La dynamique de filaments flexibles individuels en écoulement visqueux est une étape essentielle pour comprendre et contrôler la rhéologie de nombreux fluides complexes. Cette dynamique sous-tend également une multitude de processus biophysiques allant de la propulsion des micro-organismes aux écoulements intracellulaires. Cette thèse présente des expériences systématiques permettant d’étudier la dynamique de filaments flexibles browniens dans un écoulement visqueux. Nous avons choisi d’utiliser un biopolymère, l’actine, comme système modèle de filaments. Sa longueur typique varie de 1 à 100 μm, il est flexible à ces échelles avec une longueur de persistance de l’ordre de 20μm, à cause de ses petites dimensions, il est soumis aux forces Browniennes avec des fluctuations en flexion, et enfin il peut être marqué en fluorescence. Nous utilisons des dispositifs microfluidiques associés à des systèmes de contrôle d’écoulements, un microscope optique équipé́ avec une platine motorisée pour réaliser des expériences contrôlées permettant de suivre la dynamique des filaments d’actine dans un écoulement de cisaillement pur et dans un écoulement élongationnel. Pour les expériences en cisaillement pur, des simulations reproduisant les conditions expérimentales ont aussi été́ menées en utilisant la théorie des poutres inextensibles d’Euler-Bernoulli et la théorie non locale des corps élancés en présence de fluctuations Browniennes et sont en accord quantitatif avec les résultats expérimentaux. Nous montrons que la dynamique des filaments dans ce système est principalement régie par le nombre élasto-visqueux, nombre sans dimension comparant les forces de trainée visqueuses aux forces de flexion élastiques, les fluctuations thermiques ne jouant qu’un rôle secondaire. Nous présentons une caractérisation complète des différents modes de déformation subies par le filament pendant une rotation ainsi que des transitions entre les différents modes. Dans la géométrie élongationnelle, nous avons choisi un canal hyperbolique optimisé pour permettre de longs temps de résidence sous taux de déformation constant. Nous avons observé́ directement la suppression des fluctuations transverse dans la partie extensionnelle tandis que nous observons, dans la partie compressive la formation de structures hélocoïdales tridimensionnelles après le flambage du filament. Pour finir, ce manuscrit de thèse décrit des développements expérimentaux permettant de fabriquer des suspensions de filaments d’actine relativement monodisperse en taille ainsi que des résultats préliminaires sur des effets rhéofluidifiants. Au bilan, les résultats présentes dans ce manuscrit pose les premières pierres de travaux futurs en direction de l’étude de la dynamique de ces filaments dans des écoulements plus complexes comme des écoulements de Poiseuille ou oscillants. Ils permettent aussi d’envisager des études sur le lien entre déformations de particules et propriétés des suspensions diluées d’objets flexibles et Brownien, lien encore peu étudié du point de vue expérimental. / The dynamics of individual flexible filament in a viscous flow is the key to deciphering the rheolog- ical behavior of many complex fluids and soft materials. It also underlies a wealth of biophysical processes from flagellar propulsion to intracellular streaming. This thesis presents systematic exper- iments to investigate the dynamics of flexible and Brownian filaments in viscous flows. Biopolymer actin has been chosen to be our experimental model filament: its typical length can be varied from 1 to 100 μm, it is flexible at these dimensions with a persistence length in the order of 20μm, it is Brow- nian due to its small diameter with bending fluctuations and it can be labelled by fluorescent dye. Microfluidic channels and flow control systems are combined to optical microscope with automated stage to carry out well-controlled experiments on the diverse dynamics of actin filaments in shear flow and pure straining flow. In shear flow, simulations matching the experimental conditions have also been performed using inextensible Euler-Bernoulli beam theory and non-local slender body hy- drodynamics in the presence of thermal fluctuations and agree quantitatively with the experimental results. We demonstrate that filament dynamics in this flow geometry is primarily governed by a dimension- less elasto-viscous number comparing viscous forces to elastic forces with thermal fluctuations only playing a secondary role. We present a complete characterization of the different modes of defor- mation undergone by the filament while rotating as well as of the transitions between these different modes. In pure straining flow, we opt to use an optimized hyperbolic channel to allow long resi- dence time at constant strain rate to be applied. We directly observe the suppression of transverse fluctuations in the extensional part of the hyperbolic channel while we observe, in the compressive part of the flow, the formation of three dimensional helical structures subsequent to the initial buck- ling of the filament. Finally, this thesis manuscript also reports on experimental developments to fabricate suspensions of actin filament with a narrow distribution of lengths and on preliminary re- sults on shear-thinning effects. All together the results presented here pave the way of future studies towards the understanding of filament dynamics in more complex flows, as Poiseuille flows or oscil- latory flows, as well as towards establishing the link between filament deformations and rheological response in dilute suspensions of flexible Brownian filaments, which remains nearly unexplored from an experimental point of view.

