Teses / dissertações sobre o tema "Mécanique des fluides – Simulation par ordinateur"
Crie uma referência precisa em APA, MLA, Chicago, Harvard, e outros estilos
Veja os 50 melhores trabalhos (teses / dissertações) para estudos sobre o assunto "Mécanique des fluides – Simulation par ordinateur".
Ao lado de cada fonte na lista de referências, há um botão "Adicionar à bibliografia". Clique e geraremos automaticamente a citação bibliográfica do trabalho escolhido no estilo de citação de que você precisa: APA, MLA, Harvard, Chicago, Vancouver, etc.
Você também pode baixar o texto completo da publicação científica em formato .pdf e ler o resumo do trabalho online se estiver presente nos metadados.
Veja as teses / dissertações das mais diversas áreas científicas e compile uma bibliografia correta.
Pouchol, Mickaël. "Structures hiérarchiques pour la simulation de fluides". Limoges, 2010. https://aurore.unilim.fr/theses/nxfile/default/93ee02e3-c56f-4e8f-9b5a-8300d06a4c15/blobholder:0/2010LIMO4025.pdf.
Texto completo da fonteFluid flow simulation such as water or smoke is one of the most covered natural phenomena in the computer graphics community. Among these methods, lagrangian ones which divide the fluid into a set of particles, are usually favored for their efficiency, however they require suitable data structures in several steps to reduce their high computation times. We suggest to use a hierarchical hash table to handle the collision detection step with irregularly distributed and objects with variable sizes. The use of several hash levels allows to benefit from objects spatial coherence and to drastically reduce hash collisions. Adaptive methods allow to efficiently allocate computational resources to significant fluid regions by varying particles sizes during the simulation. In this case a hierarchical grid or hash table structure allows to efficiently handle nearest neighbors search by inserting each particle in the appropriate grid level, and also allows to define appropriate merging or splitting regions with local criteria. Finally the visualization step is challenging because the end-user will appreciate the final result directly. For lagrangian methods, the most common way to deal with this problem is to use blob-based methods that do not yield satisfying results ; we use the variational surface method for this task
Michel, Bertrand. "Contribution à la simulation numérique efficace des écoulements dans les prises d'air supersoniques". Paris, ENSAM, 2004. http://www.theses.fr/2004ENAM0014.
Texto completo da fonteJoubert, de la Motte Pascal. "Schéma à pas de temps caractéristique pour l'aérodynamique transsonique et conception aéroélastique optimale d'aile de grand allongement". Paris, ENSAM, 2007. https://pastel.archives-ouvertes.fr/tel-00144844.
Texto completo da fonteThe MBDA Aerolog software is commonly used for computing steady compressible flows over complex geometries. The residual-distribution scheme of Lax-Wendroff-Ni is coupled with an implicit residual smoothing technique to provide a low cost per iteration and a flexible use in multi-block computational domains. In order to improve the code efficiency for the simulation of unsteady flows a so-called characteristic timestep upwind scheme is proposed in this work: its compact stencil allows to preserve the existing boundary and interface treatments. This scheme is coupled with a new matrix-free implicit stage and applied in the framework of a dual time-step technique. The results provided by this new approach are analyzed from the viewpoint of accuracy and efficiency for steady and unsteady flows over airfoils and rudders and compared with those produced by the existing method. The optimal design of a wing is studied next in the context of static aeroelasticity: since the wings of a cruise missile are hollowed out, one seeks to take advantage of the associated aeroelastic effects to maximise the range of the missile. A fully automated aeroelastic computational chain is developed to perform a numerical study of these effects and a first parametric study is carried out to assess the relevance of the chosen parameters and the accuracy of the methods employed
Cherubini, Stefania. "Linear and non-linear global instability of attached and separated boundary-layer flows over a flat plate". Paris, ENSAM, 2010. http://www.theses.fr/2010ENAM0012.
