Die Rußbildung in der Pyrolyse von n-Hexan, Benzol und Kohlenwasserstoffen im Stoßrohr wird mit hoher Zeitauflösung absorptionsspektroskopisch beobachtet und Gasproben durch ein besonders schnelles Ventil gesammelt. Vom Beginn der Pyrolyse bis zum Einsetzen der Rußbildung wurde stets eine Induktionszeit beobachtet, deren Dauer von der Temperatur, der Kohlenstoffkonzentration und der Struktur des pyrolysierten Kohlenwasserstoffs abhängt. Dieser Zusammenhang wird mit einem Arrhenius-Ansatz beschrieben. Der Vorfaktor A ist für Aromaten eine Größenordnung kleiner als für Alkane. Die scheinbare Aktivierungsenergie beträgt (220 ± 10) kJ/mol. Das Rußmassenwachstum, das der Induktionsperiode folgt, wird mit einem Gesetz erster Ordnung beschrieben. Durch Normierung der Geschwindigkeitskonstanten auf die Kohlenstoffdichte zeigt, daß das Rußmassenwachstum in Pyrolysen und in vorgemischten Ethylenflammen vergleichbar schnell abläuft. Die Rußausbeute der Aromaten und Acetylen hat bei bei 1800 K ein Maximum. Für Ethylen und Alkane liegt diese charakteristische Temperatur um rund 100 K höher. Neben Ruß und Wasserstoff sind Acetylen gefolgt von Methan und Ethylen die wichtigsten Hauptprodukte. Die polycyclischen Aromaten tragen keine Seitengruppe und enthalten maximal einen Fünfring. Der Rußpartikeldurchmesser ist im Bereich von 30 nm. Der Einfluß von Eisenpentacarbonyl auf die Rußbildung ist gering.

O objetivo desta tese é o estudo de formação de fuligem em chamas laminares de difusão. Para o modelo de formação de fuligem é escolhido um modelo semi-empírico de duas equações para prever a fração mássica de fuligem e o número de partículas de fuligem. O modelo descreve os processos de nucleação, de crescimento superficial e de oxidação das partículas. Para o modelo de radiação, a perda de calor por radiação térmica (gás e fuligem) é modelada considerando o modelo de gás cinza no limite de chama opticamente fina (OTA - Optically Thin Approximation). São avaliados diferentes modelos de cálculo das propriedades de transporte (detalhado e simplificado). Em relação à cinética química, tanto modelos detalhados quanto reduzidos são utilizados. No presente estudo, é explorada a técnica automática de redução conhecida como Flamelet Generated Manifold (FGM), sendo que esta técnica é capaz de resolver cinética química detalhada com tempos computacionais reduzidos. Para verificar o modelo de formação de fuligem foram realizados uma variedade de experimentos numéricos, desde chamas laminares unidimensionais adiabáticas de etileno em configuração tipo jatos opostos (counterflow) até chamas laminares bidimensionais com perda de calor de etileno em configuração tipo jato (coflow). Para testar a limitação do modelo os acoplamentos de massa e energia entre a fase sólida e a fase gasosa são investigados e quantificados para as chamas contra-corrente Os resultados mostraram que os termos de radiação da fase gasosa e sólida são os termos de maior importancia para as chamas estudas. Os termos de acoplamento adicionais (massa e propriedade termodinâmicas) são geralmente termos de efeitos de segunda ordem, mas a importância destes termos aumenta conforme a quantidade de fuligem aumenta. Como uma recomendação geral o acoplamento com todos os termos deve ser levado em conta somente quando a fração mássica de fuligem, YS, for igual ou superior a 0.008. Na sequência a formação de fuligem foi estudada em chamas bi-dimensionais de etileno em configuração jato laminar usando cinética química detalhada e explorando os efeitos de diferentes modelos de cálculo de propriedades de transporte. Foi encontrado novamente que os termos de radiação da fase gasosa e sólida são os termos de maior importância e uma primeira aproximação para resolver a chama bidimensional de jato laminar de etileno pode ser feita usando o modelo de transporte simplificado. Finalmente, o modelo de fuligem é implementado com a técnica de redução FGM e diferentes formas de armazenar as informações sobre o modelo de fuligem nas tabelas termoquímicas (manifold) são testadas A melhor opção testada neste trabalho é a de resolver todos os flamelets com as fases sólida e gasosa acopladas e armazenar as taxas de reação da fuligem por área de partícula no manifold. Nas simulações bidimensionais estas taxas são então recuperadas para resolver as equações adicionais de formação de fuligem. Os resultados mostraram uma boa concordância qualitativa entre as predições do FGM e da solução detalhada, mas a grande quantidade de fuligem no sistema ainda introduz alguns desafios para a obtenção de bons resultados quantitativos. Entretanto, este trabalho demonstrou o grande potencial do método FGM em predizer a formação de fuligem em chamas multidimensionais de difusão de etileno em tempos computacionais reduzidos. / The objective of this thesis is to study soot formation in laminar diffusion flames. For soot modeling, a semi-empirical two equation model is chosen for predicting soot mass fraction and number density. The model