Cette thèse est centrée autour du développement et de l'analyse des schémas volumes finis robustes afin d'approcher les solutions du modèle diphasique compressible en milieux poreux hétérogènes et anisotropes. Le modèle à deux phases compressibles comprend deux équations paraboliques dégénérées et couplées dont les variables principales sont la saturation du gaz et la pression globale. Ce système est discrétisé à l'aide de deux méthodes différentes (CVFE et DDFV) qui font partie de la famille des volumes finis. La première classe à laquelle on s'intéresse consiste à combiner la méthode des volumes finis et celle des éléments finis. Dans un premier temps, on considère un schéma volume finis upwind pour la partie convective et un schéma de type éléments finis conformes pour la diffusion capillaire. Sous l'hypothèse que les coefficients de transmissibilités sont positifs, on montre que la saturation vérifie le principe du maximum et on établit des estimations d'énergies permettant de démontrer la convergence du schéma. Dans un second temps, on a mis en place un schéma positif qui corrige le précédent. Ce schéma est basé sur une approximation des flux diffusifs par le schéma de Godunov. L'avantage est d'établir la bornitude des solutions approchées ainsi que les estimations uniformes sur les gradients discrets sans aucune contrainte ni sur le maillage ni sur la perméabilité. En utilisant des arguments classiques de compacité, on prouve rigoureusement la converge du schéma. Chaque schéma est validé par des simulations numériques qui montrent bien le comportement attendu d'une telle solution. Concernant la deuxième classe, on s'intéressera tout d'abord à la construction et à l'étude d'un nouveau schéma de type DDFV (Discrete Duality Finite Volume) pour une équation de diffusion non linéaire dégénérée. Cette méthode permet d' avantage de prendre en compte des maillages très généraux et des perméabilités quelconques. L'idée clé de cette discrétisation est d'approcher les flux dans la direction normale par un schéma centré et d'utiliser un schéma décentré dans la direction tangentielle. Par conséquent, on démontre que la solution approchée respecte les bornes physiques et on établit aussi des estimations d'énergie. La convergence du schéma est également établie. Des résultats numériques confirment bien ceux de la théorie. Ils exhibent en outre que la méthode est presque d'ordre deux. / The objective of this thesis is the development and the analysis of robust and consistent numerical schemes for the approximation of compressible two-phase flow models in anisotropic and heterogeneous porous media. A particular emphasis is set on the anisotropy together with the geometric complexity of the medium. The mathematical problem is given in a system of two degenerate and coupled parabolic equations whose main variables are the nonwetting saturation and the global pressure. In view of the difficulties manifested in the considered system, its cornerstone equations are approximated with two different classes of the finite volume family. The first class consists of combining finite elements and finite volumes. Based on standard assumptions on the space discretization and on the permeability tensor, a rigorous convergence analysis of the scheme is carried out thanks to classical arguments. To dispense with the underlined assumptions on the anisotropy ratio and on the mesh, the model has to be first formulated in the factional flux formulation. Moreover, the diffusive term is discretized by a Godunov-like scheme while the convective fluxes are approximated using an upwind technique. The resulting scheme preserves the physical ranges of the computed solution and satisfies the coercivity property. Hence, the convergence investigation holds. Numerical results show a satisfactory qualitative behavior of the scheme even if the medium of interest is anisotropic. The second class allows to consider more general meshes and tensors. It is about a new positive nonlinear discrete duality finite volume method. The main point is to approximate a part of the fluxes using a non standard technique. The application of this ideato a nonlinear diffusion equation yields surprising results. Indeed,not only is the discrete maximum property fulfilled but also the convergence of the scheme is established. Practically, the proposed method shows great promises since it provides a positivity-preserving and convergent scheme with optimal convergence rates.

