O objetivo deste estudo foi avaliar parâmetros tais como, moagem, tamanho de partículas, homogeneidade, segregação, dissolução de amostras e amostragem de suspensão, visando o desenvolvimento de método para a determinação de Ir, Pd, Pt e Rh por espectrometria de emissão óptica com plasma indutivamente acoplado (ICP OES) em sílica mesoporosa para uso catalítico. Após a moagem em moinho eletrônico de almofariz e pistilo, a análise do tamanho e distribuição das partículas revelou que 89% das partículas apresentam tamanhos <100 &#181;m e 23% com tamanho <5 &#181;m. Na avaliação de segregação dos analitos na amostra, os resultados obtidos para amostra com moagem grosseira, indicaram que os elementos estão distribuídos de maneira heterogênea entre as frações. Entretanto, a moagem fina aumentou a homogeneidade e esta pode ser uma etapa importante e necessária no sentido de evitar segregação da amostra. A análise por difração e de Raios-X mostrou que não ocorreram reações mecanoquímicas durante a moagem e por fluorescência de Raios-X forneceu resultados discrepantes entre os laboratórios, o pode estar relacionado com a heterogeneidade que induziu erros de amostragem. A avaliação de desempenho do ICP OES foi feita a partir do estabelecimento de parâmetros de mérito como seletividade, repetibilidade, estabilidade, robutez e limite de detecção. Na otimização foram considerados os parâmetros potência, vazão de nebulização e vazão de gás auxiliar. Foi feito um estudo em solução aquosa e em meio da amostra para escolha das linhas de emissão dos elementos Ir, Pd, Pt, Rh e dos candidatos a padrões internos. Apesar dos bons resultados encontrados em solução aquosa, o aumento do sinal de fundo em meio da amostra restringiu a escolha dos comprimentos de onda. Para o estudo da dissolução das amostras foram avaliados quatro procedimentos em forno de micro-ondas com frasco fechado. Os melhores resultados foram obtidos com 150 mg de amostra, 6 mL de água régia e 210 ºC de aquecimento durante 60 min. As determinações por ICP OES usando padrões internos Ru para Pd, Yb para Pt e Dy para Rh foram aquelas que forneceram os resultados mais coerentes com aqueles obtidos por fluorescência de Raios-X. Nos estudos visando à introdução de amostra na forma de suspensão foram avaliados: a massa de amostra para o preparo da suspensão, o efeito da concentração de ácido nítrico e o uso de padrão interno na tentativa de corrigir erros de matriz. 60 mg de amostra com um volume final de 20 mL de ácido nítrico 0,5% (v v-1) foram os escolhidos. Entretanto, a determinação dos elementos via suspensão mostrou resultados muito inferiores aos valores de concentrações esperados. Irídio foi investigado somente para otimização de ICP OES porque a sílica mesoporosa para uso de catalisador automotivo tem não esse elemento em sua composição. / The aim of this study was to evaluate parameters such as grinding, particle size, homogeneity, segregation, sample dissolution and slurry sampling for method development for the determination of Ir, Pd, Pt and Rh in silica mesoporosa for catalyst using inductively coupled plasma optical emission spectrometry (ICP OES). After the grind in an electronic grider with mill and pistil, the analysis of the particle size distribution showed 89% of particles with sizes <100 &#181;m and 23% of particles with sizes <5&#181;m. The separation in different particle size fractions showed that the sample with coarse grind (as sent to the lab) is not homogenous for the elements in comparison with that submitted to grind. X-Ray Diffraction and X-Ray Fluorescence analysis did not showed chemical reactions during the grinding, however a discrepancy between results of different laboratories. The performance of ICP OES was made by using figures of merit such as selectivity, repeatability, stability, robustness and limit of detection. In the optimization of ICP OES power supply, nebulizer flow rate and auxiliary gas flow rate were considered. Studies were conducted in aqueous and in the sample solution to choose emission wavelengths for Ir, Pd, Pt, Rh and internal standard candidates. Despite the good results obtained in aqueous solution, increasing the background signal in sample media restricted the choice of the emission wavelengths. For sample dissolution four procedures using microwave oven with closed-vessel were evaluated. The best results were obtained with 150 mg of the sample, 6 mL of aqua-regia and 210 ºC of heating, during 60 min. Determinations by ICP OES using internal standards Ru for Pd, Yb for Pt, and Dy for Rh were those who provided more consistent results in comparison with those obtained by x-ray fluorescence. For slurry sampling introduction sush parameters were evaluated: the mass of sample to prepare the slurry, the effect of the nitric acid concentration and the use of internal standard to correct matrix effects. 60 mg of the sample with a final volume of 20 mL of nitric acid 0.5% (v v-1) were chosen. However, the determination of Pd, Pt and Rh via suspension showed concentrations below to those expected, as determined by ICP OES in solution. Iridium was investigated only for ICP OES optimization because the silica mesoporosa for automobile catalyst use have not this element in its composition.

