Les travaux de cette thèse s'insèrent dans le cadre de l'étude de défauts détectés par contrôle ultrason dans des pièces forgées en alliage d'aluminium destinées à l'industrie aéronautique.Cette étude a été menée sur deux pièces produites par Aubert&Duval en alliage d'aluminium 7050 fourni par Constellium. Ces pièces ont été sélectionnées sur base d'une étude statistique de l'occurrence des défauts sur cinq années de production industrielle. L'une présente un taux de rebut important du fait de ces défauts, l'autre un taux de rebut quasiment nul, bien qu'elles soient produites avec le même matériau.L'étude de l'origine des défauts a été construite autour de deux axes de recherche:• Axe thermomécanique : formation des défauts pendant la mise en forme par endommagement ductile à chaud.• Axe matériau : caractérisation des défauts préexistants dans le matériau après son élaboration et non refermés pendant la mise en forme.L'axe thermomécanique est basé sur la simulation numérique des gammes de mise en forme des pièces industrielles à l'aide du logiciel Forge® 2009. Le deuxième axe est articulé autour de la caractérisation de l'état de porosité dans un matériau modèle présentant une teneur en hydrogène élevée induisant un taux de porosité plus important que le matériau sain utilisé industriellement. De petits lopins de ce matériau modèle ont été soumis à une campagne de compression alternée simulant le forgeage industriel. La campagne de compression a été conçue par simulation numérique afin de reproduire les conditions thermomécaniques industrielles. Deux techniques de caractérisation complémentaires ont été mises en œuvre pour suivre l'évolution de la porosité : détection indirecte par contrôle ultrason à haute résolution spatiale et observations directes par microscopie électronique à balayage. Une méthode d'analyse quantitative des signaux ultrasonores a été mise au point pour ce travail.La comparaison par simulation numérique du procédé de matriçage des deux pièces industrielles étudiées a permis d'écarter l'hypothèse de la création des défauts par endommagement ductile à chaud. Lors du matriçage, la matière est soumise à des chargements globalement compressifs, et le risque d'endommagement est par conséquent très faible. De plus, comme observé lors de la caractérisation des défauts industriels, les défauts présentent un fond lisse, similaire au fond lisse observé sur les rares porosités présentes dans le matériau industriel à l'état brut d'élaboration. Le même type de fond lisse caractérise également les porosités présentes dans le matériau modèle. Cette observation favorise l'hypothèse de défauts provenant d'un défaut préexistant dans le matériau initial (en l'occurrence à partir d'une porosité). Cette hypothèse est par ailleurs cohérente avec le comportement des porosités lors de la simulation du forgeage sur petits lopins du matériau modèle.Les résultats de cette étude montrent en effet que les porosités ont tendance à s'aplatir et à se refermer pendant la déformation plastique à chaud, sans forcément cicatriser. Les faciès de rupture d'échantillons soumis à de fortes déformations cumulées et ne donnant plus de réponse ultrasonore significative continuent en effet de présenter de petites plages à fond lisse qui témoignent d'une cicatrisation incomplète. Le traitement thermique semble en outre favoriser la réouverture des porosités non cicatrisées (zones lisses) pendant la déformation et leur élargissement (zones ductiles). / The thesis project aims to study the origin of defects detected in 7XXX aluminium alloy components for aeronautical applications.This work was conducted on two different components (A and B) produced by Aubert&Duval using 7050 aluminium alloy ingots produced by Constellium. The choice of these two components is based on a five year industrial production statistical analysis. Component A has a much higher defect rate compared to component B, even though they are produced from the same ingots.This work was conducted according to two research axis:• Thermo-mechanical axis : defects occurred during the material forming as a result of ductile hot damage• Material axis: defects resulted from a pre-existent defect in the raw material that was not closed during forming.The thermo-mechanical axis is based on numerical simulations of the component's forming processes using the Forge® 2009 software. The material axis is based on the experimental characterization of porosity in a model material with high hydrogen content and thus a higher pore density than the industrial material. Small samples of the model material were experimentally forged using an alternating compression test that was designed and developed from the industrial forging process. Theses samples were characterized to describe pore evolution during hot deformation, by the combination of two characterization techniques: High Resolution Ultrasound Control (HRUS) and Scanning Electron Microscope (SEM) observations on 2D sections and fracture surfaces. An ultrasound signal processing technique was also developed in order to quantitatively compare two different deformation states using ultrasound control.The results of the numerical simulation of A and B component stamping process showed that the detected defects could not occur by ductile hot damage. Indeed, the stamping process is purely compressive and damage risks are thus very low. Moreover, the defect experimental characterization showed the presence of smooth surface areas very similar to the surface aspect of the rare observed porosity in the raw material. The same smooth aspect characterized the porosity observed in the model material. These results lead to the hypothesis that the detected defects could occur from a pre-existent defect (porosity more precisely) in the raw material. This hypothesis is consistent with the pore behaviour during the experimental hot forging of the model material.The hot forging experiment conducted on the model material, showed that pores tend to flatten as plastic strain increases. SEM observations on rupture surfaces from the highest plastic strain showed the presence of small smooth aspect areas. These areas confirm that some pores are not completely healed during deformation even though the correspondent ultrasound signal is very low. During heat treatment, unhealed pores after deformation seem to be reopened (smooth aspect surface areas) and expanded (ductile areas).