describes particle nucleation, surface growth and oxidation. For flame radiation, the radiant heat losses (gas and soot) is modelled by using the grey-gas approximation with Optically Thin Approximation (OTA). Different transport models (detailed or simplified) are evaluated. For the chemical kinetics, detailed and reduced approaches are employed. In the present work, the automatic reduction technique known as Flamelet Generated Manifold (FGM) is being explored. This reduction technique is able to deal with detailed kinetic mechanisms with reduced computational times. To assess the soot formation a variety of numerical experiments were done, from one-dimensional ethylene counterflow adiabatic flames to two-dimensional coflow ethylene flames with heat loss. In order to assess modeling limitations the mass and energy coupling between soot solid particles and gas-phase species are investigated and quantified for counterflow flames. It is found that the gas and soot radiation terms are of primary importance for flame simulations. The additional coupling terms (mass and thermodynamic properties) are generally a second order effect, but their importance increase as the soot amount increases As a general recommendation the full coupling should be taken into account only when the soot mass fraction, YS, is equal to or larger than 0.008. Then the simulation of soot is applied to two-dimensional ethylene co-flow flames with detailed chemical kinetics and explores the effect of different transport models on soot predictions. It is found that the gas and soot radiation terms are also of primary importance for flame simulations and that a first attempt to solve the two-dimensional ethylene co-flow flame can be done using a simplified transport model. Finally an implementation of the soot model with the FGM reduction technique is done and different forms for storing soot information in the manifold is explored. The best option tested in this work is to solve all flamelets with soot and gas-phase species in a coupled manner, and to store the soot rates in terms of specific surface area in the manifold. In the two-dimensional simulations, these soot rates are then retrieved to solve the additional equations for soot modeling. The results showed a good qualitative agreement between FGM solution and the detailed solution, but the high amount of soot in the system still imposes some challenges to obtain good quantitative results. Nevertheless, it was demonstrated the great potential of the method for predicting soot formation in multidimensional ethylene diffusion flames with reduced computational time.

L’aviation émet de grandes quantités de gaz et de particules dans l’atmosphère contribuant, d’une part, à la détérioration de la qualité de l’air à une échelle locale et d’autre part, au forçage radiatif atmosphérique et donc au changement climatique. Une voie envisagée pour limiter l’impact de l’aviation est l’utilisation de carburants alternatifs. Les biocarburants sélectionnés dans cette optique tendent à avoir des teneurs en soufre et en composés aromatiques réduites, ce qui induit une diminution de la quantité d’acide sulfurique formé dans les panaches d’avions ainsi que des suies émises. La modification de la nature et de la composition des carburants utilisés peut entraîner des conséquences inattendues. Il s’avère alors essentiel d’étudier et de déterminer l’évolution des aérosols dans les panaches d’avions. Pour cela, un modèle microphysique précédemment testé lors de la combustion de kérosène classique a été utilisé et amélioré. Après avoir déterminé les émissions « types » des carburants alternatifs, des simulations ont été menées afin de prédire l’évolution et le comportement des aérosols. Plusieurs processus ont nécessité des révisions tels que la congélation homogène ou encore le comportement des composés organiques. / Aircraft emit important amounts of particulate and gaseous matter in the atmosphere contributing on the one hand to local air pollution and on the other hand to the atmospheric radiative forcing and to climate change. Introducing alternative fuels in aviation can be considered as a viable option to reducing the impact of aviation, being economically and environmentally sustainable. These selected biofuels tend to have lower aromatic and sulphur contents inducing a simultaneous reduction in sulphuric acid and soot emissions. However modifying the nature and composition of the fuel used can entail unexpected consequences. It is therefore essential to study and determine the evolution of aerosols in the aircraft plume. To manage this task, a microphysical trajectory box, previously tested with standard kerosene, has been developed. After an assessment concerning the typical emissions from the combustion of biofuels in aviation, simulations have been undertaken in order to predict aerosol evolution. Several microphysical processes have been revised such as droplet homogeneous freezing or the behaviour of organic compounds.