Este trabalho desenvolveu um novo método para a detecção de contornos em imagens digitais que apresentam objetos de interesse muito próximos e que contêm complexidades associadas ao fundo da imagem como variação abrupta de intensidade e oscilação de iluminação. O método desenvolvido utiliza lógicafuzzy e desvio padrão da declividade (Desvio padrão da declividade fuzzy - FuzDec) para o processamento de imagens e detecção de contorno. A detecção de contornos é uma tarefa importante para estimar características de escoamento bifásico através da segmentação da imagem das bolhas para obtenção de parâmetros como a fração de vazio e diâmetro de bolhas. FuzDec foi aplicado em imagens de instabilidades de circulação natural adquiridas experimentalmente. A aquisição das imagens foi feita utilizando o Circuito de Circulação Natural (CCN) do Instituto de Pesquisas Energéticas e Nucleares (IPEN). Este circuito é completamente constituído de tubos de vidro, o que permite a visualização e imageamento do escoamento monofásico e bifásico nos ciclos de circulação natural sob baixa pressão.Os resultados mostraram que o detector proposto conseguiu melhorar a identificação do contorno eficientemente em comparação aos detectores de contorno clássicos, sem a necessidade de fazer uso de algoritmos de suavização e sem intervenção humana. / This work has developed a new method for digital image contour detection which can be successfully applied to images presenting objects of interest with high proximity and presenting complexities related with background abrupt intensity fluctuations. The developed method makes use of fuzzy logic and declivity standard deviation (Fuzzy Declivity Standard Deviation FuzDec) to image processing and contour detection. Contour detection is an important task to estimate two-phase flow features through bubble segmentation in order to obtain parameters as void fraction and bubble diameter. FuzDec was applied to natural circulation instability images which were experimentally acquired. Image acquisition was done at the Natural Circulation Circuit (CCN) of the Instituto de Pesquisas Energéticas e Nucleares (IPEN) in Brazil. This facility is all made up with glass tubes allowing easy visualization and imaging of one-phase and two-phase flow patterns during natural circulation cycles under low pressures. Results confirm that the proposed detector can improve contour identification when compared to classical contour detector algorithms, without using smoothing algorithms or human intervention.

O escoamento estratificado óleo-água é bastante comum na indústria do petróleo, especialmente em poços direcionas offshore, oleodutos e gasodutos. Entretanto, existem poucos trabalhos na literatura sobre a natureza da estrutura ondulatória observada no escoamento em dutos ascendentes ou descendentes a partir da horizontal. O objetivo deste trabalho foi estudar as propriedades geométricas e cinemáticas da onda interfacial, i.e, forma média, comprimento, amplitude e celeridade, e assim contribuir para a compreensão do papel da onda interfacial na dinâmica do escoamento estratificado. Um software baseado em plataforma Labview® possibilitou a automação para obtenção dos dados das ondas interfaciais extraídas de imagens de vídeos de alta resolução. Além das propriedades das ondas, também se coletaram valores de fração volumétrica in situ e de gradiente de pressão bifásico para cinco ângulos de inclinação (-20°, -10°, 0°, 10° e 20°) em diferentes pares de vazões de óleo e água. Desenvolveu-se um modelo fenomenológico considerando os termos ondulatórios do escoamento para o cálculo da fração volumétrica in situ e do gradiente de pressão bifásico, sendo comparado com modelos disponíveis na literatura e dados experimentais. A concordância do modelo proposto com os dados coletados neste trabalho se mostrou muito boa, o que sugere um avanço em comparação ao existente na literatura. O estudo da equação da onda de perturbação interfacial para o escoamento estratificado óleo-água indicou que a natureza da onda observada é cinemática e não dinâmica; e baseado na equação da celeridade da onda cinemática pode-se confrontar a celeridade experimental com a teórica, revelando boa concordância. / The oil-water stratified flow is quite common in the oil industry, especially in offshore directional wells and pipelines. However, there are few studies on the physics of the wavy structure observed in upward and downward stratified flow. The goal of this work was to study the geometric and kinematic properties of interfacial waves, i.e., the average shape, wavelength, amplitude and celerity. A homemade Labview®-based software enabled the automatic acquisition of data extracted from frames obtained via high resolution video recording. In situ volume fraction and two-phase pressure gradient data for five inclination angles (-20°, -10°, 0°, 10° and 20°) at several pairs of oil and water flow rates were also collected. A phenomenological model that takes into account the wavy structure is proposed to calculate volume fractions and two-phase pressure gradient and it was compared with available models from the literature and experimental data. The good agreement of the proposed model with the data collected in this study is promising and suggests that it may provide better predictions in comparison with models from the literature. The study of the interfacial perturbation wave equation for stratified flow indicates that the observed waves nature is kinematic and not dynamic; and based on the kinematic wave velocity equation we could compare the experimental celerity with the theoretical one, with good agreement.

