally the soot maximum more important with the EGR rates increase. High EGR rates, combined with a retarded start of injection may lead to lower soot production inside the combustion chamber. Comparison between exhaust and in-cylinder soot concentration highlights the effect of post-combustion oxidation on the particle-emissions amount.

available, unlimited, public

An original approach of laser-induced incandescence consisting in the simultaneous recording of the two-color-time-resolved and 2D LII signal is described in this paper. The application of this approach in an atmospheric pressure diffusion flame fueled with isooctane as well as inside the combustion chamber of a diesel engine is presented. Soot volume fraction and primary particle diameters are calculated, and the results are discussed. The mean diameter estimated by fitting the LII modeled curve on the experimental one is compared with the results obtained through soot sampling and microscope analyzing. The influence of the thermal accommodation coefficient and soot refractive index function is also discussed.

International audience

Laser-induced Rayleigh light scattering technique is used to determine relevant time scales in turbulent premixed flames in the flamelet regime. The major achievement of this work is the experimental determination of flamelet transit times at a given point within the flame brush. This information coupled with the usual flamelet description assumptions allows an estimate of the local mean reaction rate to be made. The statistics of instantaneous flamelet orientations are also deduced. The technique is implemented in open turbulent methane-air premixed flames stabilized on the rim of a Bunsen type burner. The variation versus the progress variable of the estimated mean reaction rate is compared with the predictions of the Bray-Moss-Champion-Libby model.

As an alternative to second generation ethanol, valeric esters can be produced from lignocellulose through levulinic acid. While some data on these fuels are available, only few experiments have been performed to analyze their combustion characteristics under engine conditions. Using a traditional spark ignition engine converted to mono-cylinder operation, we have investigated the engine performances and emissions of methyl and ethyl valerates. This paper compares the experimental results for pure valeric esters and for blends of 20% of esters in PRF95, with PRF95 as the reference fuel. The esters propagate faster than PRF95 which requires a slight change of ignition timing to optimise the work output. However, both the performances and the emissions are not significantly changed compared to the reference. Accordingly, methyl and ethyl valerate represent very good alternatives as biofuels for SI engines. Future studies will focus on testing these esters in real application engines and performing endurance tests.

International audience; Most studies of sprays produced from diesel injectors are performed at "usual" ambient temperature. These studies focus on spray development and the impact of flow inside the injector nozzles. Various mechanisms can produce initial perturbations of the jet and depend on nozzle geometry--the cavitation phenomenon through the nozzle hole is one of these. Until recently, only a few studies have examined spray development for cold temperature conditions (i.e., below 273 K). The objective of this study is to provide data regarding the phenomena occurring in a diesel fuel jet above temperature conditions, representative of a cold start for a direct-injection compression ignition engine. In this study, the discharge coefficient has been estimated by performing experiments inside a climatic chamber, from 253 to 293 K, for three different fuels, namely, n-heptane, decane, and arctic diesel fuel.

Magnesium particles (1-2 mm) are ignited by a hot wire in room-temperature C02 environment under microgravity and normal gravity conditions. In about 50% of experiments, regardless of the gravity level, a pulsating combustion regime is obtained which may be associated with the relatively low heat release in such oxidizer as C02 and high heat losses. For stable combustion regime, the burning times measured in micro gravity follows a quadratic dependence with the particle size, which is typical for the diffusion-controlled combustion. Good quantitative agreement with previous data obtained for particles of different sizes indicates that combustion mechanisms of magnesium in C02 are identical in the range of particle sizes 0.05-5 mm.

International audience; The main purpose of this paper is to determine the laminar burning velocities and Markstein lengths of α-pinene/air premixed flames using the spherical expanding flames method. Experiments are carried out in a spherical vessel at atmospheric pressure. The effects of equivalence ratio (0.7 to 1.4) and unburned gas temperature (353 to 453 K) are studied. Unstretched propagation speeds and Markstein lengths are obtained using a nonlinear methodology. A correlation is developed to calculate laminar burning speeds as a function of equivalence ratio and temperature. The PREMIX code is used to compute laminar flame speeds with the San Diego chemical kinetic mechanism of JP-10, which is an isomer of α-pinene. The measured values are compared to those of literature.

Chapter 3; International audience; Since Shell company announced its new process in 2010 to produce levulinic acid derivatives from lignocellulose, valerates have been studied in order to assess their capacity as potential fuels for car engines. Since valerates exhibit properties similar to those of engine fuels, they can be used as biofuels or blended with gasoline or diesel. Recent works have focused on determining the properties such as ignition delay, laminar burning speeds, and kinetic schemes to ensure their suitability as fuels. From few studies, it seems that blends of valerates can be used as fuels in SI or compression engines without any drawback, but more real road experiments are needed to improve the real ability of valerates to be used as substitute fuels. As propyl valerate has properties between gasoline and diesel types, studies are also needed to improve it as a potential fuel. It is still not possible to determine whether it is cost effective to produce these products and how better for the world biomass than ethanol production. Moreover, it is still necessary to decide which plant, such as giant reed, would be the most appropriate to produce valerates. This chapter redraws the scientific evolution for these biofuels from the process itself until its potential use.

International audience

International audience; The burning rate in the vicinity of the piston is estimated from the flamlet, formulation. The flame surface density Σ is measured by laser sheet tomography for three equivalence ratio methane/air flames (1.0.9, 0.8) in a spark-ignition transparent engine. Two imaging configurations are realized: five horizontal planes at different distances from the piston (0, 1, 2, 3, and 5 mm) and a vertical one at the center of the combustion chamber. To decrease the effect of the cycle-to-cycle variation on the average flame front thickness and on the determination of the mean progress variable, several methods are tested to extract the mean flame front contour. Two flame surface density methods are compared: one based on the Bray-Moss-Libby flamelet crossing density model and the other based on the burning rate formulation by Gouldin. It appears that the last one is better adapted for the determination of Σ for weakly wrinkled flames and a good compromise to estimate the burning rate locally in vicinity of the piston. All methods are also validated by applying to freely propagating spherical flames. Model adaptation is realized to include the evolution of the flamelet scalar orientation near the piston. Finally, the burning rate is determined as a function of the distance between the wall and the mean flame contour.