Les travaux présentés s’intéressent à la réponse vibratoire et au champ acoustique émis par une structure immergée et excitée par une couche limite turbulente, dans le domaine des bas nombres d’ondes et pour un nombre de Mach faible. Ce travail s’inscrit dans la problématique d’amélioration de laprédiction du bruit rayonné dans ce type de configurations, et peut trouver son application à la discrétion acoustique des navires, ou à la caractérisation du bruit rayonné par des structures externes excitées par un écoulement.Numériquement, une analyse modale de la réponse de la structure en formulation (u,p,φ) est réalisée à l’aide du code élément finis Code_Aster. L’excitation est modélisée par une somme d’ondes planes de pression dont la densité spectrale est obtenue à partir des modèles d’excitation pariétale disponibles dans la littérature. Une analyse harmonique sur base modale est réalisée pour chaque cas de chargement.Cette approche permet la prise en compte du couplage fluide-structure dans le cas d’un fluide lourd et présente l’avantage de s’affranchir des hypothèses généralement faites, de fluide léger et d’orthogonalité des déformées modales.Les résultats issus de la modélisation numérique sont comparés à des données expérimentales, concernant le comportement vibratoire d’un dispositif constitué d’une plaque plane excitée par un écoulement généré en tunnel hydrodynamique. Les résultats numériques et expérimentaux observés sont proches,qu’il s’agisse du comportement global, du niveau spectral moyen en déplacement ou du niveau de pression acoustique mesuré. En complément, l’influence de défauts, constitués de marche montantes et descendantes de hauteur inférieure à l’épaisseur de la couche limite, sur l’excitation et la réponse de la structure est explorée expérimentalement. / The following work consist in the study of the vibroacoustic response of a structure submerged in fluid, under a turbulent boundary layer flow, the response of the structure is driven by the low wavenumber behaviour, for a small Mach number. This work aims at providing better means of predicting the noise radiated in such setups, mainly regarding stealthiness of ships and submarines and noise radiated by outer structures.A numerical modal analysis based on the (u,p,φ) formulation available in the finite element software Code_Aster is performed. The pressure induced by the boudary layer is then described as a sum of plane waves and several harmonical analysis are performed on the reduced problem, projected on the (u,p,φ) modal basis, one for each term of the sum. This allows us to account for the fluid-structure interaction (inertial and acoustic) in confined and infinite fluid domains. Most numerical models found in scientific papers are making the assumption of a light fluid, or a fluid loaded plate, thus not taking clearly into account the fluid-strucure interaction or only the inertialpart. Here the interaction due to the acoustic field radiated by the plate is fully accounted for.The validity of the proposed numerical method is assesed and numerical results are compared to data obtained from an experimental setup used within a hydrodynamic tunnel. Numerically, a good reproduction of the behaviour of the plate is obtained, both in terms of displacement and spectral levels. The acoustic levels are also compared to their numerical counterparts at the position of the transducer. Moreover, an experimantal analysis is performed, for backward and forward steps of height smaller than the thickness of the boundary layer, in order to investigate the influence of such configurations on the boundary layer excitation and on the vibroacoustic response.