O escoamento bifásico é muito importante para vários ramos industriais. As misturas bifásicas podem escoar em diversas configurações, as quais são chamadas de padrões de escoamento, e que, ao longo dos anos, receberam diversas classificações. Neste trabalho foi estudado o padrão bolhas dispersas em duto anular. Os trabalhos sobre escoamento bifásico em geometria anular são mais escassos, especialmente quando se trata de dutos anulares de grande dimensão. Dentre os modelos existentes para a modelagem do escoamento bifásico, trabalhamos com o modelo de mistura (Drift Flux), pois, apesar das altas velocidades superficiais, o escoamento em bolhas não possui um comportamento homogêneo, já que a fase gasosa ainda escoa na região central e com uma velocidade maior do que a velocidade da mistura. Foi utilizado o modelo de mistura unidimensional para a modelagem do escoamento bifásico água-ar, pois, além de bastante preciso, é de baixo custo computacional e fácil implementação. Neste trabalho empregamos, a princípio, equações elaboradas para dutos circulares, as quais foram adaptadas para a geometria anular fazendo-se uso do conceito de diâmetro hidráulico. As equações foram implementadas no software Mathematica® e as previsões para o holdup e queda de pressão foram comparadas com dados experimentais próprios. O trabalho experimental foi realizado no laboratório de escoamentos multifásicos do Núcleo de Engenharia Térmica e de Fluidos (NETeF) da EESC-USP, o qual conta com uma instalação experimental em estado operacional para a simulação de escoamentos bifásicos vertical e inclinados em duto anular. Foram colhidas medidas de holdup e queda de pressão bifásica. A concordância entre os dados experimentais e as previsões foi muito satisfatória. / Two-phase flow is present in a wide range of industrial processes. Such flows occur in many geometrical configurations known as flow patterns. In this work is presented a study about bubbly flow in annular duct. Works on two-phase flow in large annular geometry are very scarce. Among the existing models for modeling two-phase flow we chose the drift flux model because the bubbly flow does not behave as a homogeneous mixture, despite of the high superficial velocities, since the gas phase flows faster in the core region in comparison with the mixture velocity. The onedimensional drift flux model besides to be accurate is of low computational cost and easy implementation. The goal of this work is to analyze the prediction of holdup and pressure drop in annular ducts of large diameter. The set of equations used was adapted from those of circular ducts by using the hydraulic diameter concept. The equations were implemented in Mathematica software and holdup and pressure drop prediction were compared with experimental data taken in the experimental facility of the Thermal-fluids Engineering Laboratory of EESC-USP. It was collected measure of holdup and pressure drop. The agreement between experimental data and model predictions is encouraging.