Une des données essentielles que recherche le clinicien ORL lors de son diagnostic concerne la fonctionnalité des fosses nasales. Dans certains cas, cette fonction n’est approchée que de manière partielle par les diverses méthodes de mesures mises à sa disposition. La prévalence des obstructions nasales chroniques dans la population générale1 (30% de la population) souligne l’intérêt que les orthodontistes doivent porter aux pathologies respiratoires. L’acquisition d’un outil de diagnostic fiable de l’obstruction nasale permettra d’effectuer un traitement plus précoce afin de guider au mieux la croissance crânio-faciale. La RAA ne peut pas être actuellement utilisée comme unique examen diagnostique, car sa corrélation avec les évaluations subjectives peut rester faible. L’objectif de ce travail est de fournir un modèle d’aide au diagnostic de l’obstruction nasale pour tous les patients. Dans les cas où la RAA ne permet pas un diagnostic certain de l’obstruction nasale, la simulation numérique permettrait, après avoir identifié les situations anatomiques particulières, d’obtenir une approche fonctionnelle objective en confirmant ou non le diagnostic de la RAA. La méthode utilisée pour ce travail de thèse consiste en une résolution complète des équations de Navier-Stokes dans des géométries reconstruites en 3D, c’est-à-dire ayant bénéficié d’un traitement numérique pour extraire les contours puis créer les maillages surfaciques et volumiques. Le développement d'outils de mesure objectifs est un enjeu capital pour déterminer les stratégies thérapeutiques optimales et pour évaluer les résultats des traitements. La modélisation par ordinateur de la dynamique des flux aériens au sein des fosses nasales à partir de reconstructions tomodensitométriques tridimensionnelles peut présenter des applications cliniques. Une de ses applications permettrait à l’ORL de visualiser l’écoulement post-opératoire virtuellement. / One of the essential data sought by the clinician in his diagnosis on the functionality of the nasal cavity. In some cases, this function is only partially approached by various measurement methods at its disposal. The prevalence of chronic nasal obstruction in the population (30% of the population) underlines the interest to orthodontists must wear respiratory disorders. The acquisition of a reliable diagnostic tool of nasal obstruction will allow for earlier treatment to guide the best craniofacial growth. The RAA can not be currently used as single diagnostic test, because its correlation with subjective assessments can remain low. The objective of this work is to provide a model for the diagnosis of nasal obstruction for all patients. In cases where the RAA does not allow a definite diagnosis of nasal obstruction, numerical simulations allow, after identifying the particular anatomical situations, to obtain a functional objective or not confirming the diagnosis of rheumatic fever. The methodology for this thesis consists of a complete resolution of the Navier-Stokes equations in geometries reconstructed in 3D, that is to say who received a digital processing to extract the contours and surface meshes and create volume. The development of objective measurement tools is a key issue in determining optimal treatment strategies and to evaluate treatment outcomes. Computer modeling of the dynamics of air flow within the nasal cavity from three-dimensional CT reconstructions may have clinical applications. One of its applications allow to visualize the flow postoperative virtually.