A program slice consists of a subset of the statements of a program that can potentially affect values computed at some point of interest. Such a point of interest along with a set of variables is called a slicing criterion. Slicing tools are useful for several applications, such as program understanding, testing, program integration, and so forth. Slicing object oriented programs has some special problems that need to be addressed due to features like inheritance, polymorphism and dynamic binding. Alias analysis is important for precision of slices. In this thesis we implement a slicing tool for sequential Java programs in the Soot framework. Soot is a front-end for Java developed at McGill University and it provides several forms of intermediate code. We have integrated the slicer into the framework. We also propose an improved technique for intraprocedural points-to analysis. We have implemented this technique and compare the results of the analysis with those for a flow-insensitive scheme in Soot. Performance results of the slicer are reported for several benchmarks.

Le transport routier est responsable d’une part des émissions de particules fines, notamment dans les grandes agglomérations. Celles-ci ayant des répercussions graves sur la santé humaine et l’environnement, des normes d’émissions des véhicules de plus en plus strictes sont mises en place. Des limites en nombre de particules sont imposées et une obligation de connaitre l’état de santé des organes de post-traitement des gaz d’échappement est également entrée en vigueur (On-Board Diagnostic : OBD).Les capteurs résistifs ont montré de bons résultats pour la mesure de la concentration massique des particules et ont l’avantage d’être simples à mettre en oeuvre, robustes et peu onéreux. Le principe de fonctionnement du capteur est basé sur la mesure de la conductance entre des électrodes de platine. Celle-ci augmente avec le dépôt de suies.L’objectif de ce travail de thèse est de parvenir à déterminer les paramètres clés qui affectent la réponse du capteur. Trois flux de suies présentant des distributions en taille différentes ont été sélectionnés et caractérisés. Un quatrième flux de suies a permis d’appréhender l’impact d’une concentration massique inférieure. La réponse du capteur à ces différents flux de suies a été étudiée. Nous avons montré que la sensibilité et le temps de réponse du capteur sont optimaux pour une tension de polarisation donnée entre les deux électrodes dont la valeur dépend de la distribution en taille des suies. Ce phénomène a été expliqué par les propriétés électriques différentes des suies et modélisé par un équilibre entre accumulation et combustion par effet Joule des suies déposées permettant de simuler la réponse temporelle du capteur. / Road transport contributes to a part of the particulate matter emissions, especially in big cities. Due to the negative effect of these pollutants on the human health and environment, more and more stringent emission standards for automotive are applied. These emissions are now limited in number of particles per kilometer and the vehicle need to indicate when there is some failure of the systems of depollution (OBD: On-Board diagnostic).Resistive sensors have shown good results to measure soot particles mass concentration. They have advantage of being a simple and robust technology that can be easily manufactured at a cheap price. The sensor principle consists of conductance measurement between two platinum electrodes. Conductance increases with soot deposition. This work aims to define the key parameters that affect the sensors response. Three particles flow with different particles size distributions (centered at 90, 70 and 50 nm) were used and characterized. A fourth flow was used to see the impact of a lower mass concentration. The sensor response exposed to these different flows was studied. It was shown that the sensor sensibility and response times are optimal for a given polarization voltage between the electrodes whose value depends on the size distribution. This phenomenon was explained by the different electrical properties of the soot particles and modeled by equilibrium of soot accumulation and their combustion by Joule heating that permitting to simulate the sensor temporal response.