A Agência Internacional de Energia Atômica (IAEA - \"International Atomic Energy Agency\") vem incentivando o desenvolvimento de sistemas passivos de refrigeração em plantas nucleares visando a simplificação e o incremento da confiabilidade em funções essenciais de segurança nos projetos de uma próxima geração de reatores nucleares refrigerados a água. O principal fundamento desses sistemas é o emprego da circulação natural como sistema de segurança aplicável em operações de desligamento do reator para manutenção ou na ocorrência de acidentes. A circulação natural é um fenômeno que surge em virtude do gradiente de temperatura em pontos diferentes do circuito de refrigeração. Em condições extremas de estabilidade têm-se o estabelecimento do escoamento bifásico gás/líquido podendo configurar-se segundo diferentes regimes. A fração de vazio é reconhecida como um dos parâmetros chave na predição da ocorrência de instabilidades do escoamento bifásico. Apresenta-se neste trabalho uma inovadora metodologia para estimativa da fração de vazio a partir de imagens digitais capturadas diretamente de circuitos experimentais que geram o escoamento bifásico. O método é baseado na aquisição de imagens, com controle da profundidade de campo, de uma seção do Circuito de Circulação Natural (CCN) presente no IPEN/CNEN-SP. A imagem é segmentada com base na inferência fuzzy de diferentes parâmetros de segmentação e ajustada ao foco utilizado na sua aquisição. Ela é varrida de um modo inédito e iterativo, utilizando máscaras de diferentes tamanhos integrando um conjunto de redes neurais com a Transformada Randomizada de Hough. Cada diferente tamanho de máscara é escolhido de acordo com os tamanhos das bolhas que são os objetos de interesse. O volume da bolha é estimado baseado em sua projeção plana capturada nas imagens digitais. O cálculo da fração de vazio considera o volume da seção geométrica do escoamento no tubo de vidro cilíndrico e a profundidade de campo utilizada e nos parâmetros geométricos inferidos para cada bolha detectada. Os resultados mostraram que a integração entre o conjunto de redes neurais e a Transformada Randomizada de Hough aumentaram a robustez das estimativas do sistema. / The International Atomic Energy Agency (IAEA) has been encouraging the use of passive cooling systems in new designs of nuclear power plants. Next nuclear reactor generations are intended to possess simpler and robust safety functions. Natural circulation based systems hold an undoubtedly prominent position among these. Natural circulation phenomenon occurrence depends only on the existence of refrigerant liquid temperature gradient in different sections of the plant refrigerator circuit. The study of limit conditions for these systems has led to instability behavior analysis where many different two-phase flow patterns are present. Void fraction is a key parameter in thermal transfer analysis of theses flow instability conditions. This works presents a new method to estimate void fraction from digital images captured at an experimental two-phase flow circuit. The method is primarily based on depth-of-field controlled image acquisition of a section of a closed loop of natural circulating water through cylindrical glass tubes. This loop is called Natural Circulation Facility (NCF) and is located at Nuclear and Engineering Research Institute in Brazil (IPEN/CNEN-SP). Image is segmented based on fuzzy inference of different segmentation parameters and adjusted to image acquisition focus. The image is then scanned in an inedited way using different-sized masks integrating a set of different artificial neural networks with a modified Randomized Hough Transform. Each different mask size is chosen in accordance to bubble sizes which are objects of interest. The bubble volume is estimated based on two-dimensional projection sizing based on digitally acquired images. Void fraction calculation takes into account the volume of the geometrical section of flow inside cylindrical glass tube considering used depth-of-field. It is also based on the summed bubble geometrical parameters inferred for each detected bubble. The results have shown that integration between artificial-neural-net sets and Randomized Hough Transforms increase system estimations robustness.

Este trabalho desenvolveu um novo método para a detecção de contornos em imagens digitais que apresentam objetos de interesse muito próximos e que contêm complexidades associadas ao fundo da imagem como variação abrupta de intensidade e oscilação de iluminação. O método desenvolvido utiliza lógicafuzzy e desvio padrão da declividade (Desvio padrão da declividade fuzzy - FuzDec) para o processamento de imagens e detecção de contorno. A detecção de contornos é uma tarefa importante para estimar características de escoamento bifásico através da segmentação da imagem das bolhas para obtenção de parâmetros como a fração de vazio e diâmetro de bolhas. FuzDec foi aplicado em imagens de instabilidades de circulação natural adquiridas experimentalmente. A aquisição das imagens foi feita utilizando o Circuito de Circulação Natural (CCN) do Instituto de Pesquisas Energéticas e Nucleares (IPEN). Este circuito é completamente constituído de tubos de vidro, o que permite a visualização e imageamento do escoamento monofásico e bifásico nos ciclos de circulação natural sob baixa pressão.Os resultados mostraram que o detector proposto conseguiu melhorar a identificação do contorno eficientemente em comparação aos detectores de contorno clássicos, sem a necessidade de fazer uso de algoritmos de suavização e sem intervenção humana. / This work has developed a new method for digital image contour detection which can be successfully applied to images presenting objects of interest with high proximity and presenting complexities related with background abrupt intensity fluctuations. The developed method makes use of fuzzy logic and declivity standard deviation (Fuzzy Declivity Standard Deviation FuzDec) to image processing and contour detection. Contour detection is an important task to estimate two-phase flow features through bubble segmentation in order to obtain parameters as void fraction and bubble diameter. FuzDec was applied to natural circulation instability images which were experimentally acquired. Image acquisition was done at the Natural Circulation Circuit (CCN) of the Instituto de Pesquisas Energéticas e Nucleares (IPEN) in Brazil. This facility is all made up with glass tubes allowing easy visualization and imaging of one-phase and two-phase flow patterns during natural circulation cycles under low pressures. Results confirm that the proposed detector can improve contour identification when compared to classical contour detector algorithms, without using smoothing algorithms or human intervention.

