No description available.

No description available.

No description available.

Doctor of Philosophy(PhD) / This body of research uses numerical and experimental investigative techniques to further the understanding of autoignition. Hydrogen/nitrogen and methane/air fuel configurations of turbulent lifted flames in a vitiated coflow burner are used as model flames for this investigation. Characterisation was undertaken to understand the impact of controlling parameters and the overall behaviour of the flames, and to provide a body of data for modelling comparisons. Modelling of the flames was conducted using the PDF-RANS technique with detailed chemistry incorporated using In-situ Adaptive Tabulation (ISAT) within the commercial CFD package, FLUENT 6.2. From these investigations, two numerical indicators for autoignition were developed: convection-reaction balance in the species transport budget at the mean flame base; and the build-up of ignition precursors prior to key ignition species. These indicators were tested in well defined autoignition and premixed flame cases, and subsequently used with the calculated turbulent lifted flames to identify if these are stabilised through autoignition. Based on learnings from the modelling, a quantitative, high-resolution simultaneous imaging experiment was designed to investigate the correlations of an ignition precursor (formaldehyde: CH2O) with a key flame radical (OH) and temperature. Rayleigh scattering was used to measure temperature, while Laser Induced Fluorescence (LIF) was used to measure OH and CH2O concentrations. The high resolution in the Rayleigh imaging permitted the extraction of temperature gradient data, and the product of the OH and CH2O images was shown to be a valid and useful proxy for peak heat release rate in autoigniting and transient flames. The experimental data confirmed the presence of formaldehyde as a precursor for autoignition in methane flames and led to the identification of other indicators. Sequenced images of CH2O, OH and temperature show clearly that formaldehyde exists before OH and peaks when autoignition occurs, as confirmed by images of heat release. The CH2O peaks decrease later while those of OH remain almost unchanged in the reaction zone.

Doctor of Philosophy(PhD) / This body of research uses numerical and experimental investigative techniques to further the understanding of autoignition. Hydrogen/nitrogen and methane/air fuel configurations of turbulent lifted flames in a vitiated coflow burner are used as model flames for this investigation. Characterisation was undertaken to understand the impact of controlling parameters and the overall behaviour of the flames, and to provide a body of data for modelling comparisons. Modelling of the flames was conducted using the PDF-RANS technique with detailed chemistry incorporated using In-situ Adaptive Tabulation (ISAT) within the commercial CFD package, FLUENT 6.2. From these investigations, two numerical indicators for autoignition were developed: convection-reaction balance in the species transport budget at the mean flame base; and the build-up of ignition precursors prior to key ignition species. These indicators were tested in well defined autoignition and premixed flame cases, and subsequently used with the calculated turbulent lifted flames to identify if these are stabilised through autoignition. Based on learnings from the modelling, a quantitative, high-resolution simultaneous imaging experiment was designed to investigate the correlations of an ignition precursor (formaldehyde: CH2O) with a key flame radical (OH) and temperature. Rayleigh scattering was used to measure temperature, while Laser Induced Fluorescence (LIF) was used to measure OH and CH2O concentrations. The high resolution in the Rayleigh imaging permitted the extraction of temperature gradient data, and the product of the OH and CH2O images was shown to be a valid and useful proxy for peak heat release rate in autoigniting and transient flames. The experimental data confirmed the presence of formaldehyde as a precursor for autoignition in methane flames and led to the identification of other indicators. Sequenced images of CH2O, OH and temperature show clearly that formaldehyde exists before OH and peaks when autoignition occurs, as confirmed by images of heat release. The CH2O peaks decrease later while those of OH remain almost unchanged in the reaction zone.

This master´s thesis solves the coal dust autoignition problems in the heatplant Karviná. The first part is devoted to the combustion of pulverized coal possibilities and technical description at the heatplant Karviná equipment. The second part deals with the phenomenon of the coal self-ignition itself. In the last part the calculation of the mill circuit heat balance is solved as well as it´s inerting protection. The final part of the thesis is devoted to the coal powder reservoir, where the fuel blaze is the most common.