O objetivo desta tese é o estudo de formação de fuligem em chamas laminares de difusão. Para o modelo de formação de fuligem é escolhido um modelo semi-empírico de duas equações para prever a fração mássica de fuligem e o número de partículas de fuligem. O modelo descreve os processos de nucleação, de crescimento superficial e de oxidação das partículas. Para o modelo de radiação, a perda de calor por radiação térmica (gás e fuligem) é modelada considerando o modelo de gás cinza no limite de chama opticamente fina (OTA - Optically Thin Approximation). São avaliados diferentes modelos de cálculo das propriedades de transporte (detalhado e simplificado). Em relação à cinética química, tanto modelos detalhados quanto reduzidos são utilizados. No presente estudo, é explorada a técnica automática de redução conhecida como Flamelet Generated Manifold (FGM), sendo que esta técnica é capaz de resolver cinética química detalhada com tempos computacionais reduzidos. Para verificar o modelo de formação de fuligem foram realizados uma variedade de experimentos numéricos, desde chamas laminares unidimensionais adiabáticas de etileno em configuração tipo jatos opostos (counterflow) até chamas laminares bidimensionais com perda de calor de etileno em configuração tipo jato (coflow). Para testar a limitação do modelo os acoplamentos de massa e energia entre a fase sólida e a fase gasosa são investigados e quantificados para as chamas contra-corrente Os resultados mostraram que os termos de radiação da fase gasosa e sólida são os termos de maior importancia para as chamas estudas. Os termos de acoplamento adicionais (massa e propriedade termodinâmicas) são geralmente termos de efeitos de segunda ordem, mas a importância destes termos aumenta conforme a quantidade de fuligem aumenta. Como uma recomendação geral o acoplamento com todos os termos deve ser levado em conta somente quando a fração mássica de fuligem, YS, for igual ou superior a 0.008. Na sequência a formação de fuligem foi estudada em chamas bi-dimensionais de etileno em configuração jato laminar usando cinética química detalhada e explorando os efeitos de diferentes modelos de cálculo de propriedades de transporte. Foi encontrado novamente que os termos de radiação da fase gasosa e sólida são os termos de maior importância e uma primeira aproximação para resolver a chama bidimensional de jato laminar de etileno pode ser feita usando o modelo de transporte simplificado. Finalmente, o modelo de fuligem é implementado com a técnica de redução FGM e diferentes formas de armazenar as informações sobre o modelo de fuligem nas tabelas termoquímicas (manifold) são testadas A melhor opção testada neste trabalho é a de resolver todos os flamelets com as fases sólida e gasosa acopladas e armazenar as taxas de reação da fuligem por área de partícula no manifold. Nas simulações bidimensionais estas taxas são então recuperadas para resolver as equações adicionais de formação de fuligem. Os resultados mostraram uma boa concordância qualitativa entre as predições do FGM e da solução detalhada, mas a grande quantidade de fuligem no sistema ainda introduz alguns desafios para a obtenção de bons resultados quantitativos. Entretanto, este trabalho demonstrou o grande potencial do método FGM em predizer a formação de fuligem em chamas multidimensionais de difusão de etileno em tempos computacionais reduzidos. / The objective of this thesis is to study soot formation in laminar diffusion flames. For soot modeling, a semi-empirical two equation model is chosen for predicting soot mass fraction and number density. The model describes particle nucleation, surface growth and oxidation. For flame radiation, the radiant heat losses (gas and soot) is modelled by using the grey-gas approximation with Optically Thin Approximation (OTA). Different transport models (detailed or simplified) are evaluated. For the chemical kinetics, detailed and reduced approaches are employed. In the present work, the automatic reduction technique known as Flamelet Generated Manifold (FGM) is being explored. This reduction technique is able to deal with detailed kinetic mechanisms with reduced computational times. To assess the soot formation a variety of numerical experiments were done, from one-dimensional ethylene counterflow adiabatic flames to two-dimensional coflow ethylene flames with heat loss. In order to assess modeling limitations the mass and energy coupling between soot solid particles and gas-phase species are investigated and quantified for counterflow flames. It is found that the gas and soot radiation terms are of primary importance for flame simulations. The additional coupling terms (mass and thermodynamic properties) are generally a second order effect, but their importance increase as the soot amount increases As a general recommendation the full coupling should be taken into account only when the soot mass fraction, YS, is equal to or larger than 0.008. Then the simulation of soot is applied to two-dimensional ethylene co-flow flames with detailed chemical kinetics and explores the effect of different transport models on soot predictions. It is found that the gas and soot radiation terms are also of primary importance for flame simulations and that a first attempt to solve the two-dimensional ethylene co-flow flame can be done using a simplified transport model. Finally an implementation of the soot model with the FGM reduction technique is done and different forms for storing soot information in the manifold is explored. The best option tested in this work is to solve all flamelets with soot and gas-phase species in a coupled manner, and to store the soot rates in terms of specific surface area in the manifold. In the two-dimensional simulations, these soot rates are then retrieved to solve the additional equations for soot modeling. The results showed a good qualitative agreement between FGM solution and the detailed solution, but the high amount of soot in the system still imposes some challenges to obtain good quantitative results. Nevertheless, it was demonstrated the great potential of the method for predicting soot formation in multidimensional ethylene diffusion flames with reduced computational time.

