Toute surface naturelle est par essence non lisse, elle est constituée de rugosités plus ou moins régulières et / ou de structures mobiles d’échelles variées. D’un point de vue mécanique des fluides, ces surfaces naturelles proposent des meilleures performances aérodynamiques en terme de réduction de traînée, d’augmentation de la portance ou de contrôle du décollement lorsqu’elles couvrent des corps en mouvement. Cela a été notament prouvé pour des écoulements de couches limites ou de sillage, autour de corps épais. La simulation numérique d’écoulements aux échelles microscopiques autour des surfaces « naturelles » demeure de nos jours encore hors de portée. En conséquence, la thèse a pour objet d’étudier la modélisation du glissement apparent de l’écoulement sur ce genre de surface, modélisée comme un milieu poreux, appliquant la théorie de la moyenne-volumique de Whitaker. Ce modèle mathématique permet globalement de représenter en moyenne les détails de la micro-structure de ses surfaces, tout en conservant une description satisfaisante des phénomènes physiques induits par l’écoulement. Le premier chapitre de ce manuscrit dresse un panorama des efforts antérieurs portant sur la modélisation de ces surfaces en précisant les résultats les plus importants issus de la littérature. Le deuxième chapitre présente la dérivation mathématique des équations de Navier- Stokes en moyenne volumique (VANS en anglais) dans un milieu poreux. Dans le troisième chapitre est étudiée la stabilité de l’écoulement à l’interface entre un fluide libre et un milieu poreux, constitué d'une série de cylindres rigides. La présence de cette couche poreuse est traitée par un terme de traînée dans les équations du fluide. On montre que l'ajout de ce terme réduit les taux d’amplification de l’instabilité de Kelvin-Helmholtz sur toute la gamme des nombre d’onde et ainsi augmente la longueur d’onde du mode le plus amplifié. Dans ce même contexte a été calculée la différence entre un modèle isotrope et une approche tensorielle pour le terme de traînée, afin de déterminer l’approche la plus consistante pour une étude de stabilité de ce type d’écoulement. Cela a mené à la conclusion que le modèle le plus pertinent est celui utilisant le tenseur de perméabilité apparent. Dans le chapitre suivant le tenseur de perméabilité apparent est identifié sur la base d’une centaine de simulations numériques directes, pour un milieu poreux tridimensionnel constitué de cylindres rigides, où le problème de fermeture est abordé par la méthode VANS. Dans ces configurations ce tenseur varie en fonction de quatre paramètres : le nombre de Reynolds, la porosité et l’orientation du gradient moyen de pression définie par deux angles d’Euler. Cette paramétrisation permet de capturer les effets tridimensionnels locaux. Cette base de données ainsi constituée a permis de créer, une approche de type kriging, un métamodèle comportemental pour estimer toutes les composantes du tenseur de perméabilité apparente. Dans le cinquième chapitre sont menées des simulations des équations VANS à l’échelle macroscopique après implémentation du méta-modèle qui autorise des temps de calcul raisonnables. La validation de l’approche à l’échelle macroscopique est effectuée sur un écoulement dans une cavité fermé couverte d’une couche poreuse et une comparaison avec les résultats d’un DNS très précise, homogénéisés a posteriori montre un très bon accord et démontre la pertinence de la démarche. L’étape suivante a consisté en l’étude du contrôle du décollement pour un écoulement autour d’une bosse poreuse par cette même approche VANS macroscopique. Enfin des conclusions générales et des directions de recherche possibles sont présentées dans le dernier chapitre. / Any natural surface is in essence non-smooth, consisting of more or less regular roughness and/or mobile structures of different scales. From a fluid mechanics point of view, these natural surfaces offer better aerodynamic performances when they cover moving bodies, in terms of drag reduction, lift enhancement or control of boundary layer separation; this has been shown for boundary layer or wake flows around thick bodies. The numerical simulation of microscopic flows around "natural" surfaces is still out of reach today. Therefore, the goal of this thesis is to study the modeling of the apparent flow slip occurring on this kind of surfaces, modeled as a porous medium, applying Whitaker's volume averaging theory. This mathematical model makes it possible to capture details of the microstructure while preserving a satisfactory description of the physical phenomena which occur. The first chapter of this manuscript provides an overview of previous efforts to model these surfaces, detailing the most important results from the literature. The second chapter presents the mathematical derivation of the volume-averaged Navier-Stokes equations (VANS) in a porous medium. In the third chapter the flow stability at the interface between a free fluid and a porous medium, formed by a series of rigid cylinders, is studied. The presence of this porous layer is treated by including a drag term in the fluid equations. It is shown that the presence of this term reduces the rates