A presente dissertação trata de um estudo teórico-experimental sobre a perda de pressão em canais de diâmetro reduzido durante escoamento bifásico de refrigerantes halogenados. Trocadores de calor baseados na ebulição convectiva, em condições de micro-escala são amplamente estudados devido à intensificação da troca de calor proporcionada e a possibilidade de compactação de sistemas de resfriamento. Proporcionam também a redução do inventário de refrigerante e do material utilizado no processo de fabricação do trocador. Porém, o incremento da transferência de calor é acompanhada pelo aumento da perda de pressão, parâmetro também fundamental para o desempenho do sistema. Para o projeto satisfatório e otimizado destes dispositivos são necessários métodos de previsão de transferência de calor e perda de pressão. Entretanto, no caso de canais de diâmetro reduzido, tais ferramentas não encontram-se disponíveis e trocadores de calor baseados em escoamentos bifásicos no interior de canais de diâmetro reduzido vêm sendo desenvolvidos heuristicamente. Desta forma, inicialmente neste estudo, realizou-se uma revisão crítica da literatura envolvendo critérios de transição entre padrões de escoamento, fração de vazio superficial, perda de pressão no interior de canais com diâmetro reduzido durante escoamento bifásico e os principais métodos de estimativa da perda de pressão para macro e micro-escala. Resultados experimentais para perda de pressão levantados neste estudo em condições adiabáticas para os fluidos R245fa e R134a e tubo com 1,1 mm de diâmetro interno foram descritos e comparados aos métodos preditivos encontrados na literatura. Finalmente um novo método da previsão da perda de pressão foi proposto baseado na correlação de Müller-Steinhagen e Heck (1986), ajustando os valores do coeficiente e do expoente com base nos resultados experimentais levantados. / A theorical and experimental study on two-phase pressure drop inside micro-scale channels has been developed. Recently, the study of flow boiling in micro-scale channel have received special attention from academia and industry due to several advantages that they offer such as minimization of fluid inventory, high degree of compactness of the heat exchangers, better performance and the capacity of dissipate extremely high heat fluxes. The significant heat transfer coefficient enhancement provided by micro-scale channels comes together with a huge pressure drop penalty that impacts the efficiency of the overall cooling system. So, accurate predictive methods to evaluate the pressure drop are necessary for the appropriate design of the system and for its optimization. In the present study, firstly, a critical review on studies from literature was performed that covers criteria of transition between micro- and macro-scale flow boiling, void fraction, frictional pressure drop on micro-scale channels and the leading frictional pressure drop predictive methods. Experimental pressure drop results were acquired under adiabatic conditions for R245fa and R134a fluids and internal diameter tube of 1.1 mm. Then, the leading pressure drop predictive methods were compared against the present database. Also a new correlation based on Muller-Steinhagen e Heck (1986) method was proposed in this work by adjusting new empirical constants based on the present database together with previous results obtained by Tibiriçá et al. (2011) for a 2.3 mm ID tube.