"Fluid-structure interaction arises in a wide array of technological applications including naval and marine hydrodynamics, civil and wind engineering and flight vehicle aerodynamics. When a fluid flows over a bluff body such as a circular cylinder, the periodic vortex shedding in the wake causes fluctuating lift and drag forces on the body. This phenomenon can lead to fatigue damage of the structure due to large amplitude vibration. It is widely believed that the wake structures behind the structure determine the hydrodynamic forces acting on the structure and control of wake structures can lead to vibration control of the structure. Modeling this complex non-linear interaction requires coupling of the dynamics of the fluid and the structure. In this thesis, however, the vibration of the flexible cylinder is prescribed, and the focus is on modeling the fluid dynamics in its wake. Low-dimensional iterative circle maps have been found to predict the universal dynamics of a two-oscillator system such as the rigid cylinder wake. Coupled map lattice (CML)models that combine a series of low-dimensional circle maps with a diffusion model have previously predicted qualitative features of wake patterns behind freely vibrating cables at low Reynolds number. However, the simple nature of the CML models implies that there will always be unmodelled wake dynamics if a detailed, quantitative comparison is made with laboratory or simulated wake flows. Motivated by a desire to develop an improved CML model, we incorporate self-learning features into a new CML that is trained to precisely estimate wake patterns from target numerical simulations and experimental wake flows. The eventual goal is to have the CML learn from a laboratory flow in real time. A real-time self-learning CML capable of estimating experimental wake patterns could serve as a wake model in a future anticipated feedback control system designed to produce desired wake patterns. A new convective-diffusive map that includes additional wake dynamics is developed. Two different self-learning CML models, each capable of precisely estimating complex wake patterns, have been developed by considering additional dynamics from the convective-diffusive map. The new self-learning CML models use adaptive estimation schemes which seek to precisely estimate target wake patterns from numerical simulations and experiments. In the first self-learning CML, the estimator scheme uses a multi-variable least-squares algorithm to adaptively vary the spanwise velocity distribution in order to minimize the state error (difference between modeled and target wake patterns). The second self-learning model uses radial basis function neural networks as online approximators of the unmodelled dynamics. Additional unmodelled dynamics not present in the first self-learning CML model are considered here. The estimator model uses a combination of a multi-variable normalized least squares scheme and a projection algorithm to adaptively vary the neural network weights. Studies of this approach are conducted using wake patterns from spectral element based NEKTAR simulations of freely vibrating cable wakes at low Reynolds numbers on the order of 100. It is shown that the self-learning models accurately and efficiently estimate the simulated wake patterns within several shedding cycles. Next, experimental wake patterns behind different configurations of rigid cylinders were obtained. The self-learning CML models were then used for off-line estimation of the stored wake patterns. With the eventual goal of incorporating low-order CML models into a wake pattern control system in mind, in a related study control terms were added to the simple CML model in order to drive the wake to the desired target pattern of shedding. Proportional, adaptive proportional and non-linear control techniques were developed and their control efficiencies compared."