In object oriented programming, a Call Graph represents the calling relationships between the program’s methods. To be more precise, a Call Graph is a rooted directed graph where each node of the graph represents a method and each edge (u, v) represents a method call from method u to method v. Focus of this thesis is on building a framework for Call Graph construction algorithms which can be used in program analysis. Our framework is able to be initialized by different front-ends and constructs various Call Graph algorithms. Here, we instantiate framework with two bytecode readers (ASM and Soot) as front-ends and implement three Call Graph construction algorithms (CHA, RTA and CTA). At first, we used two above mentioned bytecode readers to read the bytecode of a specific Java program, then we found reachable methods for each invoked method; meanwhile we kept obtained details on our own data structures.  Creating data structures for storing required information about Classes, Methods, Fields and Statements, gives us a great opportunity to implement an independent framework for applying well known Call Graph algorithms. As a result of these data structures, Call Graph construction will not depend on bytecode readers; since, whenever we read the bytecode of a program, we accumulate all necessary points in pre-defined data structures and implement our Call Graphs based on this accumulated data. Finally, the result is a framework for different Call Graph construction algorithms which is the goal of this thesis. We tested and evaluated the algorithms with a variety of programs as the benchmark and compared the bytecode readers besides the three Call Graph algorithms in different aspects.

Le travail au présent surligne la cérine nanostructurée et bien-définie ; une morphologie qui promeut une haute activité catalytique de la cérine dans l’oxydation des suies. Le travail présente également l’introduction des métaux pour le dopage, tels que praséodyme et zirconium, à la surface de la cérine pour améliorer la réductibilité, la stabilité thermique, et la capacité du stockage d’oxygène. L’oxydation à température programmée a été utilisée pour analyser l’activité catalytique. Au premier étage de la recherche, on a découvert que l’oxyde en mélange équimolaire de la cérine et de l’oxyde de praséodyme en nanostructure (indiqué comme Ce50Pr50-NP) possède la quantité d’espèces oxygénées à la surface la plus haute, la réductibilité la plus haute et l’activité catalytique la plus haute dans l’oxydation normale des suies. Il a été conclu que la nanostructure soulève la fonctionnalité du praséodyme dans la cérine. Le travail introduit également des nanoparticules (NPs) de Pt stabilisée par n-octylsilane. Pendant la calcination, les ligands silyliques se transforment aux « patches » de la silice qui leur évitent le frittage. Des NPs de Cu ont été préparées avec la même façon ; néanmoins elles ont souffert de sintering. Les NPs de Pt sont très actives dans l’oxydation de tous les polluants modèles des véhicules Diesel, spécifiquement l’oxydation des suies en présence de NOx, et elles fonctionnent mieux avec la cérine nanostructurée. Comme attendu, Ce50Pr50-NP donne l’activité catalytique plus haute que les catalyseurs à base du platine. La haute conversion du NO et l’adsorption du NO2 sur la surface sont la raison majeure de l’activité marquante / The present work highlights well-defined nanostructured ceria; a morphology that bestows exceptional catalytic activity on ceria towards soot oxidation. The work includes also introduction of promoting foreign metals, such as praseodymium and zirconium, to well-defined nanostructured ceria as a means of improving reducibility, thermal stability and oxygen storage capacity of the catalyst. Temperature-programmed oxidation (TPO) has been used for analyzing catalytic activity. At the first stage of the research, nanostructured equimolar ceria-praseodymia (denoted as Ce50Pr50-NP) was found to have the highest amount of surface oxygen, the highest reducibility and the highest catalytic activity towards soot oxidation. The nanostructured morphology has been proven to raise the functionality of praseodymia as the foreign metal in ceria. The work also introduces small, silane-stabilized Pt nanoparticles. Upon calcination, silyl ligands are transformed into siliceous patches that prevent the particle from migrating/coalescing. Cu nanoparticles have been prepared the same way as Pt nanoparticles; however, they sinter even under milder thermal treatment. The small Pt-NPs are proven active towards all pollutant oxidations, including NOx-assisted soot oxidation, and they function better with nanostructured ceria as the support. Unexpectedly, Ce50Pr50-NP gives higher activity towards NOx-assisted soot oxidation than Pt catalysts. Intense NO conversion and NO2 adsorption on the surface of Ce50Pr50-NP are the reason behind its high activity