Ce travail est consacré à l’analyse et au développement de concepts et d’outils mathématiques en vue de leur application à des problématiques propres aux sites de stockage géologique profond de déchets radioactifs. La première partie porte sur l’estimation en champ lointain de la concentration de radionucléides issus du relâchement des colis de confinement, lorsque les incertitudes sur le relâchement sont prises en compte. En s’appuyant sur les travaux de A. Bourgeat et A. Piatniski sur l’homogénéisation d’une équation de convection-diffusion avec second membre aléatoire, on développe des outils numériques permettant d’approcher le comportement probabiliste du champ de concentration dans une configuration du type site de stockage. Dans une seconde partie, on s’intéresse à la migration de gaz dans et autour d’un site de stockage. Après une revue sur la modélisation physique des écoulements diphasiques de type eau/hydrogène en milieu poreux, on propose une nouvelle formulation mathématique du problème qui décrit, dans un même jeu d’équations, les écoulements à une (liquide) et deux (liquide/gaz) phases. Une étude de l’existence de solutions de cette formulation est menée à l’aide de la théorie générale des équations différentielles quasilinéaires elliptiques-paraboliques introduite par H.W. Alt et S. Luckhaus. Une méthode de résolution numérique du problème est mise en oeuvre pour la simulation de différents cas test, des plus simples au plus représentatif d’un site de stockage géologique. Enfin, l’homogénéisation périodique du modèle est effectuée et appliquée à la simulation de l’exercice Couplex-Gaz proposé par l’ANDRA / The purpose of this work is to analyze and develop mathematical concepts and tools in application to performance assessment of an underground nuclear waste disposal. The first part is concerned with estimating the far field concentration of radionuclides released by containers of waste when uncertainties on the release are taking in account. Using the work of A. Bourgeat and A. Piatniski about homogenization of a convection-diffusion equation with random source term, numerical tools are developed to approximate the random behavior of the concentration field in an underground disposal configuration. In a second part, we are interested in gas migration in and around an underground nuclear waste disposal. After a review on physical models of two-phase flow in porous media for water/hydrogen mixture, we propose a new mathematical formulation describing one- (liquid) and two- (liquid/gas) phase flow with a unique set of equation. Considering the general theory of quasilinear elliptic-parabolic differential equations introduced by H.W. Alt and S. Luckhaus, we study existence of solutions for this formulation. A numerical method to solve the problem is implemented to simulate several test cases. These test cases run from very simple situations to a representative configuration of an underground nuclear waste disposal. Finally, the periodic homogenization of the model is done and applied to simulate the Couplex-Gas exercise proposed by ANDRA.

Cette thèse propose un algorithme efficace pour la simulation numérique des interactions houle-structure avec des solveurs CFD bi-phasiques. L'algorithme est basé sur le couplage de la théorie potentielle et des équations bi-phasiques de Navier-Stokes. C'est une extension de la méthode Spectral Wave Explicit Navier-Stokes Equations (SWENSE) pour les solveurs CFD bi-phasiques avec une technique de capture d'interface. Dans cet algorithme, la solution totale est décomposée en une composante incidente et une composante complémentaire. La partie incidente est explicitement obtenue avec des méthodes spectrales basées sur la théorie des écoulements potentiels ; seule la partie complémentaire est résolue avec des solveurs CFD, représentant l'influence de la structure sur les houles incidentes. La décomposition assure la précision de la cinématique des houles incidentes quel que soit le maillage utilisé parles solveurs CFD. Une réduction significative de la taille du maillage est attendue dans les problèmes typiques des interactions houle structure. Les équations sont présentées sous trois formes : la forme conservative, la forme non conservative et la forme Ghost of Fluid Method. Les trois versions d'équations sont implémentées dans OpenFOAM et validées par une série de cas de test. Une technique d'interpolation efficace pour reconstruire la solution des houles irrégulières donnée par la méthode Higher-Order Spectral (HOS) sur le maillage CFD est également proposée. / This thesis proposes an efficient algorithm for simulating wave-structure interaction with two-phase Computational Fluid Dynamics (CFD) solvers. The algorithm is based on the coupling of potential wave theory and two phase Navier-Stokes equations. It is an extension of the Spectral Wave Explicit Navier-Stokes Equations (SWENSE) method for generalized two-phase CFD solvers with interface capturing techniques. In this algorithm, the total solution isdecomposed into an incident and acomplementary component. The incident solution is explicitly obtained with spectral wave models based on potential flow theory; only the complementary solution is solved with CFD solvers, representing the influence of the structure on the incident waves. The decomposition ensures the accuracy of the incident wave’s kinematics regardless of the mesh in CFD solvers. A significant reduction of the mesh size is expected in typical wave structure interaction problems. The governing equations are given in three forms: the conservative form, the non conservative form, and the Ghost of Fluid Method (GFM) form. The three sets of governing equations are implemented in OpenFOAM and validated by a series of wave-structure interaction cases. An efficient interpolation technique to map the irregular wave solution from a Higher-Order Spectral (HOS) Method onto the CFD grid is also proposed.