Ce travail concerne la modélisation numérique et expérimentale de la variabilité des propriétés thermo-hydriques de matériaux cimentaires en vue de l’évaluation de son impact sur la prédiction du comportement hygrothermique de parois de bâtiments. Une approche probabiliste qui prend en compte la variabilité spatiale des propriétés de matériaux lors des transferts couplés de chaleur et d’humidité a été développée. Elle est basée sur la génération, par la décomposition modale de Karhunen-Loève, de champs aléatoires spatialement corrélés. Une implémentation d’un modèle de transfert hygrothermique dans un code de simulation numérique a été ensuite réalisée en adoptant cette démarche stochastique. Cette dernière, qui considère comme variables d’entrée des champs aléatoires, permet de quantifier l’incidence de cette variabilité sur le comportement hygrothermique d’une paroi de bâtiment. Une étude préalable, dédiée à l’évaluation de l’incidence de la variabilité aléatoire du coefficient de diffusion, a été entreprise en considérant une variabilité de ±30% pour un mortier et de ±20% pour un BHP suivant une loi de distribution normale. Aussi, nous avons relevé un certain nombre d’incertitudes possibles de la teneur en eau à saturation tout en montrant leurs effets sensibles sur le résultat de la prédiction du comportement hygrothermique. Ces études ont permis de mettre en exergue l’importance de la prise en compte des incertitudes sur les données du matériau lors des simulations numériques des transferts hygrothermiques. Sur le plan expérimental, une campagne d’évaluation de la variabilité spatiale des paramètres les plus influents a été menée. Cette campagne a été réalisée sur un voile de dimension 2x1,2 m fabriqué au laboratoire. À l’issue de ce programme expérimental, l’espérance, la variance et la longueur de corrélation des propriétés étudiées (porosité à l’eau, perméabilité à la vapeur, isotherme de sorption et perméabilité au gaz) ont été déterminées. Ces trois paramètres sont indispensables pour la bonne mise en œuvre de la décomposition de Karhunen-Loève. Aussi, une autre campagne de caractérisation expérimentale a été menée sur des pâtes de ciment, mortiers et béton. Elle a été divisée en trois grandes parties selon les propriétés étudiées : (i) Les propriétés microstructurales et d’hydratation où l’on retrouve les mesures des porosités à l’eau et au mercure ainsi que les distributions de la taille des pores et une analyse de l’effet du taux d’hydratation de matériaux cimentaires sur leurs propriétés hygrothermiques. (ii) Les propriétés hydriques : dans cette partie, une analyse sous différents angles (évolution en fonction de l’âge des matériaux, en fonction de la température, effet des constituants des matériaux, etc.) a été réalisée sur les isothermes de sorption et sur la perméabilité à la vapeur d’eau. (iii) Les propriétés thermiques où des mesures de conductivités thermiques et de chaleurs spécifiques ont été effectuées. Les résultats de l’étude ont mis en exergue les limites des approches déterministes suite à leurs confrontations avec les résultats obtenus par l’approche probabiliste, mise en œuvre dans le cadre du présent travail. / This study deals with the experimental and the numerical modeling of the variability properties of cement based materials to evaluate their effects on the prediction of hygrothermal behavior of building envelops. A probabilistic approach taking into account the spatial variability of the materials properties during the coupled heat and mass transfer has been developed. It is based on the generation of spatially correlated random fields by the Karhunen Loève decomposition. The stochastic model’s program has been implemented in a numerical simulation code. Using this tool that considers the input variables as random fields, the impact of this variability on the hygrothermal behavior of building envelops was quantified. A prior study dealing with the assessment of the effect of the diffusion coefficient random variability was carried out by considering a variation of ±30% for mortar and ±20% for high performance concrete (HPC) according to a normal distribution. Also, we have identified some possible uncertainties of the water content at saturation and showed their significant impact on the prediction of hygrothermal behavior of the material. These studies highlight the importance of considering the data uncertainties of building materials during numerical simulation of hygrothermal transfers. At the experimental level, the spatial variability of the most influential parameters was evaluated. It was carried out by manufacturing a concrete wall in lab. At the end of this experimental program, the expected value, standard deviation and the correlation length of the studied properties (water porosity, water vapor permeability, sorption isotherm and gas permeability) were determined. These three parameters are important for the successful implementation of Karhunen Loeve decomposition. Also, another experimental program was conducted on cement pastes, mortars and concrete. It was divided into three parts according to the studied properties:(i) Hydrations and microstructural properties which include the measurement of water and mercury porosity, the pore size distributions and an analysis of some techniques for stopping cement hydration.(ii) Hydric properties: where an analysis of the sorption and the water vapor permeability was performed considering their evolution with materials ages, temperature…(iii) Thermal properties where measurement of specific heat and thermal conductivity were performed. The result of the study highlighted the limits of deterministic approaches after their confrontation with the obtained results using the probabilistic one developed in this work.