Pour améliorer le rendement des moteurs essence, une méthode efficace est le downsizing qui consiste en la diminution de la cylindrée moteur compensée par l’ajout d’un compresseur pour maintenir la puissance. Lorsque le niveau de downsizing est trop important les fortes pression et températures rencontrées favorisent l’apparition de phénomènes d’auto-allumage de type cliquetis ou rumble néfastes pour l’intégrité du moteur. Ce type de phénomène, aujourd’hui encore mal compris, constitue une limite à l’utilisation du downsizing. Dans cette thèse la Simulation aux Grandes Echelles est utilisée pour étudier ce type de combustion dite anormale. L’objectif est de proposer une méthodologie numérique capable de reproduire leurs apparitions pour en étudier les mécanismes. L’auto-allumage est un mode de combustion sensible aux variations des conditions thermodynamiques locales. Des méthodes numériques précises et des modèles appropriés, en particulier pour la thermique paroi doivent donc être utilisés. La première partie de ce manuscrit présente la méthodologie numérique proposée et en particulier deux aspects développés lors de cette thèse: un modèle d’auto-allumage qui permet de reproduire le délai d’auto-allumage des gaz frais avec un schéma cinétique réduit et une méthodologie de couplage entre la chambre de combustion et la culasse permettant de définir des champs de températures paroi réalistes. La seconde partie de ce manuscrit présente les résultats de deux études numériques reproduisant certains points de fonctionnement d’un moteur expérimental. La première étude est réalisée à l’aide de modèles de combustion de la littérature et vise à reproduire le comportement expérimental pour diverses variations paramétriques influant sur la combustion. La seconde étude est réalisée à l’aide des modèles développés dans cette thèse afin d’étudier l’impact de la thermique paroi dans les mécanismes d’apparition des combustions anormales. / In order to improve the efficiency of gasoline engines, one efficient solution resides in engine downsizing which consists in the diminution of the engine size with the adjunction of a compressor to keep the power output. When the downsizing level is important, the high pressure and temperature levels promote auto-ignition phenomena such as knocking or rumble that can damage the engine. This kind of combustion, still misunderstood, is a limit to further use downsizing. In this thesis, Large Eddy Simulation is used to study this kind of abnormal combustions. The objective is to propose a numerical methodology able to reproduce its apparition and to understand its mechanisms. Auto-ignition is a combustion regime very sensitive to the variations of local thermodynamic conditions. Precise numerical methods and appropriate models, especially for thermal boundary conditions must be used. The first part of this manuscript presents the proposed numerical methodology and in particular two aspects implemented during this thesis: an auto-ignition model that permits to reproduce auto-ignition delays with reduced kinetic schemes and a coupling methodology between combustion chamber and cylinder head in order to obtain realistic temperature fields for the boundary conditions. The second part of this manuscript presents the results of two numerical studies that reproduce some operating points from an experimental engine database. The first study is performed using combustion models from the literature and aims at reproducing experimental behavior for various parametric variations impacting the combustion. The second study is performed thanks to the numerical models implanted in this thesis in order to evaluate the impact of the thermal boundary conditions on the mechanisms leading to abnormal combustions.