Texto completo da fonteLe but de cette thèse est de décrire en détail la dynamique linéaire et non linéaire d'une couche limite attachée sur une plaque plane à bas nombre de Reynolds. La dynamique linéaire, pilotée par les interactions entre les vecteurs propres non-orthogonaux, est étudiée à travers deux méthodes différentes d’instabilité globale : une analyse globale aux vecteurs propres et une optimisation directe-adjointe. Dans ces analyses globales, aucune structure spatiale n’est imposée à priori pour la perturbation, les effets convectifs dus au fort non parallélisme de l’écoulement sont pris en compte. Pour le cas de la couche limite décollée, le déclenchement des instationnarités a été clarifié : i) pour une forte amplification des perturbations de nature convective et bidimensionnelle ; ii) pour des effets de non normalité longitudinale engendrant le phénomène du flapping ; iii) pour une forte sensibilité vis-à-vis d’un forçage harmonique ; iv) pour un monde tridimensionnel globalement instable. Pour une couche limite attachée, le but a été d’identifier les perturbations localisées caractérisées par des fréquences multiples dans les directions longitudinales et transversales de l’écoulement, induisant une amplification de l’énergie des perturbations. Pour évaluer les effets les effets de la non linéarité dans les mécanismes d’instabilité identifiés par les analyses de stabilité globale, des simulations numériques directes ont été réalisées pour les écoulements de couche limite attachées ou décollées, bidimensionnelles et tridimensionnelles. La dynamique des perturbations permettant une plus rapide vers la turbulence a été étudiée. Différents scénarios de transition ont été observés, les différents mécanismes de transition ont été analysés
Mougin, Guillaume. "Interactions entre la dynamique d'une bulle et les instabilités de son sillage". Toulouse, INPT, 2001. http://www.theses.fr/2001INPT042H.
Texto completo da fonteCastaldi, Stéphane. "Étude de la turbulence cinématiquement homogène isotrope et massiquement inhomogène par simulation numérique directe". Toulouse, INPT, 1999. http://www.theses.fr/1999INPT008H.
Texto completo da fonteSergent, Emmanuel. "Vers une méthodologie de couplage entre la simulation des grandes échelles et les modèles statistiques". Ecully, Ecole centrale de Lyon, 2002. http://www.theses.fr/2002ECDL0019.
Texto completo da fonteThe numerical simulation in turbulent fluids mechanics is based on three different approaches. Direct Numerical Simulation (DNS), Large Eddy Simulation (LES) and statistical RANS methods. The work completed in this Phd aims at defining a coupling methodology between LES and RANS. These two approaches are indeed complementary for industrial applications. RANS methods will calculate the total flow field and LES will be used as "zoom" in a precise zone of the geometry. With an aim of carrying out the coupling, three approaches are adopted : one for a coupling in the direction RANS towards LES (named MV), one in the direction LES towards RANS and for tangent coupling. After a series of tests aiming at optimizing the choice of the parameters involved in the MV, these three techniques were applied to an academic case : the flow between two parallel plane plates. Finally the MV method was tested on a more complex flow : the flow around a forward-backward facing step
Agbaglah, Gbémého Gilou. "Dynamique et instabilité des nappes liquides". Paris 6, 2011. http://www.theses.fr/2011PA066195.
Texto completo da fonteVintimiglia, Anne-Sophie. "Implémentation d'une cinétique de cristallisation des polymères dans une simulation numérique bidimensionnelle par éléments finis : application à l'injection des thermoplastiques et au procédé push-pull". ENSMP, 1999. http://www.theses.fr/1999ENMP0868.
Texto completo da fonteMoussière, Sandrine. "Etude par simulation numérique des écoulements turbulents réactifs dans les réacteurs d'oxydation hydrothermale : application à un réacteur agité double enveloppe". Aix-Marseille 3, 2006. http://www.theses.fr/2006AIX30045.
Texto completo da fonteSupercritical water oxidation is an innovative process to treat organic liquid waste which uses supercritical water properties to mix efficiency the oxidant and the organic compounds. The reactor is a stirred double shell reactor. In the step of adaptation to nuclear constraints, the computational fluid dynamic modeling is a good tool to know required temperature field in the reactor for safety analysis. Firstly, the CFD modeling of tubular reactor confirms the hypothesis of an incompressible fluid and the use of k-[oméga] turbulence model to represent the hydrodynamic. Moreover, the EDC model is as efficiency as the kinetic to compute the reaction rate in this reactor. Secondly, the study of turbulent flow in the double shell reactor confirms the use of 2D axisymetric geometry instead of 3D geometry to compute heat transfer. Moreover, this study reports that water-air mixing is not in single phase. The reactive turbulent flow is well represented by EDC model after adaptation of initial conditions. The reaction rate in supercritical water oxidation reactor is mainly controlled by the mixing
Martin, Vincent. "Simulations multidomaines des écoulements en milieu poreux". Paris 9, 2004. https://portail.bu.dauphine.fr/fileviewer/index.php?doc=2004PA090006.
