Uma análise dos processos colisionais homonucleares e heteronucleares, responsáveis por perdas em uma armadilha magneto-óptica, foi realizada neste trabalho. Resultados experimentais inéditos para as taxas de perdas, no sistema Rb-Cs, foram obtidos. Os dados experimentais em conjunto com os resultados do modelo sugerem que o processo de escape radiativo é dominante. O modelo utilizado é muito sensível à profundidade da armadilha, a qual depende de outros parâmetros (intensidade, dessintonia, gradiente de campo magnético, etc.). A colaboração com pesquisadores italianos possibilitou uma análise mais detalhada das taxas de perdas heteronucleares em outras regiões de intensidade. Estudamos a dependência das taxas heteronucleares com a razão das massas do par atômico, e uma razoável concordância foi observada. O estudo de processos colisionais em armadilhas magneto-ópticas é importante na obtenção do condensado de BoseEinstein com duas espécies atômicas distintas, e também em experiências de espectroscopia fotoassociativa. / In this work, we had investigated the heteronuclear trap loss rate in a two-species Rb-Cs magneto-optical trap. The experimental results suggest that radiative escape is the main collisional process responsible for heteronuclear losses. An addapted Gallagher - Pritchard model is compared with the data. The model is very sensitive to the trap depth, which depends on the trap parameters (intensity, detuning, magnetic - field gradient, etc.) This observation is also supported by experimental results from a Pisa\' group. We have compared experimental of heteronuclear rates as a function of the masses ratio of the atomic pair with the model, and a reasonable agreement is observed. These studies are relevant for high resolution atomic / molecular spectroscopy, and for the production of mixed - species Bose-Einstein condensates.

O presente trabalho é um estudo para a determinação de índices de ajuste em medidas efetuadas por sensores de temperatura do ar utilizadas em algoritmos de controle do sistema de automação do sistema de condicionamento de ar do tipo \"Teto Frio\". É um trabalho experimental baseado na coleta de dados de diversas variáveis ambientais em um laboratório que utiliza este tipo de sistema. Esse laboratório possui um sistema de climatização misto composto por circuitos de água gelada que percorre serpentinas instaladas em placas metálicas de forro, para troca de calor por radiação (retirada de calor sensível) e circuitos de ar, com distribuição de ar pelo teto, para troca de calor por convecção (retirada de calor latente). Seu funcionamento é controlado por um sistema de automação (hardware e software) específico. Este estudo pretende determinar índices de ajuste que, introduzidos nos algoritmos de controle do sistema de automação, corrijam as leituras de temperatura do ar realizadas no ambiente por sensores do sistema posicionados em locais que, embora esteticamente sejam os mais adequados, funcionalmente, não o são. Os resultados, concluídos através de método estatístico e recursos de representação gráfica, promoverão além do melhor entendimento das condições térmicas do ambiente oferecidas pelo sistema de \"teto frio\", a otimização da operação do sistema, isto é, a melhoria do conforto para os ocupantes e a conservação de energia. / The present work is a study to determine adjustment indexes in measurements made by air temperature sensors used in control algorithms of automation system applied to \"Cooling Ceiling\" air conditioning system. It is an experimental work based on data collection of several environmental variables at a laboratory which uses this kind of system. This laboratory owns a mixed air conditioning system compounded by chilled water circuits which pass through coils installed on metal ceiling tiles, for heat exchange by radiation (sensitive heat withdrawal), and air circuits with distribution by ceiling, for heat exchange by convection (latent heat withdrawal). Its functioning is controlled by a specific automation system (hardware and software). This study intends to determine adjustment indexes which, added in system control algorithms correct the air temperature readings accomplished in the environment by system sensors located in places that, although aesthetically are more suited, functionally are not. The results, concluded through statistic methods, and graphic representation researches, will promote beyond better understanding of thermal conditions of environment offered by this cooling ceiling system, optimization of system operation, that are a better comfort for the occupants and the energy saving.