of amplification of the Kelvin-Helmholtz instability over the whole range of wavenumbers, thus leading to an increase of the wavelength of the most amplified mode. In this same context, the difference between the isotropic model and a tensorial approach for the drag term has been evaluated, to determine the most consistent approach to study these flow instabilities. This has led to the conclusion that the model that uses the apparent permeability tensor is the most relevant one. In the following chapter, based on this last result, the apparent permeability tensor, based on over one hundred direct numerical simulations carried out over microscopic unit cells, has been identified for a three-dimensional porous medium consisting of rigid cylinders. In these configurations the tensor varies according to four parameters: the Reynolds number, the porosity and the direction of the average pressure gradient, defined by two Euler angles. This parameterization makes it possible to capture local three-dimensional effects. This database has been set up to create, based on a kriging-type approach, a behavioral metamodel for estimating all the components of the apparent permeability tensor. In the fifth chapter, simulations of the VANS equations are carried out on a macroscopic scale after the implementation of the metamodel, to get reasonable computing times. The validation of the macroscopic approach is performed on a closed cavity flow covered with a porous layer and a comparison with the results of a very accurate DNS, homogenized a posteriori, has shown a very good agreement and has demonstrated the relevance of the approach. The next step has been the study of the passive control of the separation of the flow past a hump which is placed on a porous wall, by the same macroscopic VANS approach. Finally, general conclusions and possible directions of research in the field are presented in the last chapter.

[pt] Devido às suas aplicações nas mais diversas áreas do conhecimento humano e em atividades vitais para o mundo moderno, como a prospecção e produção de petróleo, o escoamento de fluidos em meios porosos tem sido objeto de inúmeros estudos experimentais e teóricos. Por exemplo, na recuperação de campos de petróleo, depara-se com o desafio de prever o comportamento de fluidos a serem injetados na formação (gás, água ou águacom adição de polímero), com o objetivo de aumentar a pressão e otimizar a recuperação de óleo. A abordagem adequada desta questão, implica na necessidade de profundo conhecimento do meio poroso e do comportamento dos fluidos envolvidos, bem como na utilização de modelos experimentais e/ou teóricos que propiciem simulações desses processos com razoável fidelidade. Neste trabalho o trajeto do fluido entre os espaços vazios de um meio poroso é representado por um capilar com contração/expansão (garganta). A partir de relações geométricas convenientemente definidas, onde apenas um parâmetro varia, foram geradas oito diferentes geometrias de análise. Foram feitas simulações numéricas de fluido newtoniano e pseudo-plástico utilizando-se código CFD desenvolvido na PUC-Rio. São estabelecidas relações entre o escoamento em capilares com garganta e o escoamento em capilares equivalentes, com raio constante. São apresentadas e comparadas duas metodologias, determinando raios equivalentes analíticos e ajustados (Req an e Req aj). Para fluidos newtonianos, a aproximação é muito boa em todas as geometrias, sendo a maior diferença 6,2%. Para fluido power law a diferença também é satisfatória, embora maior. / [en] Due to its application on a large number of fields and activities which are vital for the modern world, being oil prospection one of them, the flow of fluids on porous media have been studied in many experimental works and theoretical pieces. For instance, regarding the recovery of oil fields, it is challenging to predict the behavior of fluids that will be injected on the formation (gas, water or water added with polymer), with the objective to increase the pressure and optimize the oil recuperation. To approach this matter adequately a deep knowledge of the porous media and the behavior of the fluid in question are required, in addition to the use of the experimental and/or theorical models which provide simulations of this process with reasonable fidelity. On this experiment the path of the fluid through the empty spaces of a porous medium is represented by a capillary with contraction/expansion (throat). Starting with the conveniently defined geometrical relations where only one parameter varies, we generated eight different analysis geometries. Numerical simulations of the Newtonian and Pseudo-Plastic fluids were done using the CFD code developed at PUC-Rio. Relationships were established between the flow on capillaries with throat and equivalent capillaries, with constant radius. Two methodologies are presented and compared, determining equivalent analytical radius and adjusted (Req an e Req aj). For the case of Newtonians fluids, the approximation works very well on all geometries, and the largest difference is 6.2%. Although larger, the approximation still works in a satisfactory manner when dealing with Power Law fluids.