Non-metallic inclusions in molten steel have received worldwide attention due to their serious influence on both the steel product quality and the steel production process. These inclusions may come from the de-oxidation process, the re-oxidation by air and/or slag due to an entrainment during steel transfer, and so on. The presence of some inclusion types can cause a termination of a casting process by clogging a nozzle. Thus, a good knowledge of the inclusion behavior and deposition rate in steel flows is really important to understand phenomena such as nozzle clogging. In this thesis, inclusion behaviors and deposition rates in steel flows were investigated by using mathematical simulations and validation by experiments. A ladle teeming process was simulated and Ce2O3 inclusion behavior during a teeming stage was studied. A Lagrangian method was used to track the inclusions in a steel flow and to compare the behaviors of inclusions of different sizes. In addition, a statistical analysis was conducted by the use of a stochastic turbulence model to investigate the behaviors of different-sized inclusions in different nozzle regions. The results show that inclusions with a diameter smaller than 20 μm were found to have similar trajectories and velocity distributions in the nozzle. The inertia force and buoyancy force were found to play an important role for the behavior of large-size inclusions or clusters. The statistical analysis results indicate that the region close to the connection region of the straight pipe and the expanding part of the nozzle seems to be very sensitive for an inclusion deposition. In order to know the deposition rate of non-metallic inclusions, an improved Eulerian particle deposition model was developed and subsequently used to predict the deposition rate of inclusions. It accounts for the differences in properties between air and liquid metals and considers Brownian and turbulent diffusion, turbophoresis and thermophoresis as transport mechanisms. A CFD model was firstly built up to obtain the friction velocity caused by a fluid flow. Then, the friction velocity was put into the deposition model to calculate the deposition rate. For  the  case  of  inclusion/particle  deposition  in  vertical  steel  flows,  effects  on  the deposition rate of parameters such as steel flow rate, particle diameter, particle density, wall roughness and temperature gradient near a wall were investigated. The results show that the steel flow rate/friction velocity has a very important influence on the rate of the deposition of large particles, for which turbophoresis is the main deposition mechanism. For small particles, both the wall roughness and thermophoresis have a significant influence on the particle deposition rate. The extended Eulerian model was thereafter used to predict the inclusion deposition rate in a submerged entry nozzle (SEN). Deposition rates of different-size inclusions in the SEN were obtained. The result shows that the steel flow is non-uniform in the SEN of the tundish. This leads to an uneven distribution of the inclusion deposition rates at different locations of the inner wall of the SEN. A large deposition rate was found to occur at the regions near the SEN inlet, the SEN bottom and the upper region of two SEN ports. For the case of an inclusion/particle deposition in horizontal straight channel flows, the deposition rates of particles at different locations of a horizontal straight pipe cross- section were found to be different due to the influence of gravity and buoyancy. For small particles with a small particle relaxation time, the gravity separation is important for their deposition  behaviors  at  high  and  low  parts  of  the  horizontal  pipe  compared  to  the turbophoresis. For large particles with a large particle relaxation time, turbophoresis is the dominating deposition mechanism. / <p>QC 20150326</p>

The Sun ionizes a small fraction of Earth's atmosphere above roughly 60 km, producing the plasma that constitutes the ionosphere. Radio signals passing through the ionosphere scatter off of plasma density structures created by the Farley-Buneman instability (FBI). While numerous studies have characterized the FBI's intrinsic nature, its evolution within the broader context of the surrounding plasma remains enigmatic. This dissertation answers two fundamental questions about the FBI: How does it interact with density gradients? How does its non-linear evolution depend on the background plasma? The fourth chapter examines the combined development of the FBI and the gradient drift instability (GDI) using a 2-D simulation of the equatorial ionosphere. A half-kilometer wave perturbs a plasma layer perpendicular to the ambient magnetic field, causing the perturbed layer to develop GDI waves along the gradient aligned with the ambient electric field, as well as FBI waves in a region where the total electric field exceeds a certain threshold. Early radar observations suggested that these two instabilities were distinct phenomena; the reported results illustrate their coupled nature. The fifth chapter presents 2-D simulations in which a one-kilometer plasma wave develops an electric field large enough to trigger meter-scale waves. Such large-scale waves arise via the GDI within the daytime ionospheric gradient around 100-110 km. Typical ionospheric radars only observe meter-scale irregularities but observations show meter-scale waves tracing out larger structures. Simulated meter-scale FBI in the troughs and crests of kilometer-scale GDI matches radar observations of the daytime equatorial ionosphere, answers a question about electric-field saturation raised by rocket observations in the 1980s, and predicts an anomalous cross-field conductivity important to magnetosphere-ionosphere (M-I) coupling. The sixth chapter of this dissertation presents 3-D simulations of the FBI at a range of altitudes and driving electric fields appropriate to the auroral ionosphere, where it plays a role in M-I coupling. Research has thoroughly established the linear theory of FBI but rigorous analysis of radar measurements requires an understanding of the turbulent stage. These simulations explain the change in instability flow direction with altitude, with regard to the direction of background plasma flow.