Oscillating-foil hydrokinetic turbines have gained interest over the years to extract energy from renewable sources. The influence of the sweep angle on the performance of a fully-passive oscillating-plate hydrokinetic turbine prototype was investigated experimentally in the present work. The sweep angle was introduced to promote spanwise flow along the plate in order to manipulate the leading edge vortex (LEV) and hydrodynamically optimize the performance of the turbine. In the present work, flat plates of two configurations were considered: a plate with a 6° sweep angle and an unswept plate (control), which were undergoing fully-passive pitch and heave motions in uniform inflow at the Reynolds numbers ranging from 15 000 to 30 000. The resulting kinematic parameters and the energy extraction performance were evaluated for both plates. Planar (2D) particle image velocimetry (PIV) was used to obtain patterns of the phase-averaged out-of-plane vorticity during the oscillation cycle. The circulation in the wake was then related to the induced-forces on the plate by calculating the moments of vorticity of the LEV with respect to the pitching axis of the plate. Tomographic (3D) PIV was implemented in evaluating the influence of the spanwise flow on the dynamics of the vortex structure in three-dimensional space. The rate of deformation of the vortex length was quantified by calculating the deformation terms embedded in the vorticity equations, then linked to the stability of the vortex. The results show evidence of delay of the shedding of LEV and increased vortex stability, in the case of the swept plate. The manipulation of the LEV by the spanwise flow was related to the induced kinematics exhibited by the prolonged heave forces experienced by the swept plate, which led to the higher power extraction performance at high inflow velocities. In the presence of spanwise flow, positive vortex stretching along the vortex line increased the stabilization of the vortex core and prevented the onset of helical vortex breakdown, observed in the case of the unswept plate. The use of the sweep profile on the plate has led to the improvement of energy extraction performance of the fully-passive hydrokinetic turbine. / Graduate

<div>Positive displacement machines are used in a wide variety of applications, ranging from fluid power where they act as a transmission of power, to lubrication and fluid transport. As the core of the fluid system responsible for mechanical--hydraulic energy conversion, the efficiencies of these units are a major driver of the total efficiency of the system. Furthermore, the durability of these units is a strong decider in the useful life of the system in which they operate.</div><div><br></div><div>The key challenge in designing these units comes from understanding their working principles and designing their lubricating interfaces, which must simultaneously perform a load carrying and sealing function as the unit operates. While most of the physical phenomena relevant to these machines have been studied previously in some capacity, the significance of their mutual interactions has not. For this reason, the importance of these mutual interactions is a fundamental question in these machines that this thesis answers for the first time. In analysis of two different machine types, it is confirmed that mutual interactions of both physical phenomena and neighboring fluid domains of the unit contribute significantly to the overall performance of the machine. Namely, these analyses demonstrate load sharing owing to mutual interactions on average of 20% and as high as 50%, and mutual flow interactions of at least 10%.</div><div><br></div><div>In this thesis, the behavior of the thin films of fluid in the lubricating interfaces of the units, the bodies that make up these films, and the volumes which interface with them will be considered. The resulting coupled problem requires a model that can consider the effects of motion of all floating bodies on all films and volumes, and collect the resulting loads applied by the fluid as it responds. This will require a novel 6 degree of freedom dynamics model including the inertia of the bodies and the transient pressure and shear loads of all interfaces of the body and the fluid domain.</div><div><br></div><div>During operation, fluid cavitation and aeration can occur in both the displacement chambers of the machine and its lubricating interfaces. To capture this, a novel cavitation algorithm is developed in this thesis, which considers the release of bubbles due to both gas trapped within the fluid and vaporization of the operating fluid in localized low pressure regions of the films. In the absence of cavitation, this model will also be used to find the pressures and flows over the film, communicating this information with the remainder of the fluid domain.