Este projeto teve como objetivo a construção e automatização de um espectrômetro de baixo custo que tem como principal aplicação a caracterização de meios porosos. Para realizar a caracterização de um meio poroso é necessário que o campo magnético externo aplicado não seja intenso para diminuir os problemas relacionados com a diferença de susceptibilidade magnética na interface fluido-solido. Por este motivo, foi construído um magneto resistivo de baixo campo, operando na frequência de próton de 2MHz (0.047 Tesla) e de simples design. Para manter o equipamento com custo mais baixo possível foram construídos em laboratório a fonte de alimentação do magneto, um amplificador ajustado para 2MHz e os componentes λ/4 e os softwares necessários. O software desenvolvido tem a finalidade de controlar a placa de geração e aquisição de radiofrequência que foi adquirida pela empresa Spincore, denominada RadioProcessor. Para esta finalidade optou-se pela construção de uma nova API na linguagem C++, utilizando a API original do RadioProcessor como base, de forma a montar um conjunto de funções simples e intuitivas para a utilização do espectrômetro. Além disso, foi realizada a integração de uma linguagem script à biblioteca, que dá acesso ao usuário a todas as funções básicas de programação, permitindo que um usuário com pouca habilidade de programação possa criar um programa para o controle total do equipamento de forma simples e também tornando o software maleável, sendo possível que novas funções sejam integradas. Isso abre espaço para que o equipamento possa ser utilizado para outras aplicações além das citadas anteriormente. Somada à nova biblioteca, uma interface gráfica foi desenvolvida, a qual compreende o programa de controle do equipamento e também funções para tratamento e visualização de dados, como a transformada inversa de Laplace e transformada de Fourier. / The objective of this project was the assembling and automatization of a low cost NMR spectrometer for the characterization of porous media. The study of porous media through NMR spectroscopy requires a low intensity magnetic field in order to avoid the problems related to the difference of magnetic susceptibility on the fluid-solid interface. For that reason, a low field electromagnet, operating on the proton frequency of 2MHz (0.047 Tesla) and of simple design was built. As to keep all the equipment as low cost as possible, the following components were also assembled in laboratory: the magnets power supply, a 2MHz amplifier, the quarter-wave components and the necessary softwares. The goal of the software developed was to control the RF generation and acquisition board, which was acquired from the enterprise Spincore, further named RadioProcessor. To achieve this goal, it was decided the development of a new API, in the C++ language, which uses the original boards API as a base in order to assemble a group of functions that are simpler and more intuitive to the end user. In addition, a script language was integrated to the APIs library, allowing users with little programming knowledge to write a program which controls the whole equipment in an easy way and also adding flexibility to the software, as it becomes easy to add new functions. This permits that the equipment be easily changed to include other applications in the future. Finally, a graphical interface was developed, which comprehends the equipment control software and the visualization of the data.