Marangoni flows are free-surface flows driven by gradients of surface tension. Because surface tension depends on chemical composition, Marangoni flows may be generated by the uneven distribution of surface-active species at an interface. The primary goal of this thesis is to develop a rigorous computational framework for the simulation of the fluid dynamical and interfacial phenomena underlying the physics of Marangoni flows. The focus is on characterizing the different dynamical regimes generated by the presence of surface-active species (surfactants) at an interface. The computational framework was developed using direct numerical simulation, that is, by simultaneously solving the full system of partial differential equations governing the free-surface flow and the surfactant transport on a continually deforming interface. Results from the simulations enabled detailed examination of the interfacial mechanisms of surfactant transport and provided a comprehensive picture of the free-surface flow. Analysis of the results established limits of applicability of scaling solutions previously proposed in the literature, calculated the necessary corrections, and also lead to the discovery of previously unobserved scaling laws in viscous Marangoni flows. New findings from this research not only enhance the fundamental understanding of the physics of Marangoni flows, but also the ability to accurately predict the behaviour of Marangoni flows and the associated transport of surface-active species, which is critical to the understanding of important natural and biomedical processes, ranging from the surfactant-driven propulsion of insects and microorganisms to the spreading of drugs and natural surfactants (proteins) in the eye and lungs. Controlled Marangoni transport of chemical species is also relevant to a wide range of environmental and technological processes, with applications ranging from cleaning of oil spills to coating of microfluidic devices.

[pt] Em projetos de barragens de rejeitos de mineração, normalmente existe uma grande preocupação relacionada aos aspectos construtivos e de segurança, em função do processo de lançamento de material sobre depósitos fofos de rejeitos saturados. Os maiores desafios consistem na avaliação da resistência e da deformabilidade do material de fundação, na possibilidade de ocorrência de liquefação, e no desenvolvimento e velocidade de dissipação das poropressões. Neste contexto, a Rio Paracatu Mineração (RPM) construiu um aterro experimental instrumentado em grande escala,com material de empréstimo siltoarenoso sobre o depósito de rejeito na Mina do Ouro. Este aterro experimental foi projetado e construído com 10,0 m de altura, talude frontal de 1,5H:1,0V e um talude suave com uma berma de estabilização de 4,0 m de altura no talude oposto. A fundação do aterro foi instrumentada com piezômetros elétricos de corda vibrante, inclinômetros e placas de recalque. Além da instrumentação de campo, há uma base de dados de ensaios de laboratório realizados nas universidades de Viçosa (UFV), São Carlos (USP/SC) e PUC-Rio. Todos estes dados permitem a caracterização do material do rejeito, além da determinação de parâmetros de resistência, deformabilidade, permeabilidade e adensamento. O objetivo desta Tese é contribuir para uma melhor compreensão do comportamento de barragens de rejeitos de mineração com base nas evidências laboratoriais e na retroanálise do comportamento do aterro experimental de larga escala. Para tanto, foi realizada a simulação numérica do processo construtivo do aterro experimental, buscando-se o modelo constitutivo e os parâmetros mais adequados para a representação do comportamento de solos granulares saturados. As análises numéricas foram executadas com o programa computacional Plaxis V11, de elementos finitos. Uma vez reproduzido o comportamento do aterro experimental, procedeu-se à previsão dos recalques e deslocamentos horizontais da barragem de ouro da Rio Paracatu Mineração. Os resultados mostraram que o modelo Hardening Soil permitiu a reprodução do comportamento do aterro experimental durante as etapas construtivas. A boa seleção de parâmetros, com base em um extenso programa experimental, permitiu a previsão do comportamento da barragem principal. / [en] In projects of mining tailings dams, there is usually a big concern and constructive aspects related to security, according to the release process of material on saturated tailings deposits cute. The biggest