Le rayonnement dans les jets issus de moteurs à propergol solide constitue un phénomène essentiel à l’estimation des flux aux parois et à la prédiction de la signature radiative des engins. A haute altitude, de l’ordre de 100 km, ces jets sont caractérisés par des écoulements compressibles diphasiques, à fort aspect raréfié dans certaines régions, compos ´es de particules d’alumine et de gaz de combustion. Le transfert radiatif y joue un rôle important dans la mesure où il influence fortement le refroidissement et le changement de phase des particules. Afin de simuler numériquement les jets et leur rayonnement, différents modèles ont été développés. Le rayonnement des gaz a été pris en compte à l’aide de modèles statistiques à bandes étroites. Le phénomène de surfusion qui régit le changement de phase de l’alumine et les champs de température associés aux différentes tailles de particules, a été pris en compte. Enfin, une méthode de splitting des puissances radiatives a été mise en œuvre afin de permettre le couplage entre le rayonnement et l’écoulement dans des milieux en des ‘équilibre thermique gaz/particules. Ces modèles ont été implémentes dans une plateforme de calcul, permettant de coupler un solveur fluide utilisant une approche Navier-Stokes, un solveur eulérien pour traiter la phase dispersée et un solveur radiatif qui utilise une méthode de Monte Carlo. L’outil numérique développe a été partiellement validé en comparant nos résultats aux mesures obtenues dans le cadre de l’expérience BSUV2. Dans les conditions de cette expérience, le rayonnement des particules est prédominant mais la contribution des gaz s’avère non négligeable. Des simulations sous différentes hypothèses ont permis de mettre en évidence le rôle primordial du transfert radiatif, couplé au phénomène de surfusion, dans l’établissement des champs de température des particules. La dernière partie de ces travaux s’est attachée à l’étude du déséquilibre vibrationnel de la phase gazeuse et de son impact sur le rayonnement dans les jets. Il est montré que le gel partiel des niveaux de vibration de la molécule CO2 durant la détente du jet peut augmenter de façon significative son rayonnement. / Radiation from solid propellant rocketplumes is important for the prediction of thermalfluxes on vehicle walls and of plume signature. Athigh altitudes, of approximately 100 km, those plumesare characterized by two-phase compressible flows,highly rarefied in some regions, composed of aluminaparticles and exhaust combustion gases. Radiativetransfer plays an important role in the cooling and thephase change of the particles.In order to carry out numerical simulations of rocketplumes and their radiation, several models have beendeveloped. The radiation of the gas phase is takeninto account using statistical narrow bands models.The supercooling phenomenon has been modeled todeal with the phase change of alumina and to obtaincorrect temperature fields for the different size classesof particles. Finally, a splitting method of the radiativepower has been established to enable the couplingbetween radiation and the flow field under gas/particlethermal non-equelibrium. These models have beenimplemented in a calculation platform, enabling tocouple a Navier-Stokes solver for the gas phase, anEulerian solver dealing with the dispersed phase anda radiative solver based on a Monte Carlo method.The developed numerical tool has been partly validatedcomparing our results with the measurementsobtained during the BSUV2 experiment. In the conditionsof this experiment, particle radiation is shownto be predominant but the contribution of the gasphase is found to be non-negligible. Simulations underdifferent hypotheses have put the emphasis onthe importance of radiative transfer, coupled with thesupercooling phenomenon, for an accurate evaluationof particle temperature fields.The last part of this work focuses on the study ofgas vibrational non-equilibrium and its impact on radiationfrom high altitude plumes. It is shown thatthe slow deexcitation of vibrational levels of the CO2molecule during the plume expansion may increasesignificantly its radiation.