Growing concerns about climate change have led to the world experiencing an unprecedented push towards renewable energy. Economic drivers and government policies mean that small, distributed forms of generation, like solar photovoltaics, will play a large role in our transition to a clean energy future. In this thesis, a novel DC-DC converter known as the Coupled Inductors Combined Cuk-SEPIC' (CI-CCS) converter is explored, which is particularly attractive for these photovoltaic applications. A topological modification is investigated which provides several benefits, including increased power density, efficiency, and operational advantages for solar energy conversion. The converter, which is based on the combination of the Cuk and SEPIC converters, provides a bipolar output (i.e. both positive and negative voltages). This converter also offers both step-up and step-down capabilities with a continuous input current, and uses only a single, ground-referenced switching device. A significant enhancement to this converter is proposed: magnetic coupling of the converter's three inductors. This can substantially reduce the CI-CCS converter's input current ripple - an important benefit for maximum power point tracking (MPPT) in photovoltaic applications. The effect of this coupling is examined theoretically, and optimisations are performed - both analytically and in simulations - to inform the design of a 4 kW prototype CI-CCS converter, switched at a high frequency (100 kHz) with a silicon carbide (SiC) MOSFET. Simulation and experimental results are then presented to demonstrate the CI-CCS converter's operation and highlight the benefits of coupling its inductors. An efficiency analysis is also undertaken and its sources of losses are quantified. The converter is subsequently integrated into a domestic photovoltaic system to provide a practical demonstration of its suitability for such applications. MPPT is integrated into the CI-CCS DC-DC converter, and a combined half bridge/T-type converter is developed and paired with the CI-CCS converter to form an entirely transformerless single-phase solar energy conversion system. The combination of the CI-CCS converter's bipolar DC output with the combined half bridge/T-type converter's bipolar DC input allows grounding at both the photovoltaic panels and the AC grid's neutral point. This eliminates high frequency common mode voltages from the PV array, which in turn prevents leakage currents. The entire system can be operated in grid-connected mode - where the objective is to maximise power extracted from the photovoltaic system, and is demonstrated in stand-alone mode - where the objective is to match solar generation with the load's power demands.

Cette thèse porte sur le développement d'un traducteur ultrasonore multi-élément capacitif à couplage air (MEACUT) et son utilisation dans le domaine du contrôle non destructif (CND) de matériaux. Un modèle est employé pour simuler numériquement ce traducteur, et pour optimiser sa conception. Un prototype est ensuite fabriqué, puis caractérisé expérimentalement pour quantifier ses performances. Son originalité réside dans le fait qu'il possède une large bande passante en fréquence, tout en offrant la possibilité d'une focalisation variable. Ce prototype est alors employé pour la détection d'un endommagement causé par impact, dans une plaque composite. Il est clairement constaté que la résolution spatiale du procédé d'inspection employé (C-scan) est fortement améliorée grâce aux performances techniques du MEACUT. Enfin, un modèle hybride 3D est développé pour simuler, rapidement et intégralement, ce procédé de CND. Le très bon accord obtenu entre prédictions numériques et mesures expérimentales laisse présager que cet outil de simulation pourra servir à mettre au point d'autres expérimentations de CND, qui pourront à leur tour exploiter les performances du MEACUT. / This thesis focuses on the development of a Multi-Element Air-coupled Capacitive Ultrasonic Transducer (MEACUT) and its use in the field of non-destructive testing (NDT) of materials. A numerical model is used to simulate the translator, and to optimize its design. A prototype is then built and experimentally characterized to quantify its performance. Its originality lies in the fact that it has a broad frequency bandwidth while offering the possibility of a dynamic focusing. This prototype is then used for the detection of damage caused by impact, in a composite plate. It is clearly found that the spatial resolution of the inspection process employed (C-scan) is greatly improved thanks to the technical performance MEACUT. Finally, a 3D hybrid model is developed to simulate quickly and fully, the process of NDT. The good agreement obtained between numerical predictions and experimental measurements suggests that this simulation tool can be used to develop other NDT experiments, which may in turn exploit the performance of MEACUT.