Les matelas fibreux à base de fibres de verre produits par voie humide ont des applications industrielles en tant que séparateur de batterie et matériaux d’isolation (cœur de panneaux d’isolation sous vide). Ces matériaux utilisent principalement des fibres submicroniques, relativement chères et pouvant présenter des risques pour la santé. Ce projet est une contribution à l'élaboration d'un procédé de fabrication par voie humide de matelas de fibres de verre visant à valoriser des fibres plus grossières, le produit final devant respecter un cahier des charges précis. Nous avons été amenés à étudier le comportement des fibres de verre dans les différentes étapes du procédé et à caractériser les matelas résultants. Nous nous sommes intéressés en particulier au comportement physico-chimique des suspensions aqueuses de fibres de verre. Nous avons caractérisé les propriétés de contexture, la résistance mécanique en traction, la compressibilité et la conductivité thermique des matelas fibreux. Les travaux expérimentaux nous ont permis de proposer une formulation de la composition optimisée et des conditions opératoires du procédé afin que le matelas final soit conforme au cahier des charges. Cette nouvelle composition intègre des fibres de renfort en faible quantité. Elle permet d’améliorer les caractéristiques mécaniques sans affecter les autres propriétés.Enfin, nous avons quantifié les coûts de production et les avons comparés à ceux du procédé actuellement utilisé avec des fibres grossières (production par voie sèche).Mots-clés : Génie papetier, physico-chimie, milieu poreux, fibres de verre, caractérisation. / Glass fibre - based mats produced by a wetlaid process have industrial applications as a battery separator and insulation materials (core of vacuum insulation panels). These materials are mainly made with sub-micron fibres which relatively expensive and can present a risk to health. This project is a contribution to the production of glass fibre-based mats by a wet-laid process to add value to coarser fibres, the final product should respect precise specifications. We have been led to study the behavior of glass fibers in different stages of the process and to characterize the resulting mats. We investigated especially the physico-chemical behavior of aqueous suspensions of glass fiber. We have characterized structure properties, the mechanical resistance to traction, the compressibility and the thermal conductivity of fibrous mats. The experimental work has allowed us to give a formulation of the optimized composition and operational conditions of the process so that the final mattress conforms to the specifications. This new composition includes reinforcement fibres in small quantities. It allows improving the mechanical characteristics without affecting the other properties. Finally, we quantified the production costs and compared them to those of the process currently used with coarse fiber (dry-laid production).Keywords: Paper engineering, physical chemistry, porous media, glass fibre, characterization.

The objective for this master thesis project was to perform simulations for high and low Reynolds number for a porousmedia modelled as quadratic packed cylinders. The simulations were to be performed by using the open source codeOpenFOAM and the turbulence should be modelled by the LES turbulence model in order to resolve details about theturbulence. In Hellström et al. (2009) simulations were performed for a similar setup but for the k-Omega SST turbulencemodel. In this project simulations were performed not only for LES but also with the k-Omega SST model and for the laminar description in order to validate the LES problem setup. For the simulations Reynolds number was calculatedbased on the porosity and the cylinder diameter and was denoted Re’. The simulations were performed for a set of Re’and for each Re’ the permeability was calculated based on Darcy’s law. The calculations of the permeability showed,in line with the results from Hellström, that at Re’ around 20 there is a drop in the permeability both for the LESand SST simulations. From the permeability simulations it was also seen that for low Re’ flows the LES and laminardescriptions agreed with each other, but that the SST simulations yielded lower values of the permeability. Accordingto Hellström the permeability drop was due to onset of inertia effects. When looking at the resulting flow field plots itgave the idea that the amount of turbulence could also have influence on the resulting permeability.