challenges consist in evaluating the strength and deformability of the foundation material, the possibility of liquefaction and speed the development and dissipation of poropressures. In this context, the Rio Paracatu Mineração (RPM) built an instrumented experimental landfill on a large scale, with sandy silty material loan on the deposit of waste in the Gold Mine. This experimental landfill was designed and constructed with 10.0 m height, front slope of 1.5 H: 1.0 V and a gentle slope with a stabilizing berm of 4.0 m high on the slope opposite. The foundation of the embankment was instrumented with vibrating string electric piezometers, inclinometers and settlement plates. Besides the instrumentation field, there is a database of laboratory tests conducted at the Universities of Viçosa (UFV), San Carlos (USP / SC) and PUC-Rio. All these data allow the characterization of tailings material, besides the determination of parameters of resistance, deformability, permeability and density. The goal of this dissertation is to contribute to a better understanding of the behavior of mining tailings dams based on laboratory evidence and backcalculation behavior of large-scale experimental landfill. Therefore, we performed a numerical simulation of the construction process of the experimental landfill, seeking the constitutive model and parameters best suited to represent the behavior of granular soils saturated. The numerical analyzes were performed with the computer program Plaxis V11, finite elements. Once reproduced the experimental behavior of the landfill, proceeded to the prediction of settlements and horizontal displacements of the dam the Rio Paracatu gold mining. The results showed that the model Hardening Soil allowed playback behavior of the experimental landfill during the construction stages. A good selection of parameters based on an extensive experimental program allowed the prediction of the behavior of the main dam.

On s’intéresse dans ce travail à la convection thermique en présence de champ magnétique, en lien avec les essais Vulcano du CEA de Cadarache. Ces essais ont pour but de reproduire en laboratoire les écoulements générés par un chauffage volumique dues aux matières radioactives issues d’accidents nucléaires. Le chauffage par induction électromagnétique de ces fluides permet de simuler en laboratoire la puissance volumique des désintégrations nucléaires en évitant l’utilisation de matériaux radioactifs. Néanmoins, ce forçage électromagnétique génère un couplage entre l’écoulement fluide et le champ magnétique, par la force de Lorentz d’une part, et d’autre part le chauffage volumique par effet Joule se concentre dans l’épaisseur de peau. De plus, il peut y avoir dans ces essais présence d’une phase dispersée métallique qui risque de perturber le chauffage inductif de la phase continue (oxyde). Il est nécessaire d’étudier comment ces effets d’origine électromagnétique modifient l’écoulement. À plus petite échelle, les phénomènes magnétiques en jeu peuvent être reproduits grâce à des métaux liquides à température ambiante. Lorsqu’un champ magnétique harmonique est appliqué à la frontière d’un métal liquide, l’effet stabilisant de la force de Lorentz sera prépondérant devant l’effet Joule à basses fréquences, alors que l’effet Joule devient significatif à hautes fréquences. On considère alors plusieurs situations canoniques permettant d’analyser l’effet d’un champ magnétique constant DC et alternatif AC sur un écoulement de convection naturelle monophasique, ou en présence d’une phase dispersée très conductrice. Une première partie est consacrée à l’étude expérimentale d’une méthode de vélocimétrie acoustique sur une cellule de convection naturelle en eau. Par comparaison avec des mesures de vitesse par imagerie de particules (PIV) et des simulations numériques directes (DNS), cela a permis de valider la méthode acoustique en vue de futures expériences en métal liquide. Une seconde étude est dédiée à l’analyse numérique de l’instabilité de Rayleigh-Bénard en présence d’un champ magnétique DC vertical. Les taux de croissance sont déterminés par analyse de stabilité linéaire pour des nombres de Hartmann 0 _ Ha _ 100 et des nombres de Rayleigh 103 _ Ra _ 1.5 × 105. Ces prédictions sont confirmées par DNS. En régime stationnaire, l’analyse des DNS a permis de mettre en évidence un effet marqué de la force de Lorentz sur les structures, à travers leur nombre d’onde et leur orientation. La troisième