[pt] O escoamento de bolhas de gás em fluidos não-newtonianos é muito importante em diversas aplicações, tais como biorreatores, processamento de alimentos e operações de perfuração e cimentação de poços de petróleo, contudo poucos estudos abordam este tema. No presente trabalho é analisado o escoamento ascendente de bolhas de gás em fluido viscoplástico, modelado aqui como fluido Herschel-Bulkley. Utilizou-se uma abordagem bidimensional, aproximada por placas paralela. O escoamento multifásico foi numericamente simulado fazendo uso do método de volumes finitos VOF (volume of fluid), o qual resolve as equações de conservação de massa e momento aclopado a uma equação para a fração volumétrica dos fluidos. A influência de fatores tais como tensão limite de escoamento, dimensão da bolha, número de bolhas e distância entre as bolhas escoando em um fluido viscoplástico foram investigadas. Os resultados indicaram que a tensão limite de escoamento tem grande impacto na velocidade de deslocamento da bolha. No caso de mais de uma bolha escoando foi também observado que o deslocamento de uma bolha altera a velocidade de ascenção das outras, fazendo com que elas se unam, e a medida em que a distância entre as bolhas aumenta a interferência é eliminada e as mesmas escoam como bolhas individuais. Além disso, foi verificado que, dependendo do tamanho da bolha escoando, a parede interfere na sua velocidade de ascenção. Por fim foi analisada a mudança no formato da bolha, podendo-se observar que em número de Reynolds menores a mesma apresenta um formato esférico e a medida em que este parâmetro aumenta, a bolha vai se deformando e adquire uma forma elipsoidal. Os resultados qualitativos do presente estudo numérico foram comparados com alguns trabalhos experimentais encontrados na literatura e corresponderam relativamente bem. / [en] The gas bubble displacement in non-Newtonian fluids is important in many applications, including bioreactors, food processing and drilling and cementing of oil wells however a few studies have investigated this issue. The motion of gas bubbles in a viscoplastic fluid, modeled as a Herschel-Bulkley fluid, is analysed in the present work. A bidimensional approach was used, approximated by parallel plates. The multiphase flow was numerically simulated using the finite volume technique VOF (volume of fluid), which solves the conservation equations of mass and momentum coupled to an equation for the volume fraction of the fluids. Is was investigated the influence of factors such as fluid yield stress, bubble size, number of bubbles rising in the viscoplastic fluid and also the distance between them. The results indicated that the yield stress has great impact on the bubble rising velocity. In the case of multiple bubbles flowing it was also observed that the displacement of one bubble influences the rising velocity of the others, causing them to join together. As the distance between the bubbles increase the interference is eliminated and the bubbles flow separately. Furthermore, it was found that there was an wall interference depending on the size of the bubble. Lastly, it was analysed the change in the shape of the bubble. It could be observed that for lower Reynolds number the bubbles presented an spherical shape and as this parameter increases it turns into an elipsoidal shape. The qualitative results of the present study agreed relatively well with experimental works found in the literature.