Este trabalho foi desenvolvido utilizando-se rochas reservatório siliciclásticas provenientes de afloramentos brasileiros, pertencentes ao Grupo Guaritas (Caçapava do Sul e Pinheiro Machado - RS) e à Formação Pirambóia (Rio Claro - SP), com o objetivo de se obter suas características petrofísicas. As amostras são provenientes de diferentes fácies eólicas e pluviais, com características semelhantes a reservatórios de água e hidrocarbonetos. Foram utilizados experimentos de Ressonância Magnética Nuclear (RMN), intrusão de mercúrio, susceptibilidade magnética, análise de lâmina petrográfica e Espectroscopia de Fluorescência de Energia Dispersiva de Raios-X (EDX) para a análise das amostras. Apesar de serem provenientes de sistemas deposicionais similares, a correlação entre os resultados demonstrou que a diagênese tem importante papel no desenvolvimento dos reservatórios, influenciando na porosidade e permeabilidade das amostras. Foi demonstrado que a susceptibilidade magnética possui correlação com as quantidades de Fe 203, Al 203 e SiO2, onde o Al 203 está relacionado com a quantidade de argila das amostras (aluminossilicatos). Através dos resultados de RMN foi possível estimar a susceptibilidade magnética das amostras através da largura de linha espectral em baixo campo. As estimativas de porosidade e tamanhos de poros via RMN demonstraram compatibilidade com as análises das lâminas e experimentos de intrusão de mercúrio. Com os experimentos de RMN bidimensionais T1xT2, foi mostrado que há uniformidade na relaxação, devido aos mapas de correlação serem aproximadamente paralelos à diagonal dos mapas 2D. Nos experimentos DxT2 há a correlação entre a distribuição de tempos de relaxação e coeficientes de difusão, demonstrando que, em muitos casos, há restrição de difusão translacional das moléculas do fluido. Para estimar a permeabilidade das rochas foi utilizado o modelo desenvolvido por pesquisadores da Schlumberger Doll Research (SDR), amplamente aplicado na previsão da permeabilidade através de resultados obtidos pela RMN. Aplicando-o de maneira clássica, os resultados não foram plenamente satisfatórios, em função das características magnéticas das rochas estudadas. O coeficiente relaxatividade superficial foi calculado pela convolução entre a distribuição de tamanhos de poros da intrusão de mercúrio e a distribuição de tempos de relaxação transversal. Este parâmetro foi utilizado como correção ao modelo SDR, entretanto, os resultados obtidos continuaram não satisfatórios. Foi proposta então a correção através da susceptibilidade magnética, tanto medida pelo susceptômetro quanto estimada pela RMN. As utilizações dessas duas informações culminaram em melhores resultados para o valor das permeabilidades, com destaque para a RMN, desta maneira, propondo um novo método para estimar esse importante parâmetro das rochas reservatório. Com estes resultados podemos verificar a coesão entre os aspectos geológicos das amostras e a RMN, porém, as técnicas de RMN existentes empregadas ao estudo devem ainda ser aperfeiçoadas e novas devem ser desenvolvidas para o estudo de rochas reservatório que apresentam alta susceptibilidade magnética, um problema comum encontrado no estudo de rochas sedimentares por RMN. / The petrophysical characterization of siliciclastic reservoir rocks from Brazilian outcrops was the objective of this work. The samples were obtainded from different eolian and pluvial facies of the Guaritas Group (Caçapava do Sul and Pinheiro Machado - RS) and Pirambóia Formation (Rio Claro -SP). Different methods were applied in order to study these samples: Nuclear Magnetic Resonance (NMR), mercury injection, magnetic susceptibility, petrographic thin sections analysis and Energy Dispersive X-ray Fluorescence Spectroscopy (EDX). Despite the depositional similarities, the results indicate that permeability and porosity are affected by diagenetic processes which show that the diagenesis plays an important role at the development of these reservoirs. It was shown the correlation between magnetic susceptibility and the amounts of Fe 2O3 , Al 203 and Si0 2 were the amount of Al 203 is related with the quantity of clay (aluminosilicates). Magnetic susceptibilities were estimated using the linewidth obtained from NMR spectra (by FFTs of the FIDs) measured at low and homogeneous magnetic fields, obtaining good agreement with the bulk magnetic susceptibility. Porosity and pore size distributions determined by NMR, mercury injection and petrographic analysis were compatible. In the two-dimensional NMR experiment T 1xT2 a diagonal behavior of relaxation time distributions was observed, indicating uniformity of relaxation mechanisms inside the samples. With DxT 2 measurements, the transverse relaxation times were correlated with diffusion coefficients showing restricted diffusion in most samples. In order to predict permeability the model proposed by researchers from Schlumberger Doll Research (SDR), which uses NMR parameters, was applied. With the classical permeability SDR predictor the obtained results were not fully satisfactory, as consequence of the magnetic characteristics of the samples. In order to correct this predictor, surface relaxivity coefficients were determined by the superposition of T2 and mercury injection pore diameter distributions nevertheless, these results were again unsatisfactory. Thus, a new correction method for the SDR permeability predictor was proposed using NMR and bulk magnetic susceptibilities. This approach improved considerably the permeability prediction, especially when using NMR-derived magnetic susceptibilities. The consistency between NMR and geological properties of the samples was verified nevertheless, the used NMR techniques still need to be improved and new ones must be developed, in order to allow the study of reservoir rocks with high magnetic susceptibility.