Em eletroforese capilar, a reprodutibilidade depende essencialmente da manutenção do sinal e no tempo de migração das espécies. A alteração do fluxo eletrosmótico (EOF) entre corridas é um dos principais fatores que afetam esses parâmetros. Nesse trabalho, é proposto um sistema de monitoramento do EOF baseado em marcas térmicas (TMs), que consistem em sinais presentes em um eletroferograma devido a uma perturbação do eletrólito de corrida ocasionada por um breve aquecimento (da ordem de 100 ms) de uma pequena porção do capilar (aproximadamente 1 mm), sob ação do campo elétrico. O dispositivo responsável pelo aquecimento que se mostrou mais eficiente consistiu em um resistor SMD de 15 &#937; sobre o qual 5 V eram aplicados para gerar aquecimento. Estudos sobre a origem das TMs sugerem que o fenômeno está relacionado às mudanças do número de transporte das espécies devido à variação de temperatura. Essas conclusões foram baseadas na comparação do perfil dos sinais de TM com valores de número de transporte, além de estudos envolvendo simulação computacional. Diversas aplicações para as marcas térmicas foram propostas: correção de variação de tempo de migração devido a alterações do EOF, medida direta do EOF e para otimização de parâmetros do detector condutométrico sem contato. As TMs revelaram-se, então, um sistema prático e de fácil implementação, que pode ser utilizado como uma espécie de padrão interno. Adicionalmente, o emprego das TMs em um novo sistema de injeção de amostra, chamado de injeção térmica, foi também proposto. A vantagem desse tipo de injeção consiste na minimização de problemas de contínua introdução de amostra no canal de separação em microchip-CE. Para realizar essa injeção, o capilar é preenchido com a amostra diluída em um eletrólito de corrida e a injeção ocorre quando uma TM é gerada. Estudos com soluções de NaCl apresentaram faixa de resposta linear entre 10 &#181;mol L-1 e 1 mmol L-1 para sódio. Entretanto, problemas de interferência sobre a sensibilidade e a mobilidade dos analitos foram obstáculos enfrentados. Também foi alvo de estudo o emprego da injeção térmica em uma técnica de multiplexação para aumento de relação sinal/ruído, baseado em um novo algoritmo. Embora o modelamento matemático tenha sido eficiente para decodificação de sinais, não foi possível obter o aumento na relação sinal/ruído almejado, pois o tempo de análise seria excessivamente longo. No entanto, a utilização da injeção térmica nesse tipo de estratégia mostrou-se bastante adequada, pois ela requer que diversas injeções sejam realizadas sequencialmente sem a interrupção do potencial de corrida, característica que esse sistema de injeção permite. Posteriormente, a implementação de estratégias que gerassem TMs em tempos mais curtos poderiam viabilizar a técnica. Finalmente, a utilização de TMs como ferramenta para obtenção de constantes físico-químicas, como pKa e mobilidade por exemplo, utilizando a eletroforese capilar mostra-se como uma grande perspectiva para a técnica / In capillary electrophoresis, reproducibility depends essentially on the maintenance of the signal and the migration time when the analysis of the same sample is repeated. Variations in the electroosmotic flow (EOF) between runs are one of the major factors affecting these parameters. In this work, a new approach to monitor the EOF based on thermal marks (TMs) is proposed. TM consists in a signal present in the electropherogram caused by heating (typically 100 ms) a small portion of the capillary (approximately 1 mm), while the electric field is applied. The most effective device to promote the heating was a 15-&#937; SMD resistor, powered at 5 V. Studies about the origin of TMs suggest that the phenomenon is related to variations in the transport number of the species due to alterations in the temperature. This conclusion was based on comparison between the TM profile and the transport numbers values, as well as results from computer simulation. The proposed applications for TM include: correction in variations of migration times with alterations on EOF, EOF measurement and optimization of the parameters of contactless conductivity detection. Additionally, a new sample injection procedure, called thermal injection, was also proposed. The advantage of this injection consists in minimization of leakage problems related to the continuous introduction of the sample in the separation channel in microchip-CE. To perform the thermal injection, the capillary is completely filled with the sample diluted in a BGE and the injection occurs with the generation of a TM. Studies with NaCl solutions presented extensive linear response range from 10 &#181;mol L-1 to 1 mmol L-1. However, interference problems on sensibility and analytes mobilities appeared. The use of low-concentration solutions diminishes these problems. However, the analytical signal is also diminished, needing a strategy to raise the sensitivity. Thus, a multiplexing technique based on a new algorithm was also introduced in order to improve signal-to-noise ratio. Although the efficiency of the mathematic modeling on the signals decoding, the desired improvement of signal/noise ratio was not obtained, because the analysis time would be excessively high. However, the employment of the thermal injection seems to be very suitable for multiplexing, due to the possibility of performing several sequential injections with no interruption of the electric field. Afterwards, the implementation of fast velocity devices to generate TMs would enable the technique. Finally, the great perspective to applications of TM concerns in the use of TMs to obtain physical chemical constants, such as pKa and ionic mobilities. The present work describes values of ionic mobilities calculated to monopropyl carbonate, monoisopropyl carbonate, and hydronium.