I present a spectroscopic study of the molecular gas in the Whirlpool Galaxy (M51) and the Antennae Galaxies (NGC 4038/39) using data from the Herschel Space Observatory (Herschel) and the Atacama Large Millime- ter/submillimeter Array (ALMA). Using data from the Herschel Spectral and Photometric Imaging REceiver (SPIRE) Fourier Transform Spectrometer (FTS), I perform an excitation analysis to determine the physical characteris- tics (temperature, density, column density) of the cold and warm molecular gas across both systems. I do not find significant variation in the cold molecular gas across an individual system or between the systems. The warm molecular gas temperature is greater in NGC 4038/39 than in M51, while the density in both M51 and the nucleus of NGC 4038 is greater than the rest of the Anten- nae system. Both galaxies exhibit a similar fraction of warm to cold molecular gas. I compare Herschel SPIRE-FTS data to models of photon dominated regions (PDRs) to determine the strength of the background far ultraviolet field (G0) within both systems and find little variation across each system. I find that PDRs alone can explain the observed Herschel SPIRE-FTS data in both systems. Using ALMA observations of dense molecular gas tracers in NGC 4038/39, I investigate the physical processes affecting the dense molecular gas. Ratios of various molecular gas tracers suggest that the contributions of mechanical heating relative to PDR heating are similar across the entire system. The dense gas fraction in the nucleus of NGC 4038 and the nucleus of NGC 4039 is higher than in the overlap region, which I attribute to an increase in the stellar potential within the two nuclei. Furthermore, I find evidence for an increased cosmic ray rate in the overlap region of NGC 4038/39 relative to the two nuclei, which I attribute to an increased supernova rate in the overlap region. Most of the molecular gas in M51 and NGC 4038/39 is in the form of PDRs, while the increased temperature of the warm molecular gas in NGC iii 4038/39 compared to M51 is likely due to an increase in the mechanical heating from both supernova and stellar winds and the ongoing merger. Furthermore, a comparison of these results to previous studies of the interacting galaxy M82 and the late-stage merger Arp 220 suggests that mergers and interactions have a greater effect on the warm molecular gas compared to the cold molecular gas. The results from this thesis help to further our understanding of the effects of merging and interacting galaxies on molecular gas, while helping understand differences between interacting galaxies and merging galaxies. / Thesis / Doctor of Philosophy (PhD)

Excessive fluid-borne noise in hydraulic systems is a problem the fluid power industry has long struggled to address. Traditional noise control devices such as Helmholtz resonators, tuning coils, and Herschel-Quincke tubes are generally too large for fluid power systems unless the speed of sound in the device can be reduced. A compliant lining can achieve this effect, but compliance (and lossy compliance) has had little attention in noise control in general, and in fluid power in particular. One means to achieve compliance in these devices, especially at elevated pressures, is through a liner made of syntactic foam, which in this case is a urethane host matrix with embedded hollow, polymer microspheres. The material properties at elevated pressure are unknown by the liner manufacturer, but are known to be pressure- and temperature-dependent. Therefore, the effect of hydrostatic pressures from 2.1-21 MPa and temperatures from 20-45 C on the liner properties, thus the device performance, are studied. For a Helmholtz resonator, a theoretical model is fit to experimentally-measured transmission loss of the device using a least-squares routine, which solves the inverse problem for the complex bulk modulus of the liner. These material properties are used to compare a predictive model of a tuning coil to experimental data, and in a parameter study of a Herschel-Quincke tube. The compliance of the liner is found to lower the effective sound speed by an order of magnitude and decrease the volume of the cavity of a Helmholtz resonator by up to two orders of magnitude. This work is expected to result is more compact noise control devices for fluid power systems.