Este trabalho apresenta e consolida uma metodologia alternativa para a determinação de taxas de reação nuclear ao longo da direção radial das pastilhas combustíveis sem necessidade de intensos fluxos neutrônicos. Esta técnica se baseia na irradiação de um disco de UO2 inserido no interior de uma vareta combustível desmontável no núcleo do reator IPEN/MB-01. Após a irradiação são realizadas várias espectrometrias gama do disco utilizando um detector HPGe alternando sequencialmente 6 colimadores de chumbo com diâmetros diferentes. Consequentemente, as reações nucleares de captura radiativa que ocorrem nos átomos de 238U, juntamente com as fissões que ocorrem em ambos 235U e 238U são mensuradas em função de 6 regiões radiais distintas do disco combustível. As correções de eficiência geométrica devido à introdução dos colimadores no sistema de detecção HPGe são determinadas através do código MCNP-5. As medidas de taxa de fissão são realizadas utilizando o 99Mo como radionuclídeo traçador. Esse radionuclídeo foi estudado e provou-se ideal para estas medidas por possuir um comportamento linear de formação, alto rendimento de fissão e principalmente por emitir fótons de baixa energia. As medidas foram efetuadas irradiando discos de UO2 (com enriquecimento de 4,3%) na posição central do reator IPEN/MB-01 a potência de 100 Watts durante uma hora. Algumas medidas foram realizadas utilizando uma luva de cádmio envolta na vareta combustível para determinar as taxas de reação nuclear na faixa de energia epitérmica. Os resultados experimentais obtidos são comparados a cálculos de taxa de reação nuclear via MCNP-5 utilizando a biblioteca de dados ENDF/B-VII.0, os quais apresentaram discrepâncias de no máximo 9% para as taxas de captura no 238U e 14% para as taxas de fissão no U na faixa epitérmica. Foram obtidos valores máximos de 4,5% para incertezas relativas as taxas de captura total e epitérmica e para as taxas de fissão total e epitérmica valores máximos de 11,3%. / This research presents and consolidates an alternative methodology for determining nuclear reaction rates along the radial direction of the fuel pellets which does not require high neutron flux. This technique is based on irradiating a thin UO2 disk inserted into a removable fuel rod at the IPEN/MB-01 reactor core. Several gamma spectrometries are performed after irradiation using a HPGe detector. Six lead collimators with different diameters are sequentially alternated during this process, thus, the nuclear radioactive capture which occurs in 238U and the fissions which occur in both 235U and 238U are measured according to six different radial regions of the fuel disk. Geometric efficiency corrections due to the introduction of collimators in HPGe detection system are determined by MCNP-5 code. The fission rate measurements are performed using the 99Mo. This radionuclide was studied and proved ideal for these measurements because it is formed in linear behavior in the reactor core, have a high yield fission and emits low-energy photons. Measurements were performed irradiating UO2 disks (with 4.3% enrichment) in the central position of the IPEN/MB-01 core at 100 watts power level during one hour. Some measurements were performed using a cadmium glove wrapped in the fuel rod to determine the nuclear reaction rates in the epithermal energy range. The experimental results obtained are compared with nuclear reaction rate calculations by means of MCNP-5 with ENDF/B-VII.0 data library showing discrepancies of up to 9% in 238U capture rates and 14% for U fission rates for epithermal energies. Uncertainties regarding the nuclear capture rates have maximum values of 4.5% and the fission rates has maximum values of 11.3%.