In this thesis, we consider four different problems in membrane filtration, using a different mathematical approach in each instance. We account for the fluid-driven deformation of a filtercake using nonlinear poroelasticity in Chapter 2. By considering feeds with very high and very low particle concentrations, we introduce a quasi-static caking model that provides a suitable approximation to the full model for the physically realistic concentration regimes. We illustrate the agreements and differences between our model and the existing conventional cake-filtration law. In Chapter 3, we introduce a stochastic model for membrane filtration based on the quantised nature of the particles and show how it can be applied for feeds with different particle types and membranes with an interconnected pore structure. This allows us to understand the relation between the effects of clogging on the level of an individual pore and on the macroscopic level of the entire membrane. We conclude by explaining the transition between the discrete and continuous model based on the Fokker--Planck equation. In Chapter 4, we consider the inverse problem of determining the underlying filtration law from the spreading speed of a particle-laden gravity current. We first couple the theory of gravity currents with the stochastic model developed in Chapter~3 to determine a filtration law from a given set of experiments. We then generalise this idea for the porous medium equation, where we show that the position of the front follows a power law for the conventional filtration laws, which allows us to infer the clogging law in certain instances. We conclude the thesis by showing in Chapter 5 how we can combine experimental measurements for the clogging of a depth filter and simple fluid dynamics to accurately predict the pressure distribution in a multi-capsule depth filter during a filtration run.

With conventional oil reserves in decline, energy companies are increasingly turning to enhanced oil recovery (EOR) processes to extend the productive life of oilfield wells. Laboratory-scale core floods, in which one fluid displaces another from the pore space of a rock core, are widely used in petroleum research for oilfield evaluation and screening EOR processes. Achieving both macro- and pore-scale understandings of such fluid displacement processes is central to being able to optimise EOR strategies. Many of the mechanisms at play, however, are still poorly understood. In this thesis nuclear magnetic resonance (NMR) has been used for quantitatively, non-invasively and dynamically studying laboratory core floods at reservoir-representative conditions. Spatially-resolved relaxation time measurements (L-T1-T2) have been applied to studying a special core analysis laboratory (SCAL) protocol, used for simulating reservoir oil saturations following initial oil migration (primary drainage) and characterising core samples (capillary pressure curves). Axial heterogeneities in pore filling processes were revealed. It was demonstrated that upon approaching irreducible water saturation, brine saturation was reduced to a continuous water-wetting film throughout the pore space; further hydrocarbon injection resulted in pore pressure rise and wetting film thinning. L-T1-T2 techniques were also applied to a xanthan gum polymer-EOR flood in a sandstone core, providing a continuous measurement of core saturation and pore filling behaviours. A total recovery of 56.1% of the original oil in place (OOIP) was achieved, of which 4.9% was from xanthan. It was demonstrated that deposition of xanthan debris in small pores resulted in small-pore blocking, diverting brine to larger pores, enabling greater oil displacement therein. L-T1-T2, spectral and pulsed field gradient (PFG) approaches were applied to a hydrolysed polyacrylamide (HPAM)-EOR flood in a sandstone core. A total recovery of 62.4% of OOIP was achieved, of which 4.3% was from HPAM. Continued brine injection following conventional recovery (waterflooding) and EOR procedures demonstrated most moveable fluid saturation pertained to brine, with a small fraction to hydrocarbon. Increases in residual oil ganglia size was demonstrated following HPAM-EOR, suggesting HPAM encourages ganglia coalescence, supporting the "oil thread/column stabilisation" mechanism proposed in the literature. NMR relaxometry techniques used for assessing surface interaction strengths (T1/T¬2) were benchmarked against an industry-standard SCAL wettability measurement (Amott-Harvey) on a water-wet sandstone at magnetic field strengths comparable to reservoir well-logging tools (WLTs). At 2 MHz, T1/T2 was demonstrated to be weakly sensitive to the core wettability, although yielded wettability information at premature stages of the Amott-Harvey cycle. This suggests the potential for NMR to deliver faster wettability measurements, in SCAL applications or downhole WLT in situ reservoir characterisation.