</div><div><br></div><div>Due to the high pressures that form in these units, the bodies deform. The resulting deformation changes the shape of the films and therefore its pressure distribution. This coupled effect will be captured in one of two ways, the first relying on existing geometric information of the unit, and the other using a novel analytical approach that is developed to avoid this necessity. In either case, the added damping due to the shear of the materials will be considered for the first time. Additionally in regions of low gap height, mixed lubrication occurs and the effects of the surface asperities of the floating bodies cannot be neglected. Accurate modeling of this condition is necessary to predict wear that leads to failure in these units. This work will then develop a novel implementation for mixed lubrication modeling that is directly integrated into the cavitation modeling approach.</div><div><br></div><div>Finally, effort is made to maintain a generic tools, such that the model can be applied to any positive displacement machine. This thesis will present the first toolbox of its kind, which accounts for all the mentioned aspects in such a way that they can be captured for any machine. Using both multithreaded and sequential implementations, the tool will be capable of fully utilizing a machine on which it is run for both low latency (design) and high throughput (optimization) applications respectively. In order to make these applications feasible, the various modules of the tool will be strongly coupled using asynchronous time stepping. This approach is made possible with the development of a novel impedance tensor of the mixed universal Reynolds equation, and shows marked improvements in simulation time by requiring at most 50% of the simulation time of existing approaches.</div><div><br></div><div>In the present thesis, the developed tool will be validated using experimental data collected from 3 fundamentally different machines. Individual advancements of the tool will also be verified in isolation with comparison to the state of the art and commercial software in the relevant fields. As a demonstration of the use of the tool for design, detailed analysis of the displacing actions and lubricating interfaces of these same units will be performed. These validations demonstrate the ability of the tool to predict machine efficiencies with error averaging around 1% over all operating conditions for multiple machine types, and capture transient behavior of the units. To demonstrate the utility as a virtual optimization tool, design of a complete external gear machine design will be performed. This demonstration will start from only analytical parameters, and will track a route to a complete prototype.</div>

Das Innenohr des Säugetieres ist ein hochspezialisiertes sensorisches System, das durch ein komplexes mechanisches Verhalten gekennzeichnet ist. Neben der komplizierten Morphometrie und Geometrie kommen auch dem richtungsabhängigen Materialverhalten eine wesentliche Bedeutung zu. Es zeigt sich, daß im Cortischen Organ mit der äußeren Haarsinneszelle ein Zelltyp vorliegt, der durch seine physikalischen Eigenschaften das Gesamtverhalten des Innenohres maßgeblich beeinflußt. Wie jede tierische Zelle weist die äußere Haarsinneszelle als biomechanisches System eine heterogene Mikrostruktur auf. Vom mechanischen Standpunkt aus gesehen, ist neben der mehrschichtigen basolateralen Zellwand jede Einzelzelle durch ein viskoses inneres Fluid (Zellplasma) und einen Zellkern (Nukleus) gekennzeichnet. Die resultierenden mechanischen Eigenschaften des Gesamtsystems ”äußere Haarsinneszelle” können durch Experimente und eine geeignete Modellierung determiniert werden. In dieser Arbeit wird ein neuer Ansatz zur Bestimmung der viskoelastischen Materialeigenschaften der basolateralen Wand vorgestellt. Durch Anwendung einer effektiven Fluid-Struktur-Interaktion wird das Gesamtsystem geschlossen untersucht und eine umfangreiche Materialparameterstudie durchgeführt. Dabei werden im Rahmen der Kontinuumsmechanik gültige Materialgesetze angewendet. Das durch partielle Differentialgleichungen formulierte mechanische Feldproblem wird im Rahmen der Finiten-Elemente-Methode approximiert, was zu einem linearen Gleichungssystem führt. Auf dieser Grundlage wird ein Finite-Elemente-Modell der äußeren Haarsinneszelle entwickelt. Die zur Beschreibung notwendigen Systemmatrizen – insbesondere die Dämpfungsmatrix – basieren dabei vollständig auf einem viskoelastischen Materialgesetz. Die benutzte Methodik läßt weiterhin eine effiziente Berechnung im Frequenzbereich zu. Es zeigt sich, daß eine spezielle Dämpfungsformulierung die experimentell bestimmten dynamischen Eigenschaften der Zelle adäquat widerspiegelt. Eine Analyse auf Materialgesetzebene zeigt, daß dafür reine Schubdämpfung und damit eine spezielle Anisotropie im Viskositätstensor verantwortlich ist. Diese Eigenschaft bestimmt das dynamische Verhalten der äußeren Haarsinneszelle bis mindestens 10 kHz und liegt damit im Hörbereich. Der Modellierung der Zelle geht eine angepaßte Auswertung der experimentell ermittelten Daten voraus. Die mechanisch geeignete Auswertung der zugrundeliegenden Experimente weist dabei auf mögliche Fehlerquellen bei der Analyse der Rohdaten hin. Das hat zur Konsequenz, daß der experimentellen Umgebung die gleiche Aufmerksamkeit geschenkt werden muß wie dem Meßobjekt selbst. Nur so kann eine geeignete Extraktion der für das Meßobjekt spezifischen Eigenschaften erfolgen.