A avaliação dos efeitos dos aerossóis em relação ao balanço radiativo local e global bem como o impacto sobre a saúde humana, principalmente em grandes centros urbanos, demanda que se conheça de forma precisa a sua concentração e distribuição espaço-temporal. Neste contexto, o monitoramento acurado de longo prazo por estações instaladas ao redor do mundo tornou-se um desafio e uma necessidade para várias áreas do conhecimento. Classicamente as estações utilizam a técnica de fotometria solar para inferir a concentração e a distribuição de tamanho dos aerossóis através da profundidade óptica do aerossol (POA) e do coeficiente de Ångström . No presente trabalho é sugerida uma metodologia para estimar POA e do coeficiente de Ångström através de quatro canais espectrais de radiômetros do tipo Multi-Filter Rotating Shadowband Radiometer (MFRSR) que operam em São Paulo desde 1999. Estas estimativas foram avaliadas em relação à rede AErosol RObotic NETwork (AERONET) entre os anos de 2004 a 2006. Os resultados mostram que, em termos de variabilidade temporal da POA, há boa concordância entre os diferentes instrumentos. Entretanto, são encontradas diferenças médias sistemáticas da ordem de 0,03 na magnitude da POA em três dos quatro canais analisados, enquanto que o valor sugerido pela OMM para uma atmosfera limpa é de no máximo 0,02. Este resultado aponta que a metodologia empregada nos MFRSR fornece valores de boa qualidade. Ainda neste trabalho, foram relacionadas variáveis meteorológicas coletadas pela Estação Meteorológica do IAG-USP com a POA e o do coeficiente de Ångström estimados neste trabalho. Observou-se uma relação linear entre o coeficiente de Ångström e a umidade relativa (UR), indicando um crescimento do tamanho médio dos aerossóis que integram um grupo com 0,20<POA<=0,60 no canal 415nm com o aumento da umidade relativa. Acredita-se que este grupo tenha predominância de aerossóis de fontes locais. As medidas de longo prazo permitiram também caracterizar, em conjunto com estimativas espaciais de POA pelo sensor Moderate Resolution Imaging Spectroradiometer (MODIS) a bordo do satélite Terra, a influência do transporte de aerossóis de queimadas da região amazônica e do Brasil central em São Paulo, na série de dados analisados. Estes eventos contribuem para o aumento significativo da POA sobre a região metropolitana e trazem partículas ligeiramente menores que as emitidas por fontes locais. / The evaluation of the aerosol effects to the local and global radiation budget, as well as the impact on human health, particularly in large urban centers, demands knowing accurately their concentration and spatial-temporal distribution. In this context, the accurate long term monitoring from ground based stations installed around the world has become a challenge and a necessity for various areas of knowledge. Classically, the stations use the technique of Sun photometry to infer the concentration and size distribution of aerosols through the aerosol optical depth (AOD) and Ångström coefficient . In this paper we suggest a methodology to estimate AOD and Ångström coefficient through four channels of Multi-Filter Rotating Shadowband Radiometers (MFRSR) operating in Sao Paulo since 1999. These estimates were compared to the results of the AErosol RObotic NETwork (AERONET) between the years of 2004 to 2006. The results showed that the MFRSR can represent well the temporal variability of the AOD, but systematic differences were found with mean values of about 0.03 in AOD, in three of the four analyzed channels. The value suggested by WMO for clean air is at most 0.02. This result indicates that the methodology employed in MFRSR provides values of quality good. Also in this study, meteorological influences on the aerosol optical properties were analyzed. The meteorological variables were monitored at the Meteorological Station of the IAG-USP There was a linear relationship between Ångström coefficient and relative humidity (RH), indicating a growth of the average size of aerosols within a group with 0.20<AOD<=0.60, in the channel 415 nm, with increasing relative humidity. It is believed that this group has predominance of aerosols from local sources. Long term measurements also helped to characterize, together with spatial estimates of AOD by the Moderate Resolution Imaging Spectroradiometer (MODIS) on board the Terra satellite, the influence of the transport of aerosol particles from biomass burning from the Amazon region to São Paulo. It was observed that these events contribute to significant increase in AOD on the metropolitan region and bring slightly smaller particles than the emitted by local sources.