Texto completo da fonteThis thesis is mainly concerned with the multidomain simulations of flow in porous media. Three different themes are considered. First, we study a domain decomposition method with non-matching meshes using Robin type interface conditions, for the mixed finite elements. Second, this method is implemented in parallel using the parallel system OcamlP3l, written in Ocaml by computer scientists. In OcamlP3l, the user develops and debugs sequentially, and obtains the parallel code with a mere recompilation. A realistic 3D simulation is given to validate the procedure. Finally, we present a new model for flow in a porous medium containing large fractures that may have very large and/or very small permeabilities. In this model, the fractures are treated as interfaces between subdomains. Existence and uniqueness of the solution is proved and an error estimate is obtained. Some numerical experiments show the quality of the results
Joly, Cécile. "Simulations numériques d'un jet rond turbulent et de son interaction avec un tourbillon de sillage". Université de Marne-la-Vallée, 2002. http://www.theses.fr/2002MARN0147.
Texto completo da fonteThe general context of this study concerns the impact of contrails, these famous white plumes frequently observed in the aircraft wake, on the atmosphere. From an aerodynamic point of view, the formation of the contrails is characterised by the interaction between a turbulent jet and a wing-tip vortex. The aim of this thesis is to contribute to a better understanding of the thermal and dynamic phenomena involved in this flow. This work is based on the numerical resolution of the three-dimensional, unsteady and compressible Navier-Stokes and energy conservation equations. Two approaches are considered: the direct numerical simulation and the large eddy simulation. A temporal simulation of the transition to turbulence of a non-isothermal jet is performed without accounting for the vortex flow field. A the end of this simulation, a vortex model is superimposed on the jet flow field. The first part of this thesis describes the two approaches, the different subgrid models chosen for the large eddy simulations, and the numerical techniques employed. The second part is devoted to the jet flow simulation, and here the objective is to determine the subgrid model appropriated to this flow configuration. The third part is dedicated to the simulation of interaction between the jet and the vortex. Results are compared to experimental data. The simulations have demonstrated the development of large scale structures all around the vortex core. The temperature field concentrates in the large scale structures
Helin, Lionel. "Développement de modèles algébriques explicites pour les fluides viscoélastiques". Lille 1, 2006. https://pepite-depot.univ-lille.fr/LIBRE/Th_Num/2006/50376-2006-Helin.pdf.
Texto completo da fonteOtheguy, Mendionde Pantxika. "Dynamique de l'appariement tourbillonnaire en milieu stratifie et stratifie tournant". Palaiseau, Ecole polytechnique, 2005. http://tel.archives-ouvertes.fr/docs/00/06/19/56/PDF/Pantxika_Otheguy.pdf.
Texto completo da fonteOne of the main characteristics of geophysical flows is the presence of stable stratification, i. E the presence of a vertical density gradient. Such flows are generally organized into thin horizontal layers. This PhD thesis investigates numerically, theoretically and experimentally the dynamics of a pair of co-rotating vertical vortices in a stratified fluid in order to understand the mechanism of formation of such layered structure. Such flow is of high interest because co-rotating vortex pairs play an important role in two-dimensional turbulence through the merging process which leads to the energy transfer from small to large scales. A three-dimensional instability has been discovered. It is similar to the zigzag instability observed on a pair of counter-rotating vortices in a stratified fluid and it decorrelates the vortex merging along the vertical. A numerical three-dimensional stability analysis has shown that the zigzag instability dominates for strong stratification whereas the elliptic instability dominates for weak stratification. The zigzag instability bends symmetrically the two vortices without significant internal deformation. It selects a wavelength proportional to Fh b, where Fh is the horizontal Froude number and b the separation distance between the vortices. Its growth rate is proportional to the strain rate. An asymptotic analysis has recovered these scaling laws and has shown that the zigzag instability comes from the coupling between vortex bending modes and the strain field that each vortex generates on the other. Experimental observations and Direct Numerical simulations have also shown that the vertical lengthscale selected by the instability agrees with the predictions. Of the stability analyses. The non-linear effects do not slow down the instability development but tends to recorrelate the fluid within each layers and to expel vertical gradients in between these layers. In the final part, the effect of planetary rotation has been investigated. It is shown that the zigzag instability remains active with a growth rate independent of the Coriolis parameter. In the rapidly rotating regime (quasi-geostrophic), the zigzag instability is similar to the "tall-column" instability with the most unstable wavelength proportional to Fh b / Ro, where Ro is the Rossby number. This instability may therefore affect geophysical flows for a large range of parameters both in the oceans and the atmosphere
Bailly, Pascal. "Contribution à l'étude de l'interaction turbulence - combustion dans les flammes turbulentes de prémélange à l'aide de modèles du second ordre". Poitiers, 1996. http://www.theses.fr/1996POIT2279.