L'enjeu des travaux contenus dans cette thèse est la caractérisation rhéologique de suspensions concentrées d'objets viscoplastiques suspendus dans une matrice viscoplastique. Trois contraintes physiques se présentent : l'hétérogénéité des objets, la taille des objets au regard de la taille du système étudié, et la proximité entre la consistance de la phase suspendante et celle des objets. Dans notre contexte alimentaire, il faut ajouter des contraintes d'ordre biologique : les faibles durées de vie des produits, les prélèvements in vivo et la conservation des conditions de températures et d'humidité. L'ensemble de ces contraintes aboutit à une rhéométrie spécifique, aux frontières de la mécanique des milieux continues, avec pour objectif de mesurer des propriétés rhéologiques pertinentes. Pour cela, deux instrumentations originales ont été construites pour effectuer la caractérisation rhéologique le plus vite et au plus près de la génération des échantillons.Appliquée aux bols alimentaires issus de fromage, leur rhéologie vient gouverner les écoulements en bouche et induire la création de surface d'échange en bouche, au travers de la mastication. Au travers de ces interfaces, les arômes et les sapides sont libérés et transportés vers les récepteurs sensoriels. Les résultats montrent le lien entre le seuil de contrainte à l'écoulement du bol et la libération d'arômes de différents hydrophobicités, fonction de la fermeté et du taux de matières grasses du fromage. Plus particulièrement, la phase suspendante fluide, semble jouer un rôle majeur dans la création et la persistance des surfaces d'échanges. Appliquée aux jus gastriques issus de pain, l'enjeu est de savoir si des modifications des contenus en glucides pouvaient impacter la rhéologie des jus gastriques, modifier les cinétiques de vidange en sortie d'estomac, et diminuer l'excursion glycémique. Les résultats montrent l'effet d'un ajout d'amylose sur l'augmentation de la viscosité aux temps de digestion longs, lorsque la rhéologie n'est plus gouvernée par l'encombrement des particules de pain gonflées d'eau. Cependant, l'amylose ne semble pas avoir d'impact sur les cinétiques de vidange, ce qui vient conforter que le débit en sortie d'estomac est régulé par la charge énergétique contenue dans le repas. / The challenge of this work is to carry out the rheological characterization of concentrated suspensions of viscoplastic objects suspended in a viscoelastic matrix. Three structural constraints are faced: the heterogeneity of the particles, the size of particle in relation to the size of the whole system studied, and the proximity between the consistency of the suspending phase and that the consistency of the particles. In the food context, some biological constraints are added: small lifetimes of bolus, in vivo sampling and test conditions of temperature and humidity. All these constraints lead to a compromise in rheometry, at the edge of continuum mechanics, aiming to measure useful rheological properties. To carry out the rheological characterizations quickly and efficiently, two original devices has been designed trough this work. Applied to food bolus from cheese, rheology comes to govern flows in the mouth and induce the creation of exchange area in the mouth, during the chewing process. Through these interfaces, the savors and flavors are released and transported to the sensory receptors. The results show the relationship between the yield stress of the bolus and flavors release of different hydrophobic aromas, depending on the firmness and fat content of the cheeses. Specifically, fluid suspending matrix appears to play a major role in the creation and persistence of exchange area.Applied to gastric juice from bread, the issue is whether changes in carbohydrate content could impact the rheology of gastric juice, modify the kinetics of stomach flow rate, and decrease the glycemic excursion. The results show the effect of amylose addition on increasing the viscosity for long digestion time, when the rheology is no longer governed by the jamming effect of the particles made of water-swollen bread. However, amylose does not seem to have any impact on the kinetics of emptying, which reinforces that the stomach flow rate is regulated by the caloric charge contained in the meal.

Orientador : Geraldo de Freitas Maciel. / Banca: Podalyro Amaral de Souza / Banca: Adriano Souza / Resumo: Em função da concentração de material sólido, misturas resultantes de argila e água conferem, via de regra, propriedades não-Newtonianas. Essa concentração passa a variar com a granulometria e a composição do mineral do sedimento, bem como com a qualidade da água. Algumas tentativas foram feitas na busca de melhor explicar a reologia e a dinâmica global dessas misturas nessa última década, notadamente no contexto de lavas torrenciais e misturas hiperconcentradas. Essa dissertação de mestrado tem por objetivo desenvolver uma caracterização reológica de misturas sólido-líquido (água+argila caulinítica e água+areia fina+argila caulinítica) com diferentes concentrações em volume (Cv) para as quais são determinadas, a partir de ensaios de reometria de precisão, as propriedades tais como a viscosidade aparente, tensão crítica de cisalhamento e curva de tensão de cisalhamento x taxa de deformação. Verificou-se, em condições controladas, a influência da temperatura, pH e concentração em volume (Cv) sobre a curva reológica dessas misturas, além de eventuais fenômenos de tixotropia e modo de preparo. O modelo reológico melhor ajustado para as argilas foi o de Herschel-Bulkley ( ( )n τ =τ c + k γ& ), onde: τ é a tensão de