configuration étudiée est le chauffage inductif d’une couche de métal liquide en imposant un champ magnétique harmonique au niveau de la paroi basse pour des nombres de Rayleigh proportionnels à la puissance déposée 1.1×104 _ Ra _ 1.2×105 et pour des épaisseurs de peau inférieures à la moitié de l’épaisseur de liquide. Dans ce cas, les courants induits sont dissipés par effet Joule sur l’épaisseur de peau. La prédiction des taux de croissance requiert l’utilisation de méthodes adaptées car ici le développement de l’instabilité est concomitant à la conduction instationnaire de la chaleur. Malgré la perte de symétrie des équations introduite par le terme source d’effet Joule, l’écoulement présente une symétrie de réflexion apparente, que ce soit en régimes transitoire ou stationnaire. Cela est lié au brassage conséquent par convection naturelle. Enfin une situation modèle de particules métalliques immergées dans un liquide transparent au champ magnétique est étudiée. Ici, l’énergie magnétique est dissipée sous forme thermique dans les particules, qui transmettent toute leur chaleur au fluide. Des mouvements convectifs se mettent alors en place. La convection est décrite par la concentration relative en particules. Selon les valeurs des paramètres, on observe la formation d’amas de particules en réponse au panache qu’elles génèrent. / In this work, we study thermal convection in the presence of magnetic field in connection with the Vulcano tests at CEA Cadarache. These tests aim at reproducing in the laboratory the behavior of fluids that result from a severe nuclear accident, while avoiding the use of radioactive materials. Induction heating is used to mimic in the laboratory the volume power of nuclear disintegrations. However, there is a parasitic coupling between the flow and the magnetic field (Lorentz force) on the one hand, and on the other hand the concentration of Joule dissipation in the skin layer, while it is homogeneous in a real case. Moreover, the presence of a dispersed metallic phase may interfere with the induction heating of the continuous phase (oxide). At smaller scales, the magnetic phenomena at play can be simulated with liquid metals at room temperature. When a magnetic field is applied at the boundary of a liquid metal, the stabilizing role of the Lorentz force will dominate at very low frequencies, whereas the Joule effect will be significant at high frequencies. Here, we consider several generic configurations that allow to analyze the action of DC or AC magnetic fields on natural convection, for a single phase flow or in the presence of a dispersed phase. The first part is devoted to the experimental study of an acoustic velocimetry method in the case of natural convection in water. The comparison of particle imaging velocimetry (PIV) data along with direct numerical simulations (DNS) allowed the validation of the acoustic method. In the future, experiment in liquid metal will be performed. A second part is dedicated to the numerical analysis of the Rayleigh-Bénard instability in the presence of vertical DC magnetic field. Growth-rates are determined by linear stability analysis for Hartmann numbers 0 _ Ha _ 100 et Rayleigh numbers 103 _ Ra _ 1.5×105. These predictions are confirmed by DNS. In the steady regime, DNS showed a strong effect of the Lorentz force on flow structures, through the modulation of the wavenumber and the structures orientation. The third configuration is the induction heating of a liquid metal layer by impressing an AC magnetic field at the bottom, for power Rayleigh numbers 1.1×104 _ Ra _ 1.2×105 and for skin depths lower than half of the liquid thickness. The prediction of the growth-rates requires to use an adapted method to account for the simultaneous development of the transient heat conduction. Although the Joule dissipation source term breaks the mirror symmetry of the governing equations, the flow exhibits an apparent reflectional symmetry, both in the transient and in the stationary regimes. This is a consequence of the mixing induced by the natural convection. Finally, a model situation of metallic particles immersed in a fluid transparent to the magnetic field is studied. Here the magnetic energy is dissipated in thermal form in the particles, which then transfer their heat to the fluid. As a result, buoyant motion then sets in. Depending on the parameters values, we observe the formation of clusters in response to the thermal plume that they generate.