Texto completo da fonteRahmani, Ariane. "Simulation des grandes Échelles pour les incendies en tunnels". Lyon 1, 2006. http://www.theses.fr/2006LYO10032.
Texto completo da fonteDautriat, Jérémie. "Comportement hydromécanique de roches réservoir sous contraintes : relations entre évolution de perméabilité et échelles des mécanismes d'endommagement". Palaiseau, Ecole polytechnique, 2009. http://tel.archives-ouvertes.fr/docs/00/38/57/77/PDF/Manuscrit_These_Dautriat.pdf.
Texto completo da fonteThe Understanding and the prevention of damage mechanisms, which have an impact on the kinetics of production and the rate of recovery, remain an outstanding issue in reservoir engineering. The aim of this study, based on the characterization of the hydromechanical behaviour of sandstone and carbonate rocks in 'reservoir conditions', is the identification of the local mechanisms responsible for changes in permeability measured macroscopically. The experimental work was performed with an original triaxial set-up, which allows measurements of the stress-induced permeability evolutions in the principal directions of the stress tensor. A first experimental campaign, consisting in hydrostatic and uniaxial compression tests, has been performed on Fontainebleau and Bentheim sandstones. In one hand, we showed that experimental end-effects might affect significantly the ‘classical' axial permeability measurements, and in the other hand, we have determined the impact of brittle failure on directional permeabilities. Compression experiments, following different stress-paths, were also carried out on a carbonate, the Estaillades limestone. In elastic deformation regime, the reduction of permeability was modelized by pore network simulations, based on 3D reconstructions of μ-tomography RX and a simplified representation of the pore space. While brittle fracture of carbonate samples induced slight permeability evolutions, at higher effective pressure, permeability drops can reach 90% to the initial values and are associated with mechanisms of pore collapse. The post-mortem analysis of deformed samples, coupled with digital image correlation methods, using both SEM and Optical acquisition devices, provide a better understanding of the role of the heterogeneities, identified at different scales, in the strain localization and their potential impacts on permeability changes at the sample scale
Dridi, Walid. "Influence de l’acoustic streaming sur les instabilités des écoulements chauffés latéralement". Lyon 1, 2008. http://n2t.net/ark:/47881/m6mc8x5j.
Texto completo da fonteWe study the effect of acoustic streaming (steady flow generated by progressive acoustic waves) on the stability of convective flows associated with crystal growth from melt in horizontal Bridgman configurations. We consider two simple configurations: an extended fluid layer submitted to a horizontal temperature gradient and a laterally heated parallelepipedic cavity. In both cases, the dependence of the critical Grashof number Grc on the acoustic intensity (monitored through the acoustic parameter A) is determined for given values of the Prandlt number and of the dimension of the acoustic source Hb. In the case of the extended fluid layer, it is shown that for rather small beam widths Hb, the acoustic streaming destabilizes the buoyant flows, but for a large beam width, a range of acoustic intensities A is found for which the buoyant flows are stabilized. An adequate decentring of the beam can also enhance the stabilization. In the case of the parallelepipedic cavity of length Ax, both the structure of the flows and their stability properties are determined. The flows have different symmetries, but these symmetries are generally broken at the first bifurcation points. Bifurcation diagrams are obtained which show that the flows become oscillatory periodic at a Hopf bifurcation, either directly on the primary steady solution branch, or on a secondary branch which bifurcates from the primary branch at a steady bifurcation point. The critical Grashof numbers for these bifurcation points are calculated as a function of the cavity length Ax, the Prandtl number Pr and the acoustic streaming parameter A. The thresholds are generally found to increase when the acoustic streaming contribution is enhanced, which indicates a stabilizing effect induced by acoustic streaming and may explain the observed improvement of the crystal quality when ultrasound waves are applied during the growth process. Destabilization effects are, however, found in some parameter range
Aubard, Guillaume. "Simulation des grandes échelles des instationnarités basses fréquences d'une interaction onde de choc couche limite sur plaque planeTexte imprimé". Paris, ENSAM, 2012. http://www.theses.fr/2012ENAM0019.