cisalhamento; τc é a tensão crítica (ou rigidez inicial); k é um termo de consistência (viscosidade aparente); n é o índice de escoamento; e dy γ& = du é a taxa de deformação. Obtém-se também, a partir de uma análise quantitativa global, leis de evolução dos diversos parâmetros reológicos em função da concentração em volume. Após a etapa de caracterização física e reológica das misturas, realizaram-se, sob condições controladas, ensaios dinâmicos através de escoamentos de lamas em canal inclinado. Nesses ensaios foram medidos os parâmetros lâmina normal, vazões... (Resumo completo, clicar acesso eletrônico abaixo). / According to solid concentration material, clay-water mixture can exhibit non- Newtonian properties. Mixture concentrations vary with the grain size distribution and the composition of the mineral sediment, as well as, with the water quality. In the last decade some attempts were realized to explain the rheology and the global dynamics of those mixtures, mainly in the context of mudflows and high mud concentrated suspensions. The purpose of this work is develop a rheological characterization of solid-liquid mixture (water + clay kaolinitic and water + fine sand + clay kaolinitic) with different solid volume concentrations (Cv) for which are determined, using a R/S Rheometer (Brookfield), rheological properties such as the apparent viscosity, yield stress and shear stress x shear rate curve. It was verified, under controlled conditions, the influence of the temperature, pH and solid volume concentration (Cv) on the rheological curve those mixtures, besides eventual tixotropy phenomenon and way of prepare. The rheological model adjusted for the clays was Herschel-Bulkley ( ( )n τ =τ c + k γ& ), where: τ is shear stress; τc is yield stress; k is a consistency term (apparent viscosity); n is flow index; and dy γ& = du is shear rate. It is also obtained, from a global quantitative analysis, evolution laws of the several rheologics parameters in function of the solid volume concentration. After the stage of rheological and physical characterization of mixing samples, dynamics experiments were made, under controlled conditions, than the flow of mud was observed in the open channel. These analyses were made in order to examine the behavior of parameters such as: normal depth, discharge and velocities. In fact, the experience showed that is possible to deduce wall friction laws from the rheological proprieties based on high mud concentrated suspensions and the dynamic of these muds. / Mestre

New methodologies and concepts for developing compact and energy efficient automotive exhaust systems have been studied. This originates in the growing concern for global warming, to which road transportation is a major contributor. The focus has been on commercial vehicles—most often powered by diesel engines—for which the emission legislation has been dramatically increased over the last decade. The emissions of particulates and nitrogen oxides have been successfully reduced by the introduction of filters and catalytic converters, but the fuel consumption, which basically determines the emissions of carbon dioxides, has not been improved accordingly. The potential reduction of fuel consumption by optimising the exhaust after-treatment system (assuming fixed after-treatment components) of a typical heavy-duty commercial vehicle is ~4%, which would have a significant impact on both the environment and the overall economy of the vehicle. First, methodologies to efficiently model complex flow duct networks such as exhaust systems are investigated. The well-established linear multiport approach is extended to include flow-acoustic interaction effects. This introduces an effective way of quantifying amplification and attenuation of incident sound, and, perhaps more importantly, the possibility of predicting nonlinear phenomena such as self-sustained oscillations—whistling—using linear models. The methodology is demonstrated on T-junctions, which is a configuration well known to be prone to self-sustained oscillations for grazing flow past the side branch orifice. It is shown, and validated experimentally, that the existence and frequency of self-sustained oscillations can be predicted using linear theory. Further, the aeroacoustics of T-junctions are studied. A test rig for the full determination of the scattering matrix defining the linear three-port representing the T-junction is developed, allowing for any combination of grazing-bias flow. It is shown that the constructive flow-acoustic coupling not only varies with the flow configuration but also with the incidence of the acoustic disturbance. Configurations where flow from the side branch joins the grazing flow are still prone to whistling, while flow bleeding off from the main branch effectively cancels any constructive flow-acoustic coupling. Two silencer concepts are evaluated: first the classic Herschel-Quincke tube and second a novel modified flow reversal silencer. The