Depuis le milieu des années 90, la minéralisation in situ du CO2 est envisagée comme une solution durable et efficace pour limiter ses émissions anthropiques vers l'atmosphère. Il s'agit de récupérer le CO2 émis par certaines industries pour le piéger en profondeur sous forme minérale (carbonates) dans les aquifères mafiques et ultramafiques naturels (de type basaltes et péridotites). La carbonatation du CO2 a été largement décrite dans les systèmes naturels où elle apparait à travers une série de réactions chimiques couplées au transport des espèces réactives dans le fluide. Plusieurs expériences en réacteurs fermés ont été menées depuis une quinzaine d'années afin de mieux comprendre les paramètres physico-chimiques contrôlant ces réactions. Mais très peu d'études n'ont encore caractérisé les processus de transport-réactif au cours de l'injection et de la minéralisation in situ du CO2 dans ces roches.Ces travaux visent à répondre à 3 principaux objectifs : (i) caractériser l'évolution des chemins réactifs lors de l'injection de CO2 dans des roches (ultra-)mafiques, (ii) mesurer les effets en retour des réactions sur les propriétés hydrodynamiques du milieu et (iii) quantifier le rendement et la pérennité des processus sur le long terme. Ils s'appuient sur le développement de protocoles expérimentaux pour (i) reproduire l'injection de CO2 dans les roches (ultra-)mafiques et (ii) caractériser les réactions à l'aide d'une série d'outils géochimiques et analytiques de l'échelle atomique à centimétrique. Trois séries d'expériences de percolation réactive ont été réalisé sur des agrégats (ultra-)mafiques relativement simples (olivines de San Carlos et d'Hawaii) et plus complexes (basaltes de Stapafell) dans des conditions de P-T-confinement in situ (Ptot=10-25 MPa ; T=180-185°C;Pconf=15-28 MPa).Les résultats obtenus ont permis de différencier plusieurs chemins réactifs dans ces systèmes en fonction du transport du fluide, de la porosité du milieu, des hétérogénéités locales de la roche, de la minéralogie et/ou des variations locales de la composition chimique du fluide. Les calculs du bilan de masse ont révélé une minéralisation efficace du CO2 contrôlée par les propriétés chimiques et hydrodynamiques du milieu. Mais certaines réactions associées à l'altération des roches (ultra-)mafiques (hydratation) ont des effets en retour négatifs sur les propriétés réservoirs de la roche (porosité, perméabilité) pouvant compromettre la pérennité du stockage du CO2 dans les aquifères naturels sur le long terme.Ces nouvelles données permettront aux modèles numériques de mieux simuler la carbonatation des roches (ultra-)mafiques en connaissant les propriétés hydrodynamiques du milieu et les hétérogénéités structurales du réservoir. Elles suggèrent aussi qu'un meilleur contrôle de certains paramètres d'injection, comme le débit ou la composition du fluide injecté (ex: pCO2), permettrait d'améliorer le taux et le rendement de la carbonatation. / Since the mid-90s, in situ mineralization of CO2 has been considered as a safe and efficient solution to mitigate its anthropogenic emissions to the atmosphere. It is to recover the CO2 emitted by some industries and trap it in the mineral form (carbonates) in mafic and ultramafic aquifers (e.g. basalts and peridotites). The carbonation of CO2 has been widely described in natural systems where it occurs through a series of complex chemical reactions coupled to the transport of reactive species in the fluid. Numerous experiments have been conducted in batch reactors over the past fifteen years to better understand the physico-chemical parameters controlling the carbonation of (ultra-)mafic rocks. But few studies have further characterized the coupling reactive-transport processes during the injection and in situ mineralization of CO2 in these rocks.This work aims to meet 3 main objectives: (i) characterize changes in reaction paths during the injection of CO2 in (ultra-)mafic systems, (ii) measure the feedbacks effects of chemical reactions on the hydrodynamic rock properties and (iii) quantify the efficiency and sustainability of such processes over long time periods. It is based on the development of experimental protocols to (i) reproduce the injection of CO2 into (ultra-)mafic rocks and (ii) characterize the reactions using a series of geochemical and analytical tools from the atomic to the centimetric scale. Three series of reactive percolation experiments have been performed on (ultra-)mafic aggregates from relatively simple (olivines from San Carlos and Hawaii) to more complex samples (basalts from Stapafell) under in situ P-T-containment conditions (Ptot=10-25 MPa; T=180-185°C; Pcont=15-28 MPa).The results allowed us to differentiate several reactions paths in these systems depending on the fluid transport, rock porosity, local hydrodynamic properties, mineralogy and/or local changes in the fluid composition. Mass balance calculations have revealed an efficient mineralization of CO2 in the samples. It is controlled by the chemical and the hydrodynamic properties of the rock at the pore scale. But some reactions associated with the alteration of (ultra-)mafic rocks (e.g. hydration) have negative feedbacks effects on the reservoir rock properties (porosity and permeability) that may compromise the sustainability of CO2 storage in natural aquifers in the long term.These new supporting data will allow numerical models to better simulate the carbonation of (ultra-)mafic rocks knowing the hydrodynamic properties and the structural heterogeneities of the reservoir. They also suggest that a better control of some injection parameters, such as the flow injection rate and the injected fluid composition (e.g. pCO2), would improve the rate and yield of CO2 mineralization in these systems.