Texto completo da fonteShock wave / turbulent boundary layer interactions arising on aeronautical vehicles are prone to unsteadiness with a low frequency of few hundred Hertz, which can induce severe aerodynamic loads. The present study is done in the framework of the ANR project SPICEX, which deals with the numerical modeling of these low-frequency phenomena. The unsteady dynamics of a shock wave / turbulent boundary layer interaction (SWTBLI) on a flat plate is investigated. A numerical tool based on Large Eddy Simulation (LES) is developed for the compressible regime. In particular, efficient strategies for shock capturing and turbulent inflow generation are implemented. A particular attention is devoted to the validation of the present algorithms, notably concerning the LES models. Turbulent boundary layer and SWTBLI simulations are conducted, and give confidence in the present numerical strategy. LES of an unsteady SWTBLI covering dozens of low-frequency cycles is performed. The results suggest that the low-frequency dynamics is related to a global synchronization of the flow, characterized by a cyclic breathing motion of the recirculation bubble, associated with a fore-and aft motion of the reflected shock, together with a modulation of the mixing layer. The study shows that the coupling between the restoring force due to the incident shock and the disturbances generated by the incoming turbulence and the mixing layer is responsible for the low-frequency broadband modulation of the vortex shedding in the downstream flow, corresponding statistically to the low frequency of interest
Renard, Paul-Henri. "Etudes d'interactions flamme-tourbillon : implications pour la combustion turbulente". Châtenay-Malabry, Ecole centrale de Paris, 1999. http://www.theses.fr/1999ECAP0650.
Texto completo da fonteParnaudeau, Philippe. "Etude numérique d'un écoulement cisaillé turbulent complexe à basse vitesse : application à la protection rapprochée". Poitiers, 2004. http://www.theses.fr/2004POIT2311.
Texto completo da fonteComplex wakes are studied by Direct and Large Eddy Simulations. In order to take account of the obstacle, a method of forcing was established in numerical code which uses numerical methods of high precision on a Cartesian grid. In order to keep accurately best possible in modelling the flow around a cylinder, the method of forcing underwent modifications. Thus using a flow mirror, definite within the obstacle, we improved this method. The DNS around a cylinder highlight at these improvements. In particular, we highlighted the importance of using high order space discretizations schemes within the framework of our study. The study of a wake of cylinder for a Reynolds number of 3900 was used to validate our approach within the framework of the LES. Finally a preliminary study of a mixing layer in interaction with a wake is proposed
Nguyen, Thi Phuong. "Simulation numérique et analyse physique des couches limites turbulentes compressibles sous l'influence des gradients de pression". Toulouse, INPT, 1998. http://www.theses.fr/1998INPT041H.
Texto completo da fonteNguyen, Van Thinh. "Modélisation de l’interaction entre hydroliennes et le courant dans un courant de marée comme celui du Raz Blanchard". Caen, 2015. http://www.theses.fr/2015CAEN2036.
Texto completo da fonteAmong the existing marine renewable energies, the exploitation of tidal current raises a particular interest in the Cherbourg peninsula where the Raz Blanchard, the world's second strongest current, is located. The reliability of the methods used to optimize the arrangement of tidal turbines within a tidal farm must be improved. A particular attention should be paid to the modelling of the turbulence because it strongly affects the interactions between the turbines and the local currents. In this thesis, a methodology based on the Actuator Disk (AD) concept is established. The tidal turbine is represented by a porous disk having a given thrust. The model results are compared to experimental measurements carried out in the wake of a porous disk placed in a laboratory flume. Several turbulence models such as the Standard k-ε and Realizable k-ε models, the SST k-ω model and the Reynolds Stress Model (RSM) are compared. Simulations are performed without and with a source term for Turbulent Kinetic Energy respectively. Simulations are then carried out with one turbine placed in a flow representative of the hydrodynamic conditions of the Raz Blanchard. It permits to investigate the effect of a time-varying current direction on the energy produced by a pilot farm. Finally, a method similar to the AD associated to the Reynolds Averaged Navier-Stokes equations (RANS) is applied to simulate a Darrieus turbine. The model successfully predicts the forces exerted on the blades and the velocities in the wake. It constitutes a first step towards realistic applications intending to optimize the arrangement of vertical axis turbines in a tidal farm
Coppens, Fabien. "Simulations numériques sur le développement de la turbulence dans un tourbillon". Toulouse, INPT, 1998. http://www.theses.fr/1998INPT014H.