Herschel-Quincke tube is capable of providing effective attenuation with very low pressure loss penalty. The attenuation conditions are derived and their sensitivity to mean flow explained. Two implementations have been modelled using the multiport methodology and then validated experimentally. The first configuration, where the nodal points are composed of T-junctions, proves to be an example where internal reflections in the system can provide sufficient feedback for self-sustained oscillation. Again, this is predicted accurately by the linear theory. The second implementation, with nodal points made from Y-junctions, was designed to allow for equal flow distribution between the two parallel ducts, thus allowing for the demonstration of the passive properties of the system. Experimental results presented for these two configurations correlate well with the derived theory. The second silencer concept studied consists of a flow reversal chamber that is converted to a resonator by acoustically short-circuiting the inlet and outlet ducts. The eigenfrequency of the resonator is easily shifted by varying the geometry of the short circuit, thus making the proposed concept ideal for implementation as a semi-active device. Again the concept is modelled using the multiport approach and validated experimentally. It is shown to provide significant attenuation over a wide frequency range with a very compact design, while adding little or no pressure loss to the system. / QC 20110208

Fourier transform spectroscopy (FTS) is one of the premier ways to collect source information through emitted radiation. It is so named because the principal measurement technique involves the analysis of spectra determined from the Fourier transform of a time-domain interference pattern. Given options in the field, many space- and ground-based instruments have selected Fourier transform spectrometers for their measurements. The Herschel Space Observatory, launched on May 14, 2009, has three on-board instruments. One, SPIRE, comprises a FTS paired with bolometer detector arrays. SCUBA-2 (Submillimetre Common User Bolometer Array) and FTS-2 have recently been commissioned and will be mounted within the collecting dish of the James Clerk Maxwell Telescope by Fall, 2010. The use of FTS in these two observatories will be examined. While work towards each project is independently useful, the thesis is bound by the commonality between the two, as each seeks similar answers from vastly different viewpoints. / xvii, 123 leaves : ill. (some col.) ; 29 cm

Les radio galaxies sont les candidats préférentiels pour comprendre la formation et l'évolution des galaxies sur une grande échelle de temps. Observées jusqu'à z>5 en raison de leur brillance, elles sont abritées par des galaxies elliptiques géantes. L'émission radio révèle la présence d'un trou noir supermassif. Un tore de poussière entourant le noyau actif de galaxie (AGN) agit comme un coronographe naturel permettant alors l'étude de la galaxie hôte. L'objectif de cette thèse est de déterminer l'évolution de la composante stellaire en présence d'un AGN. La décomposition est faite à partir de la distribution spectrale d'énergie (SED) de l'UV au submillimétrique en utilisant le code d'évolution de galaxies PEGASE.3 et un code d' AGN, les deux modélisant l'émission de la poussière par transfert radiatif.En premier lieu, nous présentons le projet HeRGE, 70 radio galaxies observées avec Herschel, qui permet de mesurer leurs luminosités totales infrarouges, comparables à celles des ULIRG. Une décomposition de la luminosité infrarouge entre l'émission AGN et un modèle de starburst est proposée pour l'ensemble de l'échantillon. Ces luminosités élevées sont interprétées en termes de taux d'accrétion et de formation d'étoiles, favorisant la croissance du trou noir par rapport à la galaxie hôte.En second lieu, l'orientation du jet par rapport au tore est contrainte à partir de l'infrarouge moyen et du rapport des émissions radio des lobes (isotrope, 500MHz) et du coeur (anisotrope, 20GHz). Ces observations en accord avec le modèle d'unification permettent d'évaluer le facteur d'absorption Av, l'inclinaison du tore et de contraindre le facteur de Lorentz.Une sélection de 12 radio galaxies observées de l'UV au sub-mm est analysée avec PEGASE.3 et un modèle d'AGN. Une seule composante stellaire est insuffisante. Seules deux composantes (une évoluée et massive, et une jeune issue d'un starburst) permettent un ajustement significatif de la SED complète. La composante évoluée est très massive (environ 10^12 msun) formée sur une courte période de temps (<10^9 ans). La composante jeune (<4.10^7 ans), moins massive (environ 10^11 msun), confirme un processus épisodique de croissance par sursauts. Ces résultats sont des contraintes fortes pour les modèles de formation de galaxies. La relation avec le noyau actif reste encore à préciser. Les projets d'observations complémentaires, optique et mm, permettront de confirmer ces résultats.