Les travaux de cette thèse s’inscrivent dans le cadre d’une collaboration CIFRE entre Naval Group, le Laboratoire d’Hydrodynamique, d’Energétique et d’Environnement Atmosphérique (LHEEA) de l’Ecole Centrale de Nantes et de l’Institut de Génie des Matériaux (GeM) de l’ICAM de Nantes, sur la problématique de développement d’hélices marines plus efficaces, plus discrètes, et plus respectueuses de l’environnement. Une des solutions passe par le développement de structures composites plus légères et plus flexibles, capables de se déformer passivement pour s’auto-adapter à l’écoulement incident grâce à leurs propriétés spécifiques de couplage flexion-torsion. En lien direct avec cette problématique, nous avons réalisé les travaux de thèse en deux temps. Nous avons dans un premier temps mis en place, montré la faisabilité et validé une méthode de couplage numérique fluidestructure implicite fort entre les codes commerciaux de CFD Starccm+ et de CSD Abaqus pour un cas d'application issu de résultats expérimentaux disponibles dans la littérature sur deux hydrofoils flexibles déformables. Puis dans un second temps nous avons développé, réalisé et testé, dans le tunnel hydrodynamique de l’Ecole Navale, un profil portant composite spécifiquement conçu pour s’approcher du comportement d’une hélice. L’étude expérimentale et numérique de cette configuration nous a permis i) d’éprouver l'utilisation en milieu académique et industriel de nouvelles méthodes expérimentales d'instrumentation d'une pièce composite par insertion directe de fibres optiques dans les plis de composite, et d'une méthode mixte numérique - expérimentale de calibration fine d'un modèle numérique structure ; ii) d’apporter un éclairage sur la physique de l'interaction fluidestructure se produisant sur une surface portante composite ; et iii) de préciser les limitations actuelles concernant la diffusion en milieu industriel de cette méthode de couplage numérique fluide-structure. / This Ph.D is sponsored by the French company Naval Group in collaboration with LHEEA Laboratory from Ecole Centrale de Nantes and GeM Institute from ICAM de Nantes, and deals with the development of new composite marine propellers with improved efficiency, improved acoustic discretion and more environment-friendly. One of the key solutions lies in the application of composite materials to marine structures, in order to benefit from their reduced weight, increased flexibility and bend-twist coupling capacity. Indeed, the latter enables the shape-adaptability of the structure to passively adapt to the incoming flow. To meet this challenge, we first set-up a tightly coupled numerical fluid-structure method using two commercial CFD (Starccm+) and CSD (Abaqus) solvers on two flexible hydrofoils and we validated this method against experimental results available in the literature. Second, we specifically developed a composite hydrofoil to behave closely like a marine propeller and tested it in the hydrodynamic tunnel of the Ecole Navale. Thanks to the combined experimental and numerical analysis of this composite hydrofoil we reached the following conclusions: i) we helped demonstrate the industrial application of a state-of-the-art strain measurement technique using optical fibers directly embedded within the composite plies, ii) we provided some insights into the physics of the fluid-structure interaction occurring on composite hydrofoils and iii) we presented the current limitations of this coupled numerical fluid-structure method relatively to its industrial application.