La présence des particules d’aérosols dans l’atmosphère influencent le bilan radiatif de notre planète et ainsi son équilibre climatique. Selon les différents mécanismes d’émission et processus de transports atmosphériques, les aérosols peuvent être entrainés dans la troposphère libre et ainsi y résider pendant plusieurs semaines. Les contributions optiques et radiatives des aérosols de troposphère libre par rapport à celles de la colonne atmosphérique sont encore mal évaluées du fait de la difficulté d’accès et du manque de mesures sur de longues périodes. Ces travaux de thèse se sont donc appuyés sur deux sites d’altitude présentant des topographies adéquates pour l’analyse des aérosols de troposphère libre : le site ACTRIS/GAW du puy de Dôme (PUY, 1 465 m, France) et le site ACTRIS/GAW de Chacaltaya (CHC, 5 240 m, Bolivie). Ces deux sites disposent d’un large jeu de données in-situ et de télédétection. Les résultats montrent ainsi l’importance de la prise en compte de la structure verticale de l’atmosphère et de l’effet de l’humidité sur les propriétés des aérosols dans l’analyse des mesures. Pour la première fois à notre connaissance, l’utilisation de ces techniques instrumentales depuis la station de mesures météorologiques la plus haute du monde (Chacaltaya) a permis d’établir les propriétés optiques des aérosols dans cette région largement influencée par la ville de La Paz et par les émissions de la forêt amazonienne. Les résultats montrent que l’influence des feux de forêt amazoniens à la fin de la saison sèche peut accroitre les propriétés optiques de l’aérosol à cette altitude d’un facteur de 3,5 en moyenne et celles de la troposphère libre de 28 à 80%. La station est d’ailleurs régulièrement influencée par les conditions de la troposphère libre (30% du temps en journée et 60% la nuit). Ce manuscrit présente également des méthodes originales pour la détermination des contributions optiques et radiatives des aérosols de troposphère libre avec l’utilisation conjointe des mesures in-situ, photométriques et LIDAR. Les résultats établissent ainsi des contributions optiques de la troposphère libre au-dessus du puy de Dôme variant de 20%en hiver à 80% en été en moyenne. L’utilisation du modèle de transfert radiatif SBDART permet d’évaluer les contributions radiatives correspondantes qui oscillent entre 13 et 40% pour les courtes longueurs d’onde, soit des forçages radiatifs de -1 W.m-2 à -10 W.m-2. Les différentes sources d’aérosols en surface influencent donc fortement la composition de la troposphère libre, qu’il est alors nécessaire de prendre en compte dans le bilan radiatif global de notre planète. / Atmospheric aerosols impact the earth radiative budget and its climate. Depending on their emission mechanisms and atmospheric transport processes, aerosols can be injected into the free tropophere where their lifetime is increased to up to several weeks. Optical and radiative properties of free tropospheric aerosols are still poorly known because of the difficulties to access high altitudes over long periods of time. The present work is based on two high altitude sites measurements: the ACTRIS/GAW station of puy de Dôme (PUY, 1 465 m, France) and the ACTRIS/GAW station of Chacaltaya (CHC, 5 240 m, Bolivia). These two sites are equipped with a unique set of both in-situ and remote sensing measurements. Results first show the importance of taking into account the vertical atmospheric structure and the hygroscopic properties of aerosols when combining in situ and remote sensing measurements. Measurements from the highest atmospheric station in the world (Chacaltaya), provide for the first time to our knowledge, the aerosol optical properties from this region of the world, segregated into mixing layer aerosols and free tropospheric aerosols.The site is both influenced by anthropogenic emissions from the nearby city, La Paz and pristine emissions from the Amazonian forest. From these measurements, we observe that biomass burning emissions can increase column aerosol optical properties by an average factor of 3,5 and the free tropospheric aerosol optical properties between 28 and 80%. The station is regularly influenced by free tropospheric conditions (30% of the time during daytime and 60% during nighttime). This manuscript is also presenting original methods for retrieving the optical and radiative aerosol contributions from the free tropospheric layer to the total atmospheric column using a combination of in-situ, photometric and LIDAR measurements. Results show that free tropospheric contributions above puy de Dôme station vary from 20% during winter to 80% during summer. The use of the radiative transfer model SBDART allows to evaluate the corresponding shortwave radiative contributions of free tropospheric aerosols between 13 and 40% (between -1 and -10 W.m-2). Thus, the different surface aerosol emissions influence significantly the free tropospheric composition which is essential for radiative budget determination.