Une étude expérimentale de la vaporisation de goutte d'éthanol et de propan-1-ol a été réalisée. Le dispositif expérimental est constitué d’une enceinte chauffée à l’intérieur de laquelle se trouve le support de goutte. Il est formé par deux fibres en quartz croisées. La goutte d'alcool se trouve à l'intersection de ces fibres. Le diamètre initial de la goutte est contrôlé, il varie entre 300 et 600µm. L’étude est faite de 298 à 973K et à pression atmosphérique. La théorie de l’état quasi-stationnaire est utilisée pour comparer et expliquer tous les résultats expérimentaux. L'impact réel de la concentration d'eau sur la vaporisation d'une goutte d'éthanol est également examiné. Deux périodes sont observées sur les courbes en d2. Ceci montre clairement que la vaporisation d'une goutte d'éthanol est accompagnée par la condensation de la vapeur d'eau à la surface de la goutte. L’évolution des taux instantanés de vaporisation du propan-1-ol et de l'éthanol confirme ce phénomène. Les délais d’autoinflammation de l'éthanol, du propan-1-ol et des mélanges d'éthanol et de l'eau ont été mesurés dans une machine de compression rapide. Les conditions de l’étude sont : une pression de compression de 30bar, la gamme de température variant de 750 à 860K, pour des mélanges stoechiométriques carburant/air. Les délais d’autoinflammation enregistrés diminuent lorsque la température augmente. Le propan-1-ol est plus réactif que l'éthanol, ce qui se traduit par des délais d’autoinflammation plus courts. Cependant, l'addition de l'eau à l'éthanol augmente la réactivité du mélange et se traduit par une réduction des délais d’autoinflammation / Detailed investigation of the vaporization of an isolated of ethanol and 1-propanol droplet was carried out in this experimental study. The experimental set-up consists of a heated chamber with a cross quartz fibers configuration as droplet support. An alcohol droplet is located at the intersection of the cross quartz fibre with a controlled initial diameter (300-600µm). Ambient temperature is varied from 298 to 973K at atmospheric pressure. The quasi-steady theory has been used to compare and to explain all experimental results. The real impact of the water concentration on the vaporization rate of an ethanol droplet is also examined, where two ‘quasi-steady’ periods are observed on the d2-curves, clearly showing that the vaporization of an ethanol droplet is accompanied by the simultaneous condensation of water vapour on the droplet surface and thus the temporal evolution of the droplet squared diameter exhibits an unsteady behavior. The histories of the instantaneous vaporization rates of both 1-propanol and ethanol droplets confirm this phenomenon. The autoignition experimental study of ethanol, 1-propanol and blends of ethanol and water have been carried out in a rapid compression machine at a compressed pressure of 30bar over a temperature range of 750-860K for stoichiometric mixture of fuel and air. The ignition delay times recorded show a significant decrease with increasing temperature. 1-propanol is more reactive than ethanol, which results in shorter ignition delay times. However, water addition to ethanol increases the reactivity of the mixture and results in a shorter ignition delay times than 1-propanol

La stratégie de combustion par auto-inflammation d’une charge homogène en composition s’intègre dans une démarche de réduction des émissions de particules et de NOx, tout en conservant les rendements thermiques élevés des moteurs Diesel classiques. Pour contrôler ce nouveau mode de combustion, une compréhension fine des mécanismes de couplage entre l'aérodynamique et la thermochimie est nécessaire. Des simulations numériques directes d'un écoulement homogène, turbulent, réactif et subissant une compression, ont été effectués. Deux régimes d'auto-inflammation ont ainsi pu être définis. Le premier, dit quasi-homogène, est caractérisé par une auto-inflammation en masse d'un volume important du mélange réactif et s'accompagne de fortes ondes de pression. Dans le second régime, dit localisé, les noyaux s'initient de manière plus sporadique dans l'espace et dans le temps et aucune onde de pression significative n'est générée lors de l'auto-allumage. / Combustion by autoignition of a homogeneous charge aims at reducing particulate matter as well as NOx emissions, while maintaining higher thermal efficiency of conventional diesel engines. To control this new mode of combustion, a fine understanding of the mechanisms of coupling between aerodynamics and thermochemistry is required. Direct Numerical Simulations of a turbulent reactive flow, undergoing a compression, have been performed. This study led to identification of two regimes. The first, known as quasi-homogeneous, is characterized by volumetric autoignition of large zones of the reactive mixture and results in the generation of strong pressure waves, which are potentially dangerous for the structure of engines. In the second regime, called localized, hot spots are initiated more sporadically in space and time, and their topology is such that no significant pressure wave is generated.

No description available.