As moléculas que compõem um fluído, devido à autodifusão, encontram-se em constante movimento de translação num meio poroso. Observar a migração dessas moléculas de um poro para outro é fundamental para análise das conectividades do meio, resultado importante para a prospecção de óleo e gás, qualidade de membranas filtros, entre outros processos. Se há distintos tempos de relaxações transversais, T2, para os poros, ocorre quando estes apresentam características físico-químicas ou tamanhos diferentes, um experimento capaz de observar esse efeito de troca entre poros é o chamado T2-T2 Exchange proposto em 1993. Esse experimento, na sua versão original, corresponde a um experimento tridimensional que consome muito tempo para ser executado, inviabilizando, por exemplo, o uso em perfilagem de poços petrolíferos. Desta forma, o grupo propôs diminuir o tempo de execução reduzindo uma dimensão do experimento. Este novo método, novo experimento, foi nomeado de T2-Filtered T2-T2 Exchange, o qual utiliza a primeira parte da sequência de pulsos original como um filtro de T2. Com o objetivo de validar o experimento proposto, surgiu a oportunidade de estudar a técnica T2-T2 Exchange como um todo. No decorrer do trabalho, surgiu mais uma nova aplicação da técnica T2-T2 Exchange, a execução simultânea. Tanto a versão tridimensional do T2-T2 Exchange quando a bidimensional, necessitam de uma ciclagem de fase especial. Neste contexto, foi proposto também a execução simultânea dos experimentos T2-Filtered T2-T2 Exchange e Saturation-Recovery-CPMG, adquirindo os sinais separados e empregando a ciclagem de fases distintas. A fim de validar as propostas T2-Filtered T2-T2 Exchange e execução simultânea, se optou por uma amostra padrão, uma cerâmica de alumina foi saturada com água e, então, estudado os núcleos de hidrogênios. A cerâmica foi manufaturada pelo método de prensagem a seco com adição de agentes porogênicos e sinterização, a qual foi caracterizada via porosimetria por intrusão de mercúrio e imagens por microscópio eletrônico de varredura. Em um campo de 2 tesla, frequência de 85 MHz para os núcleos de hidrogênios, foram executados os experimento Saturation-Recovery-CPMG, T2-T2 Exchange e T2-Filtered T2-T2 Exchange. A análise comparativa entre os mapas de correlação T1-T2 e as curvas de trocas obtidos deste, permitiu validar as propostas. Desta forma, se concluiu que a nova sequência T2-Filtered T2-T2 Exchange proposta, é uma poderosa ferramenta com potencial para aplicação em well-logging. / The molecules that make up a fluid inside a porous media are in constant motion due to self diffusion. Observing this migration inside the media is very interesting for analyzing the connectivity between different regions of the sample, which is of great importance to the field of oil and gas prospection. In the case the sample has pores with distinct transverse relaxation times, T2, which happens when they have distinct physical-chemical properties or sizes, an experiment capable of observing the movement from one distinct pore from the other is the T2-T2 Exchange proposed in 1993. This experiment, in its original version, corresponds to a three dimensional experiment which usually takes a very long time, making it unfeasible for well logging for example. Thus, this work proposes a method to drastically reduce the duration of the experiment by eliminating one of its dimensions. It was named T2-Filtered T2-T2 Exchange. It uses the first part of the original sequence as a T2 filter. To validate the filtered technique, the T2-T2 Exchange technique was vastly studied as a whole. Some phase cycling characteristics of the techniques were observed and in the end led to the development of new proposals. The proposal consists of simultaneously performing both the T2-Filtered T2-T2 Exchange and the Saturation-Recovery-CPMG. For the experimental tests of the sequences, a water saturated alumina ceramic sample was chosen as the standard sample. The ceramic was manufactured by a process