No presente trabalho é apresentado um conjunto de medidas de seção de choque do processo de fusão nuclear nos sistemas ANTPOT. 12 C + ANTPOT. 63, 65 CU, em um intervalo de energia correspondente à 0.91.8 vezes a barreira coulombiana. O método experimental utilizado foi a detecção dos resíduos de evaporação através da técnica de tempo de vôo, acoplado à um defletor eletrostático para separar as partículas provenientes do espalhamento elástico dos produtos de fusão. As vantagens e limitações do método são discutidas em detalhes. As funções de excitação de fusão dos sistemas citados foram analisadas via modelo de penetração de barreira unidimensional , utilizando vários potenciais nucleares. Dessa análise comprovou-se que a secção de choque de fusão é subestimada em energias em torno e abaixo da barreira coulombiana. Nesse sentido foi feito um cálculo, utilizando o código computacional CCFUS, que simula o acoplamento de canais, com a finalidade de descobrir quais canais contribuem para o processo de fusão nuclear em baixas energias. As seções de choque de fusão obtidas foram comparadas com outros sistemas - ANTPOT. 12 C + ANTPOT. 63, 65 CU.- que utilizam os mesmos núcleos-alvo. A comparação revelou que os sistemas ANTPOT. 12 C + ANTPOT. 63, 65 CU possuem o maior aumento de seção de choque de fusão subcoulombiana e esse fato pode estar relacionado à deformação estática do projétil. A comparação das distribuições da velocidade dos resíduos de evaporação para os dois de sistemas revelou um canal intenso para o sistema ANTPOT. 12 C + ANTPOT. 63 CU que não está presente para o outro sistema. / In this work we present cross-section measurements for nuclear fusion in the 12C+ 63,65 Cu system, at 12C energy range from 0.9 to 1.8 times the Coulomb barrier. In order to detect and to obtain the mass identification of the evaporation residues following the fusion process, the time of flight method was adopted in conjunction with an eletrostatic deflector capable of separating the evaporation residues from the beam particles. The limitation and advantadges of this method of measurement are discussed. The excitation functions for the above systems were analysed using the unidimensional barrier penetration model with different nuclear potentials. Theoretical fusion cross-section values obtained from this analysis were systematically smaller than our measured values, in the energy region below the Coulomb barrier. In order to discover which channel enhances the fusion cross-section in this region, a coupled channel calculation was performed, with the aid of CCFUS code. The experimental data for the above reactions were compared with the systems 12C+ 63,65 Cu, measured by our group. In this comparison, It was noted that the systems 12C+ 63,65 Cu have greater fusion cross section below the Coulomb barrier. This fact can be related to the static deformation of 12C. The comparison of velocity spectra of the evaporated residues for the two systems shows that 12C+63Cu has a strong reaction channel that was not present in the 12C+ 65Cu system.

Ce travail s'inscrit dans le cadre du projet SIMED fédérant les activités de modélisation à l'échelle méditerranéenne, et plus globalement dans le programme MERMEX qui vise à étudier les cycles biogéochimiques en mer Méditerranée et leurs évolutions futures. L'étape préliminaire a été de coupler la plateforme de modélisation hydrodynamique (NEMO) à celle de modélisation biogéochimique mécaniste (Eco3M), afin de réaliser une simulation (2000-2012) utilisant les sorties hydrodynamiques de la configuration NEMO-MED12 pour forcer le modèle biogéochimique Eco3M-MED. Les nombreuses comparaisons menées dans cette thèse (chlorophylle, sels nutritifs, production primaire, etc.) ont aidé à s'assurer de la capacité du modèle à reproduire les principales caractéristiques biogéochimiques de la Méditerranée. Ce travail a permis de généraliser le rôle majeur joué par le carbone organique dissous dans la pompe biologique à l'échelle de la mer Méditerranée. Les résultats montrent que la production de carbone organique particulaire est restreinte aux régions de forte dynamique physique, tandis que l'accumulation de carbone organique dissous dans les eaux de surface est commune à la plupart des régions du bassin. Ce dernier processus s'est avéré dépendant des contenus cellulaires du phytoplancton et des bactéries hétérotrophes. Finalement d'après le modèle, la fraction dissoute du carbone organique contribuerait à hauteur d'environ 64 % à l'exportation dans le bassin Ouest, et de 90 % dans le bassin Est. Le bassin Est -en dépit de sa plus forte oligotrophie- s'avère participer à près de 60 % à l'exportation de carbone organique en mer