Cette thèse est consacrée à l'analyse mathématique et numérique d'un problème d'interaction fluide-structure stationnaire, couplant un fluide newtonien, visqueux et incompressible, modélisé par les équations de Stokes 2D et une structure déformable, décrite par les équations d'une poutre 1D. Le fluide et la structure sont couplés via une condition aux limites de type frottement à l'interface.Dans l'étude théorique, nous montrons un résultat d'existence et unicité de solutions faibles, dans le cadre de petits déplacements, du problème de couplage fluide structure avec une condition de glissement de type Tresca en utilisant le théorème de point fixe de Schauder.Dans l'analyse numérique, nous étudions d'abord, l'approximation du problème de Stokes avec la condition de Tresca par une méthode d'éléments finis mixtes à quatre champs. Nous montrons ensuite une estimation d'erreur a priori optimale pour des données régulières et nous réalisons des tests numériques. Enfin, nous présentons un algorithme de point fixe pour la simulation numérique du problème couplé avec des conditions aux limites non linéaires. / This PHD thesis is devoted to the theoretical and numerical analysis of a stationary fluid-structure interaction problem between an incompressible viscous Newtonian fluid, modeled by the 2D Stokes equations, and a deformable structure modeled by the 1D beam equations.The fluid and structure are coupled via a friction boundary condition at the fluid-structure interface.In the theoretical study, we prove the existence of a unique weak solution, under small displacements, of the fluid-structure interaction problem under a slip boundary condition of friction type (SBCF) by using Schauder fixed point theorem.In the numerical analysis, we first study a mixed finite element approximation of the Stokes equations under SBCF.We also prove an optimal a priori error estimate for regular data and we provide numerical examples.Finally, we present a fixed point algorithm for numerical simulation of the coupled problem under nonlinear boundary conditions.

<p>  </p> <p>Three key challenges in developing advanced image-based computational models of cerebral aneurysms are: (i) disentangling the effect of biomechanics and confounding clinical risk factors on aneurysmal progression, (ii) accounting for arterial wall mechanics, and (iii) incorporating the effect of surrounding tissue support on vessel motion and deformation. This thesis addresses these knowledge gaps by developing fluid-structure interaction (FSI) models of subject-specific geometries of cerebral aneurysms to elucidate the effect of coupled hemodynamics and biomechanics. A consistent methodology for obtaining physiologically realistic computational FSI models from standard-of-care imaging data is developed. In this process, a novel technique to estimate heterogeneous arterial wall thickness in the absence of subject-specific arterial wall imaging data is proposed. To address a limitation in the mesh generation workflow of the state-of-the-art cardiovascular flow modeling tool SimVascular, generation of meshes with boundary-layer mesh refinement near the blood-vessel wall interface is proposed for computational geometries with nonuniform wall thickness. Computational murine models of thoracic aortic aneurysms were developed using the proposed methodology. These models were used to inform external tissue support boundary conditions for human cerebral aneurysm subjects via a scaling analysis. Then, the methodology was applied to subjects with multiple unruptured cerebral aneurysms. A comparative computational FSI analysis of aneurysmal biomechanics was performed for each subject-specific pair of computational models for the stable and growing aneurysms, which act as self-controls for confounding clinical risk factors. A higher percentage of area exposed to low shear and high median-peak-systolic arterial wall deformation, each by factors of 1.5 to 2, was observed in growing aneurysms, compared to stable ones. Furthermore, a novel metric – the oscillatory stress index (OStI) – was defined and proposed to indicate locations of oscillating arterial wall stresses. Growing aneurysms demonstrated significant areas with a combination of low wall shear and low OStI, which were hypothesized to be associated with regions of collagen degradation and remodeling. On the other hand, such regions were either absent (or were a small percentage of the total aneurysmal area) in the stable cases. This thesis, therefore, provides a groundwork for future studies, with larger patient cohorts, which will evaluate the role of these biomechanical parameters in cerebral aneurysm growth.</p>

The aim of this master thesis is to characterize the fluid forces applied to a fuel assembly inthe core of a nuclear power plant in case of seism. The forces are studied with a simplifiedtwo-dimensional model constituted of an array of 3 by 3 infinite cylinders oscillating in aclosed box. The axial flow of water, which convects the heat in the core of a nuclear powerplant, is also taken into account. The velocity of the axial flow reaches 4m/s in the middle ofthe assembly and modifies the forces features when the cylinders move laterally.The seism is modeled as a lateral displacement with high amplitude (several cylinderdiameters) and low frequencies (below 20 Hz). In order to study the effects of the amplitudeand of the frequency of the displacement, the displacement taken is a sine function withboth controlled amplitude and frequency. Four degrees of freedom of the system will bestudied: the amplitude of the displacement, its frequency, the axial velocity amplitude andthe confinement (due to the closed box).The fluid forces exerted on the cylinders can be seen as a combination of three terms: anadded mass, related to the acceleration of cylinders, a drift force, related to the damping ofthe fluid and a force due to the interaction of the cylinder with residual vortices. The firsttwo components will be characterized through the Morison expansion, and their evolutionwith the variation of the degree of freedom of the system will be quantified. The effect ofthe interaction with the residual vortices will be observed in the plots of the forces vs. timebut also in the velocity and vorticity map of the fluid.The fluid forces are calculated with the CFD code Code_Saturne, which uses a second orderaccurate finite volume method. Unsteady Reynolds Averaged Navier Stokes simulations arerealized with a k-epsilon turbulence model. The Arbitrary Lagrange Euler model is used todescribe the structure displacement. The domain is meshed with hexahedra with thesoftware gmsh [1] and the flow is visualized with Paraview [2]. The modeling techniquesused for the simulations are described in the first part of this master thesis.