Le présent travail de thèse s’inscrit dans le contexte fondamental de la propulsion marine instationnaire. Il analyse les performances propulsives des profils de différents rapports d’aspects (envergure par rapport à la corde), en oscillation harmonique de tangage. On modélise la principale partie active d’un système propulsif naturel, tel une nageoire caudale de poisson, par des profils simples, minces et rigides quasiment identiques, à l’exception du rapport d’aspect très important du premier profil (4.8), dit bidimensionnel, et très faible du second (1/6), dit tridimensionnel. Le mouvement de tangage est imposé au tiers de corde, en partant du bord d’attaque des profils immergés dans un écoulement. L’objectif étant de faire varier le nombre de Strouhal, basé sur l’amplitude d’excursion totale du bord de fuite du profil, dans une gamme relativement faible et étroite, où évoluent plusieurs espèces aquatiques. Cette gamme est donc intéressante pour l’étude des systèmes propulsifs artificiels. L’aptitude à la propulsion a été déduite en analysant la structure de l’écoulement induit dans le sillage, et en déterminant l’effort réel exercé par le profil sur l’écoulement. En raison des effets importants de la traînée induite associée aux tourbillons marginaux et de ceux de la traînée visqueuse, inversement proportionnelle au rapport d’aspect, il apparait que les performances propulsives du profil tridimensionnel sont bien moindres que celles du profil bidimensionnel. L’influence des configurations du battement sur les performances propulsives a également été étudiée. / This thesis work is part of the fundamental context of unsteady marine propulsion. The propulsive performances of different aspect ratios (span to corde) profiles, in harmonic pitching motion are analyzed. The main active part of a natural propulsion system, such as a fish caudal fin, is modelled by simple, thin and rigid profiles that are almost identical, with the exception of the very high aspect ratio of the first profile (4.8), called two-dimensional, and very low aspect ratio of the second (1/6), called three-dimensional. The pitching motion is imposed on the third corde, starting from the leading edge of the profiles immersed in a flow. The objective is to vary the Strouhal number, based on the total excursion amplitude of the trailing edge of the profile, over a relatively small and narrow range, where several aquatic species evolve. This range is therefore interesting for the study of artificial propulsion systems. The propulsion capability was inferred by analyzing the structure of the induced flow in the wake, and by determining the actual force exerted by the profile on the flow. Due to the significant effects of the induced drag associated with marginal vortices and those of the viscous drag, inversely proportional to the aspect ratio, it appears that the propulsive performance of the three-dimensional profile is much lower than that of the two-dimensional profile. The influence of flapping parameters on propulsive performance has also been studied.

[pt] Um método para a simulação dinâmica bi-dimensional de bóias cilíndricas de superfície, sujeitas à ação de ondas e ancorada ao fundo marinho através de uma linha flexível é apresentado neste trabalho. O modelo de elementos finitos adotado por Lustosa (2002) é empregado na descrição do movimento do cabo de ancoragem, onde elementos de vigas de Euler-Bernoulli de dois nós são empregados, com a cinemática do movimento descrita através de grandezas referidas ao sistema local co-rotacionado. No modelo numérico, são consideradas as não-linearidades geométricas resultantes dos grandes deslocamentos da linha e que causam o acoplamento das rigidezes axial e flexional. Na caracterização do movimento da linha consideram-se os seguintes efeitos: peso próprio, empuxo, carregamento hidrodinâmico da correnteza marinha, deslocamentos impostos pela unidade flutuante, forças de inércia e, se presente, a ação de flutuadores. Os deslocamentos do cabo são obtidos da solução do sistema algébrico de equações não-lineares, resultante da integração temporal de Newmark das equações diferenciais temporais do movimento, cuja convergência é garantida através do método iterativo de Newton-Raphson. Para a bóia, considerada para efeito de análise como indeformável, três graus-de-liberdade são considerados: dois de deslocamentos lineares e um de deslocamento angular. As ondas de superfície, monocromáticas e bi-dimensionais, são representadas no modelo através da teoria linear de Airy associada à modificação empírica de Wheeler (Wheeler, 1969). Forças resultantes da ação das ondas sobre a bóia de superfície são obtidas através da integração numérica da equação proposta por Chitrapu et al. (1998) que compõe-se de um termo