Le schéma de régularisation de Taylor-Lagrange (TLRS) est basé sur la définition des champs en tant que distributions à valeurs d'opérateurs (OPVD). L'expression de ces OPVD implique des fonctions test qui, grâce à leurs propriétés (propriétés d'échelles, super-régularité), permettent d'étendre des distributions singulières à tout l'espace. Ce type de régularisation, que l'on peut qualifier de coupure ultra-douce, est efficace quelque soit le degré de divergence originel et produit des amplitudes finies dépendant d'une échelle intrinsèque sans dimensions. Enfin, ce schéma préserve les symétries du groupe de Poincaré et l'invariance de jauge. Après avoir présenté le formalisme TLRS, celui-ci est appliqué au calcul des corrections radiatives en QED ainsi qu'à celles à la masse du boson de Higgs dans le cadre du modèle standard de la physique des particules. Dans une dernière partie, il est appliqué au modèle de Yukawa dans le cadre de la dynamique sur le front de lumière. Les corrections radiatives et un calcul non-perturbatif d'états liés sont effectués. Ces exemples permettent de vérifier, d'une part, l'applicabilité de ce schéma dans différents cas, et d'autre part, de tester son respect des propriétés de symétrie des théories. / The Taylor-Lagrange regularization scheme (TLRS) is based on the definition of fields as operator valued distributions (OPVD). The expression of these OPVDs implies test functions which, thanks to their properties (scaling properties, super-regularity), allow to extend singular distributions to the whole space. This type of regularization, which could be qualified as an ultra-soft cut-off, is efficient for any order of divergences and produces finite amplitudes depending on an intrinsic dimensionless scale. Finally, this scheme respects the Poincaré group symmetries as well as gauge invariance. After an introduction to the TLRS, it is applied to the calculation of radiative corrections to QED and to the mass of the Higgs boson within the standard model of particle physics. In a last section, it is applied to the Yukawa model using the framework of light front dynamics. Radiative corrections and non-perturbative bound state are calculated. This examples allow to verify, on one hand, the applicability of the TLRS, and on the other hand to test its respect of the symmetry properties of the theories.

La fabrication du pare-brise automobile est réalisée par un procédé de thermoformage dans un four tunnel où des feuilles de verre subissent un chauffage différentiel par rayonnement par des centaines d'éléments chauffants électriques contrôlés individuellement. Ces travaux ont pour objectif final de répondre à une problématique industrielle formulée en tant que problème d'optimisation. Elle consiste à aider le conducteur du four à retrouver la cartographie de puissance qui permet d'obtenir le champ de température nécessaire à la surface du verre afin d'aboutir à une forme souhaitée. Pour y parvenir, un modèle du four basé sur la méthode de réseau de composants est utilisé afin de simuler le cycle de chauffage. Dans un premier temps, la précision de la température calculée est améliorée par identification paramétrique en se référant à des données de mesures effectuées in situ. Une étude de sensibilité locale et globale a été réalisée au préalable. Par la suite, dans le but d'accélérer ces calculs, une méthode d'optimisation originale est proposée. Elle consiste à combiner la méthode métaheuristique du Recuit Simulé et l'Algorithme de Re-revêtement pour identifier l'émissivité multi-bande des matériaux. Après avoir effectué une validation sur un modèle simplifié 3D de four radiatif de traitement de matériaux, la méthode originale est appliquée pour le modèle du four réel. Outre l'amélioration de la précision des résultats de la simulation, la nouvelle démarche réduit considérablement le temps de calcul. Dans la deuxième partie du travail, plusieurs méthodes métaheuristiques, telles que l'Algorithme Génétique, le Recuit Simulé, la Recherche Tabou ainsi que leur hybridation sont expérimentées pour un modèle simplifié d'une enceinte radiative. Les résultats montrent que la combinaison de l'Algorithme Génétique et du Recuit Simulé a permis d'accélérer la convergence pour atteindre les champs de températures souhaités sur la surface du produit. Cette méthode est par la suite appliquée avec succès pour inverser le modèle du four afin de retrouver les paramètres de commande du four. / The manufacturing of automobile windshield is produced by a thermoforming process in a tunnel furnace where glass undergoes differential heating radiation by hundreds of electrical heating elements individually controlled. The final purpose of this work is to answer a real industrial problem, which is formulated as an optimization problem. It aims at assisting the furnace driver to find the setting that allows obtaining the required temperature distribution on the glass design in order to achieve the desired shape. Based on the method of network components, a model of the furnace is used to simulate the heating cycle. As a first step of this work, the accuracy of the temperature calculated is improved by parametric identification by referring to the data of measurements taken in situ. A local and global sensitivity analysis was performed beforehand. Thereafter, in order to accelerate these calculations, an original and optimization method is proposed. It consists in combining the Simulated Annealing metaheuristic method and the Replating Algorithm to identify multi-band emissivity. First, the original method validation is performed on a simplified 3D model of radiative enclosure, and then applied to the real furnace model. The new approach significantly reduces the computation time while improving the accuracy of the simulation results. In the second part of this work, several metaheuristic methods, such as Genetic Algorithm, Simulated Annealing, Tabu Search, and their hybridization are tested on a simplified model of a radiative enclosure. Results show that the combination of Genetic Algorithm and Simulated Annealing has accelerated the convergence to achieve the desired temperature fields on the product surface. This new method is successfully applied to the real furnace model to find the optimal control parameters.