of dry pressing with addition of some porogenic agents and followed by sinterization. The sample was characterized by mercury injection capillary pressure and scanning electron microscopy. The NMR experiments were performed on a 2 Tesla magnet corresponding to an 85 MHz frequency for protons and the experiments performed wereT2-T2 Exchange, T2-Filtered T2-T2 Exchange and Saturation-Recovery-CPMG. The comparative analysis between the T1-T2 correlation and the exchange curves obtained allowed the validation of the proposals. It was possible to conclude that the T2-Filtered T2-T2 Exchange technique is a powerful tool with potential application in the well-logging field.

Saline water evaporation from porous media with the associated salt precipitation patterns is frequently observed in a number of industrial and environmental applications and it is important in a variety of topics including, but not limited to, water balance and land-atmosphere interaction, terrestrial ecosystem functioning, geological carbon storage, and preservation of historical monuments. The excess accumulation of salt in soil is a global problem and is one of the most widespread soil degradation processes. Thus, it is important to understand the dominant mechanisms controlling saline water evaporation from porous media. This process is controlled by the transport properties of the porous medium, the external conditions, and the properties of the evaporating fluid. During saline water evaporation from porous media, the capillary induced liquid flow transports the solute towards the evaporation surface while diffusive transport tends to spread the salt homogeneously thorough the porous medium. Therefore, the solute distribution is influenced by the competition between the diffusive and convective transport. As water evaporates, salt concentration in the pore space increases continually until it precipitates. The formation of precipitated salt adds to the complexity of the description of saline water evaporation from porous media. In this dissertation, the effects of salt concentration, type of salt, and the presence of precipitated salt, on the evaporation dynamics have been investigated. The obtained results show that the precipitated salt has a porous structure and it evolves as the drying progresses. The presence of porous precipitated salt at the surface causes top-supplied creeping of the evaporating solution, feeding the growth of subsequent crystals. This could be visualized by thermal imaging in the form of appearance and disappearance of cold-spots on the surface of the porous medium, brought about by preferential water evaporation through the salt crust. My results show that such a phenomenon influences the dynamics of saline water evaporation from porous media. Moreover, a simple but effective tool was developed in this dissertation capable of describing the effects of ambient temperature, relative humidity, type of salt and its concentration, on the evaporative fluxes. Additionally, pore-scale data obtained by synchrotron x-ray tomography was used to study ion transport during saline water evaporation from porous media in 4D (3D space + time). Using iodine K-edge dual energy imaging, the ion concentration at pore scale with a high temporal and spatial resolution could be quantified. This enabled us to reveal the mechanisms controlling solute transport during saline water evaporation from porous media and extend the corresponding physical understanding of this process. Within this context, the effects of particle size distribution on the dispersion coefficient were investigated together with the evolution of the dispersion coefficient as the evaporation process progresses. The results reported in this dissertation shed new insight on the physics of saline water evaporation from porous media and its complex dynamics. The results of this dissertation have been published in 3 peer-reviewed journal papers together with one additional manuscript which is currently under review.