Les tubes des générateurs de vapeur des centrales nucléaires vibrent sous l'effet d'écoulement eau/vapeur. Pour appréhender ce phénomène et le comprendre, des expériences à échelles réduites sont réalisées. La simulation numérique a montré son habilité à reproduire l'interaction fluide-structure sur ce type de géométrie pour des écoulements monophasiques. L'objectif est désormais de faire de même en écoulement diphasique et de caractériser les propriétés physiques du mélange liquide/gaz influant sur la vibration.Pour se faire, un code CFD avec une approche bi-fluide est utilisé. Une méthode dite de "Discrete forcing" est implémentée pour permettre le mouvement imposé de corps solides au sein d'un écoulement à plusieurs phases. Celle-ci est alros validée sur des cas simples et intégraux avec une comparaison systématique à des résultats expérimentaux ou théoriques.En se basant sur un algorithme implicite existant dans la littérature, un couplage fluide-structure utilisant cette méthode de suivi d'interface est implémenté. Validé sur des cas monophasiques et diphasiques, ce couplage offre désormais la possibilité de déplacer un solide en fonction des forces fluides diphasiques qui lui sont appliquées.Les différentes méthodes numériques présentes dans NEPTUNE_CFD sont ensuite évaluées pour un écoulement fréon/fréon au travers d'un faisceau de tubes inclinés. La nécessité d'utiliser des modèles dit "multi-régime" est mis en avant.Afin de déterminer l'influence sur l'écoulement des différentes propriétés physiques d'un mélange diphasique, plusieurs cas simples sont réalisés.Finalement, l'application industrielle cible, un écoulement eau/fréon dans un faisceau de tubes à pas carré, est simulée et comparée à un écoulement en conditions réelles (eau/vapeur à 70 bar). Les vibrations induites par écoulement monophasique puis diphasique sont correctement reproduites sur des cas dit de "faisabilité". / In nuclear power plants, steam generator tubes vibrate because of steam/water cross-flows. In order to understant this phenomenon, reduced-scale experiments are performed. Numerical simulations have shown their ability to accurately reproduce the vibration induced by a single phase flow in a tube bundle. The aim of the present work is to do the same with two-phase flow and to characterize the effect of the mixture physical properties on vibration.To do so, a CFD code based on a two-fluid approach is used. A "discrete forcing" method is implemented in order to allow solid body motion in a two-phase flow. The validation is performed with simple and industrial cases using experimental and theoretical results.Using an existing implicit algorithm, a fluid-structure coupling based on the developed interface tracking method is implemented. Validated for single and two-phase flows, it is now possible to have solid motion induced by fluid forces.The different numerical models dedicated to two-phase flows are then evaluated on a freon/freon flow across an inclined tube bundle. The use of a multi-regime model is required. In order to investigate the role of the different physical properties on the vibration, three simple studies are performed.Finally, the industrial application, a freon/water flow across a square pitch tube bundle, is performed. First, it is compared to a steam/water flow in order to characterize the discrepancies when we are using a modeling mixture. Then, the vibration induced by single- and two-phase flows is reproduced by the developed method on feasibility test cases.