Cette étude a pour finalité le développement d'une nouvelle famille de catalyseurs pour la combustion des suies Diesel afin de produire des filtres à particules (FAP) à régénération continue basse température. En effet, les régénérations périodiques des FAP actuellement commercialisés, engendrent une surconsommation plus ou moins élevée en carburant. Les catalyseurs étudiés sont des céramiques conductrices par les ions oxygènes et exempts de métal noble. L'ensemble de ces études a permis d'aboutir à plusieurs conclusions majeures. Les oxygènes actifs pour oxyder les particules de suies à basse température sont les oxygènes contenus dans le catalyseur. L'oxydation de la suie a donc lieu à l'interface solide/solide : suie/catalyseur. Un mécanisme de type électrochimique comme dans une pile à combustible mais à l'échelle nanométrique a été proposé : l'oxydation de la suie représente la réaction anodique qui se déroule aux points de contact suie / 8-YSZ, les électrons produits diffusent à travers les particules de suie vers les point triples entre les particules de suie (conductrices électroniques), la phase gaz (présence d'oxygène) et 8-YSZ (conducteur ionique) où se déroule la réaction cathodique d'incorporation de l'oxygène gazeux dans le matériau. Les paramètres clés qui gouvernent l'activité catalytique sont la surface de contact suie/catalyseur et donc la granulométrie de la poudre de catalyseur ainsi que la pression partielle d'oxygène dans la phase gaz et la mobilité de l'oxygène dans le catalyseur / This study aims to develop a new family of catalysts for diesel soot combustion to produce and optimize self-DPFs, based on ionic conducting ceramics, able to continuously burn soot particulates at low temperatures without fuel overconsumption and without the use of noble metals. The investigated catalysts are oxygen ionically conducting ceramics. Yttria stabilized Zirconia (8-YSZ containing 8 mol% of yttria) was chosen as the reference catalyst due to its high thermal and chemical stability and good ionic conductivity. A set of experiments was implemented to vary different parameters that can influence the reactivity of the reference catalyst. All of these studies have resulted in several major conclusions. Oxygen species active to oxidize soot particles at low temperature are those contained in the catalyst. An electrochemical type mechanism as in a fuel cell but at the nanoscale was proposed: the soot oxidation represents the anodic reaction which occurs at the contact points soot / 8-YSZ/O2 (gas) electrons are diffused through soot particles to triple points between the soot particles (electronic conductor), the gas phase (presence of oxygen) and 8-YSZ (ion conductor) where the cathodic reaction takes place with the incorporation of gaseous oxygen into the ceramic. The key parameters that influence the catalytic activity of 8-YSZ are soot / catalyst contact and thus the agglomerates size of the catalyst powder, the oxygen partial pressure in the gas phase and the mobility of oxygen in the catalyst

Dans une thématique de sécurité incendie, la prédiction de la propagation d’une flamme le long d’une paroi verticale solide nécessite la quantification du rayonnement. L’objectif de la thèse est de quantifier et caractériser les suies d’une flamme de PMMA. La flamme stationnaire d’un brûleur gaz sera également étudiée. La méthode optique de LII, une fois calibrée (extinction), permet d’obtenir des champs 2D de la fraction volumique ainsi que le taux de combustion en fonction de la hauteur pour la flamme de PMMA. Ensuite, la morphologie des agrégats de suie est déterminée grâce à des clichés MET et des mesures de la distribution en taille. Les propriétés optiques sont obtenues avec des mesures in-situ et ex-situ (effet de la température) de l’extinction spectrale, couplées avec une mesure de la concentration massique pour remonter à la fonction d’indice. Cette base de données permettra la perspective de calculer le rayonnement des suies et de l’utiliser dans un modèle de propagation. / In the field of fire safety, the propagation prediction of a flame along a solid vertical wall requires radiation quantification. The thesis objective is to quantify and characterize the PMMA flame soot. Gaseous burner steady flame will also be studied. The LII optical method, after calibration (extinction), provides 2D volume fraction fields and the combustion rate in function of the height for the PMMA flame. Then, the soot aggregate morphology is determined by TEM images and size distribution measurements. The optical properties are obtained with in-situ and ex-situ measurements (temperature effect) of the spectral extinction, coupled with a mass concentration measurement to obtain the index function. This database will allow the prospect to calculate the soot radiation and to use it in a propagation model.