Méditerranée. / This work is part of the SIMED project which is dedicated to basin-scale modeling of the Mediterranean Sea. It also belongs to the MERMEX program which aims at studying biogeochemical cycles in the Mediterranean Sea and their evolution. The first step of this work was to couple the hydrodynamic modeling platform (NEMO) to the mechanistic biogeochemical modeling platform (Eco3M). We ran a simulation (2000-2012) using the hydrodynamic outputs from NEMO-MED12 configuration to force the biogeochamical model Eco3M-MED. The model evaluation was conducted using numerous field measurements (chlorophyll, nutrients, primary production, etc.). The simulation strengthens and extends to the whole basin the prominent role of dissolved organic carbon in the biological carbon pump in the whole Mediterranean Sea. A comprehensive analysis of organic carbon (particulate and dissolved) production processes production was performed. Results reveal that particulate organic carbon production is restricted to the highly dynamic areas, whereas dissolved organic carbon accumulation in the surface layers is a common process in much areas of the basin. This latter process appeared to dependant on the cellular contents of phytoplancton and heterotrophic bacteria, themselved being controled by low phosphate availability. Finally, the dissolved organic carbon contribution to carbon export is around 64 % in the Western basin, and up to 90 % in the Eastern basin. When taking into account the dissolved fraction, total organic carbon export in the Eastern basin -despite its higher oligotrophy- exceeds the one in the Western basin (60% against 40 %).

No presente trabalho é apresentado um conjunto de medidas de seção de choque do processo de fusão nuclear nos sistemas ANTPOT. 12 C + ANTPOT. 63, 65 CU, em um intervalo de energia correspondente à 0.91.8 vezes a barreira coulombiana. O método experimental utilizado foi a detecção dos resíduos de evaporação através da técnica de tempo de vôo, acoplado à um defletor eletrostático para separar as partículas provenientes do espalhamento elástico dos produtos de fusão. As vantagens e limitações do método são discutidas em detalhes. As funções de excitação de fusão dos sistemas citados foram analisadas via modelo de penetração de barreira unidimensional , utilizando vários potenciais nucleares. Dessa análise comprovou-se que a secção de choque de fusão é subestimada em energias em torno e abaixo da barreira coulombiana. Nesse sentido foi feito um cálculo, utilizando o código computacional CCFUS, que simula o acoplamento de canais, com a finalidade de descobrir quais canais contribuem para o processo de fusão nuclear em baixas energias. As seções de choque de fusão obtidas foram comparadas com outros sistemas - ANTPOT. 12 C + ANTPOT. 63, 65 CU.- que utilizam os mesmos núcleos-alvo. A comparação revelou que os sistemas ANTPOT. 12 C + ANTPOT. 63, 65 CU possuem o maior aumento de seção de choque de fusão subcoulombiana e esse fato pode estar relacionado à deformação estática do projétil. A comparação das distribuições da velocidade dos resíduos de evaporação para os dois de sistemas revelou um canal intenso para o sistema ANTPOT. 12 C + ANTPOT. 63 CU que não está presente para o outro sistema. / In this work we present cross-section measurements for nuclear fusion in the 12C+ 63,65 Cu system, at 12C energy range from 0.9 to 1.8 times the Coulomb barrier. In order to detect and to obtain the mass identification of the evaporation residues following the fusion process, the time of flight method was adopted in conjunction with an eletrostatic deflector capable of separating the evaporation residues from the beam particles. The limitation and advantadges of this method of measurement are discussed. The excitation functions for the above systems were analysed using the unidimensional barrier penetration model with different nuclear potentials. Theoretical fusion cross-section values obtained from this analysis were systematically smaller than our measured values, in the energy region below the Coulomb barrier. In order to discover which channel enhances the fusion cross-section in this region, a coupled channel calculation was performed, with the aid of CCFUS code. The experimental data for the above reactions were compared with the systems 12C+ 63,65 Cu, measured by our group. In this comparison, It was noted that the systems 12C+ 63,65 Cu have greater fusion cross section below the Coulomb barrier. This fact can be related to the static deformation of 12C. The comparison of velocity spectra of the evaporated residues for the two systems shows that 12C+63Cu has a strong reaction channel that was not present in the 12C+ 65Cu system.