Cette thèse présente une étude expérimentale visant à caractériser les propriétés hydrauliquesd'écoulements à surface libre de fluides complexes: fluides viscoplastiques et mélangesfluides viscoplastiques-grains. Ce travail est motivé par la nécessité d'améliorer lesconnaissances sur les propriétés des écoulements de laves torrentielles, dans un contexte deprotection contre les risques naturels en montagne. Nous avons développé un dispositif delaboratoire original, un canal à fond mobile, permettant de générer des coulées gravitaires quisont globalement stationnaires dans le référentiel du laboratoire. Les coulées présentent unfront très raide, suivi à l'amont d'une zone d'épaisseur uniforme. Les fluides viscoplastiquesutilisés (carbopol et kaolin) ont été choisis de manière à modéliser la matrice des couléesboueuses naturelles en prenant en compte les critères de similitude adaptés. Ces fluidessuivent une loi de comportement de Herschel-Bulkley, et leurs propriétés rhéologiques ont étédéterminées au moyen d'un rhéomètre de laboratoire en accordant un soin particulier àl'établissement des incertitudes associées. Nous présentons des mesures précises de l'évolutionde l'épaisseur des coulées dans la zone uniforme en fonction de la vitesse du fond du canal etde la pente. Ces résultats sont comparés à des prédictions théoriques correspondant à unécoulement permanent et uniforme d'un fluide de Herschel-Bulkley, et nous montrons quel'accord est satisfaisant moyennant les incertitudes sur les mesures rhéologiques. Nous avonségalement analysé la forme du front des coulées et, là-aussi, l'accord avec des modèleshydrauliques (modèles de type Saint-Venant) est bon. Ces deux études permettent de validerles prédictions hydrauliques obtenues à partir de lois de comportement mesurées aurhéomètre. Enfin, dans une dernière partie, nous présentons une étude préliminaire de ladynamique de particules rigides isolées placées au sein de la coulée, en nous intéressant auxvitesses longitudinales et aux vitesses de chute pour différentes tailles et différentes densitésde particules.

La convection de Rayleigh-Bénard est étudiée expérimentalement dans une cellule circulaire. Des fluides à seuil modèles (gels aqueux de Carbopol) sont mis en œuvre. Leurs comportements rhéologiques et leurs perméabilités en relation avec leurs microstructures ont été finement caractérisés. Dans toute la thèse, les expériences ont été menées sans glissement à la paroi. L'influence du seuil d'écoulement et de la distance entre plaques chaudes et froides sur les transferts thermiques a été approfondie. Trois mécanismes sont discutés pour expliquer le déclenchement de la convection: i) les propriétés visco-élastiques au-dessous du seuil, ii) le fluage au-dessous du seuil, iii) une approche d'un milieu poreux pour les gels de Carbopol considérés comme une suspension de micro-gels. On montre que le nombre de seuil Y, représentant le rapport entre la contrainte du seuil et la contrainte de la poussée d'Archimède est un paramètre important gouvernant l'apparition de l'instabilité. Les valeurs critiques de Y^(-1) sont déterminées entre 60 et 90. La visualisation à l'aide des cristaux liquides thermo-chromiques a permis une vue globale de la cinématique. Les structures observées dans les différents états thermiques montrent l'évolution de la convection. Une analyse qualitative du champ de température est également présentée. Enfin, la simulation numérique dans une cellule carré avec un modèle d'Herschel-Bulkley régularisé dans la gamme des nombres sans dimension utilisée dans les expérience a permis de mettre en évidence les paramètres critiques et la morphologie des champs thermiques et cinématique. Les ordres de grandeurs du nombre de seuil critique prédit se comparent raisonnablement avec les valeurs expérimentales. / In this thesis, three main mechanisms proposed in a recent paper (Darbouli et al., Physics of fluids, 25(2) 2013) have been discussed to explain the onset of Rayleigh Bénard Convection in a yield fluid (Carbopol gels): i) the elasto-visco-plasticity behavior of the material below the yield stress, ii) a viscosity at low values of shear rates by creep measurements below the yield stress, iii) a microscopic viewpoint considering the fluid as a porous two phases system. No-slip conditions have been achieved for all the experiments. The results with different Carbopol gels have proved the importance of Y, the yield number which presents the report of the yield stress and the buoyancy effect, as the governing parameter. The critical value of Y^(-1) with no-slip condition has been found between 60 and 90. A visualization measurement with the utilization of thermochromics liquid crystals presents a global view from above. Different structures have been observed in different states of thermal conditions, which describe the evolution of the convection. For several cases the color of the liquid crystals can indicate the temperature field in the whole experiment cell. Numerical simulations with a Herschel-Bulkley model have also been discussed in this thesis. The dimensionless parameters are defined approaching the values obtained in the experiments, so that we can compare the numerical results with some of experimental ones.