Plusieurs méthodes existent pour évaluer de manière opérationnelle l'éclairement solaire au sol à partir d'images acquises par satellite. Durant sa thèse soutenue en 2009 à MINES ParisTech, Oumbe a conçu une nouvelle méthode, Heliosat-4, faisant appel à des modèles numériques du transfert radiatif et à des approximations d'exécution rapide. La présente thèse vise à consolider ces résultats et à effectuer la validation complète de la méthode Heliosat-4. Elle s'inscrit dans une collaboration scientifique internationale dans les projets européens MACC (Monitoring Atmosphere Composition and Climate) et MACC-II.Oumbe a proposé une approximation de l'équation de transfert radiatif s'écrivant alors comme un produit de l'éclairement par ciel clair par un terme d'extinction dû aux nuages. Nous avons établi que les erreurs liées à cette approximation sont très faibles dans les conditions usuelles et qu'elle peut donc être utilisée dans Heliosat-4, ce qui en facilitera l'implémentation informatique ainsi que son fonctionnement opérationnelle.La méthode Heliosat-4 est donc ainsi composé de deux modèles composés d'abaques : McClear pour l'éclairement par ciel clair et McCloud pour l'extinction cet éclairement due aux nuages. A l'aide de mesures in-situ d'éclairements direct et diffus de référence, nous avons analysé finement les performances de Heliosat-4 selon différentes conditions. La qualité de la première version pré-opérationnelle de Heliosat-4 est jugée satisfaisante car elle permet des estimations d'éclairement global avec une précision de l'ordre de celles des méthodes existantes mais des estimations des composantes directe et diffuse sensiblement de meilleure qualité. / Several methods have been developed to assess operationally the surface solar irradiance from satellite images. During his PhD thesis presented in 2009 at MINES ParisTech, Oumbe has designed a new method using numerical radiative transfer model and fast approximations. The present PhD thesis aimed at consolidating these results and validating Heliosat-4. This work is the international scientific collaboration framework of the European-funded projects MACC (Monitoring Atmosphere Composition and Climate) and MACC-II.As a foundation of Heliosat-4, Oumbe has proposed an approximation of the radiative transfer equation by a product of clear-sky irradiance and a term describing the cloud extinction. We have established that estimation errors due to this approximation are very small in usual conditions and that this approximation may be adopted. It allows a convenient modular development of Heliosat-4 and eases its future operational use.The Heliosat-4 method is then composed of two abacus-based models: McClear for the irradiance under clear-sky and McCloud for the irradiance extinction due to clouds. With in-situ reference measurements of direct and diffuse irradiance, we have carried out deep performance analysis of Heliosat-4, under different conditions. The quality of this first preoperational version of Heliosat-4 is judged satisfactory as it enables estimations of global irradiance with the same level of quality of other existing methods in literature but also estimations of direct and diffuse irradiances with a noticeable better quality.