viscoso - resultante da parcela de arrasto da equação de Morison - e de um termo não-viscoso - considerado como resultante da soma da força de Froude-Krylov (Chakrabarti, 1987) à parcela de inércia da equação de Morison -. A integração numérica desta equação sobre a superfície “molhada” instantânea do cilindro é realizada, a cada incremento de tempo da análise, empregando-se o método da quadratura adaptativa de Simpson. Finalmente, na integração temporal das equações de movimento da bóia de superfície utiliza-se o método de Runge-Kutta de quinta ordem, com o controle adaptável do passo temporal. O procedimento descrito foi implementado em um programa de computador e os resultados obtidos de algumas análises numéricas comparados com os fornecidos por outras simulações independentes, de forma a verificar a adequabilidade da formulação proposta no estudo da dinâmica de bóias flutuantes. / [en] In this work a method for two-dimensional dynamic analysis of a cylindrical buoy under surfaced waves and moored to the seafloor by a flexible riser is presented. The finite element model used to describe the dynamics of the maritime line was developed by Lustosa (2002). This is a two node beam element based on the Euler-Bernoulli theory, with the element kinematics referred to a co-rotacional coordinate system attached to the element. The model considers the representation of geometric non-linearities due to the line motion large displacements resulting in axial-bending couplings in the element total stiffness representation. The element displacements in the line are due to the following external loadings: self weight, buoyancy force, hydrodynamic forces due to maritime currents, surface buoy imposed displacements, the forces of inertia, and, if present, the action of floaters. The buoy two displacements and one rotation degrees-of-freedom are measured at its center of gravity and are obtained from the solution of the non-linear algebraic system resulting from the Newmark temporal integration of the differential equations of motion, which solution convergence is granted by a Newton- Rapshon iterative method. The linear wave theory - Airy Theory -, associated to the Wheeler empirical modification, is used in this work to represent the monochromatic bi-dimensional surface waves. To obtain the wave forces acting on the surface buoy while in motion, the numerical integration of the equation presented by Chitrapu et al. (1998) is employed. This is the result of a two part equation: the first is evaluated as the Morison viscous drag force and, in a second part, is the sum of the - Froude-Krylov - and the Morison inertia forces. The numerical integration of the resulting equation is evaluated over the instantaneous wet surface of the buoy, in each time-step, using the adaptive recursive Simpson´s method. Finally the temporal integration of the buoy equations of motion are evaluated using a fifth-order Runge-Kutta method, with an adaptive control of the temporal time step. This procedure has been implemented in a computer program and the results obtained from some numerical analysis are compared to the ones obtained from independent simulations, so as to verify the ability of this proposed method to describe the dynamic of floating buoys.

Orientador: José Luiz Gasche / Resumo: Na atualidade, um compressor precisa ter elevada eficiência para ser competitivo no mercado. Como o escoamento de refrigerante nas válvulas é uma das maiores fontes de perdas termodinâmicas, melhorias no seu projeto podem aumentar significativamente a eficiência do compressor. Simular computacionalmente válvulas de compressores utilizando modelos tridimensionais completos, além de computacionalmente caro, demanda muito tempo e, por essa razão, pesquisadores buscam metodologias simplificadas para realizar a tarefa de predizer a dinâmica das válvulas. Aqui avaliamos a aplicabilidade de um modelo híbrido para simular computacionalmente o comportamento dinâmico de um modelo frequentemente usado como válvula de sucção. Neste modelo, resolvemos o problema do escoamento tridimensional em regime transiente pela válvula aplicando o método de volumes finitos para escoamento compressível e isotérmico e a dinâmica da válvula usando um modelo massa-mola-amortecedor com um grau de liberdade. Comparamos os resultados do modelo híbrido com aqueles obtidos de um modelo tridimensional completo para a estrutura da válvula usando como parâmetros o tempo de processamento, deslocamento da válvula, força sobre a válvula e distribuições de pressão e de velocidade do escoamento. Validamos os modelos numéricos com resultados experimentais. A boa concordância entre os resultados indicam que a redução de tempo obtida pelo modelo híbrido é pequena (15% em média) e não justifica seu uso. / Mestre