L'ère des observations interférométriques mène à la nécessité d'une description plus précise de la structure physique et de la dynamique des régions de formation d'étoiles, des coeurs pré-stellaires et des disques proto-planétaires. L'émission moléculaire et du continuum de la poussière peuvent être décrites par de multiples composantes physiques. Pour comparer avec les observations, un modèle de transfert radiatif précis et complexe de ces régions est nécessaire. J'ai développé au cours de cette thèse une application autonome appelée GASS (Generator of Astrophysical Sources Structures, Quénard et al., soumis) à cette fin. Grâce à son interface, GASS permet de créer, de manipuler et de mélanger différents composants physiques tels que des sources sphériques, des disques et des outflows. Dans cette thèse, j'ai utilisé GASS pour travailler sur différents cas astrophysiques et, entre autres, j'ai étudié en détail l'eau et l'émission de l'eau deutérée dans le coeur pré-stellaire L1544 (Quénard et al., 2016) ainsi que l'émission des ions dans la proto-étoile de faible masse IRAS16293-2422 (Quénard et al., soumis). / The era of interferometric observations leads to the need of a more and more precise description of physical structure and dynamics of star-forming regions, from pre-stellar cores to proto-planetary disks. The molecular and dust continuum emission can be described with multiple physical components. To compare with the observations, a precise and complex radiative transfer modelling of these regions is required. I have developed during this thesis a standalone application called GASS (Generator of Astrophysical Sources Structures, Quénard et al., submitted) for this purpose. Thanks to its interface, GASS allows to create, manipulate, and mix several different physical components such as spherical sources, disks, and outflows. In this thesis, I used GASS to work on different astrophysical cases and, among them, I studied in details the water and deuterated water emission in the pre-stellar core L1544 (Quénard et al., 2016) and the emission of ions in the low-mass proto-star IRAS16293-2422 (Quénard et al., submitted).

Ce travail de thèse est dédié à la modélisation du thermoformage du verre. Le procédé consiste à déformer une plaque de verre sous l'effet de son propre poids. Posée sur un support et placée dans un four, la température de la pièce augmente et sa viscosité diminue, ce qui permet d'obtenir la forme désirée. Les simulations numériques, qui se basent sur un modèle thermomécanique, doivent permettre de mieux comprendre l'influence, sur le produit final, des différents paramètres d'essai, comme le chargement thermique, la géométrie et le matériau du moule ou encore la forme initiale de la pièce. Pour ce faire, le logiciel commercial Abaqus®, qui utilise une méthode de résolution des calculs par éléments finis, prend en charge les aspects mécaniques et conductifs. En revanche, comme le verre est un milieu semi-transparent, la modélisation du transfert radiatif est complexe et nécessite le développement d'un code se basant sur une méthode de Monte Carlo dite réciproque. La méthode a été validée en deux dimensions sur des cas-tests de la littérature scientifique. Le code a ensuite été implémenté dans le logiciel Abaqus® afin de réaliser des simulations de thermoformage sur moule et en suspension. Le verre est considéré comme un matériau élasto-visco-plastique obéissant à un modèle de Maxwell simple et la thermodépendance de la viscosité est prise en compte par une loi WLF. Une attention particulière a été accordée au modèle radiatif. Différentes hypothèses, issues de la littérature scientifique, sont testées afin de vérifier leur validité dans notre cas d'étude / This dissertation is dedicated to the modelling of the glass sagging process. This operation consists in forming a sheet or a plate of glass by heating it in a furnace. Glass temperature rises and reaches a work temperature at which viscosity is low enough to allow glass to sag under its own weight due to gravity. Numerical simulation, based on a thermomechanical model, can help to better understand the influence of the different parameters on the final product, such as the thermal loading, the shape and material of the mould or even the initial geometry of the glass plate. Thus, the commercial software Abaqus® is used to solve the problem by a finite elements method. However, it cannot render the complexity of the radiative heat transfer in glass. So, a Monte Carlo code based on a reciprocal method was developed and validated using benchmarks from the scientific literature. Then, the code was implemented into Abaqus® in order to simulate glass sagging on a mould or glass forming by the draping process. Glass is considered as an elasto-viscoplastic material which obeys a Maxwell model. Viscosity is dependant to temperature according to a WLF law. Special care was devoted to the radiative heat transfer. Different hypothesis are reviewed and performed to check their validity when applied to our numerical set-up