Texto completo da fonteSitbon, David. "Étude de stratégies de calcul multidomaine pour la conception de chambres de combustion de statoréacteurs". Toulouse, INPT, 1997. http://www.theses.fr/1997INPT049H.
Texto completo da fonteDeutsch, Emmanuel. "Dispersion de particules dans une turbulence homogène isotrope stationnaire calculée par simulation numérique directe des grandes échelles". Ecully, Ecole centrale de Lyon, 1992. http://bibli.ec-lyon.fr/exl-doc/TH_T1454_edeutsch.pdf.
Texto completo da fonteBournot, Hervé. "Modélisation numérique de la réduction de traînée d'arrière-corps par injection diphasique additionnelle". Aix-Marseille 1, 2004. http://www.theses.fr/2004AIX11058.
Texto completo da fonteMenu, Mélissa. "Champs magnétiques générés par effet dynamo dans les objets astrophysiques en rotation". Thesis, Université Paris-Saclay (ComUE), 2019. http://www.theses.fr/2019SACLX104.
Texto completo da fonteThe objective of this thesis is to investigate the generation of astrophysical magnetic fields via the dynamo effect. The impact of a global rotation and an average magnetic field imposed simultaneously, suitable configuration especially for planetary and stellar interiors, is the main issue of this study. Particular attention is paid to the temporal dynamics of the system, as well as to the topology of the generated field. In order to obtain an overview of the phenomenon, several numerical approaches were used, from the cubic model (TURBO code) focused on the study of local phenomena to the spherical model (PaRoDy code) allowing a more direct comparison with observations. These two types of geometry complement each other : the less constraining local approach makes it possible to observe precisely the consequences of a parameter, while the global approach is subject to more realistic constraints, such as thermal convection. The local model revealed an inverse cascade of hybrid helicity, the second invariant of the doubly anisotropic MHD, whose intensity depends on several parameters. The latter is more important when the polarization favored by the forcing corresponds to magnetostrophic waves (right polarization). Also, the tilt angle between the magnetic field and the rotation axes plays a major role, especially when θ≻35°, the value for which the process causing the cascade changes and weakens it. Finally, increasing the number of magnetic Prandtl (Pm) leads to a faster and more important transfer of magnetic energy to large scales. The global study also shows the key role of Pm in large-scale field generation. Indeed, the new values explored show that strong dipole fields can be maintained in more turbulent regimes. This specific behaviour corresponds to dynamos for which the Lorentz force is not negligible at large scales. The relative importance of inertia compared to magnetic forces is decisive in the topology of the observed field. Thus, the transition from a dipolar regime to a multipolar regime is delayed by the Lorentz force, the essential component of force balances in astrophysics. These results could be applied to various systems, including the geomagnetic field during reversals
Andro, Jean-Yves. "Aérodynamique d'un profil d'aile battante à bas nombre de Reynolds". Poitiers, 2008. http://theses.edel.univ-poitiers.fr/theses/2008/Andro-Jean-Yves/2008-Andro-Jean-Yves-These.pdf.
Texto completo da fonteRecent progress made in the domain of microtechnologies allow the design of very small sized Micro Air Vehicles (MAVs), whose wingspan is inferior than 15 cm, suitable for observation or intervention in hazardous environments. A biomimetic approach using flapping wing locomotion seems to be promising because it could allow hovering flight and great manoeuvrability in confined environments. However, flapping wings aerodynamics at low Reynolds numbers differs from classical applications aerodynamics due to the vortex shedding phenomenon and the superimposition of various unsteady mechanisms. In a first time, we have studied the vortex shedding on an airfoil by characterising theoretically and experimentally its apparition and then by simulating it numerically. Experiments in a water tank have finally specified the three dimensional effects and Reynolds number effects on the efforts generated by the vortex shedding. In a second time, thanks to direct numerical simulation (DNS), we have studied the fundamental movements of the flapping flight, i. E. Pure heaving and pitching movements, so as to emphasize the various unsteady mechanisms, to differentiate the various flapping flight strategies and to propose simplified models that could describe instantaneous efforts
Meunier, Alain. "Simulations numériques de suspensions de sphères dans un fluide visqueux". Nice, 1990. http://www.theses.fr/1990NICE4375.
Texto completo da fonteCarpy, Sabrina. "Contribution à la modélisation instationnaire de la turbulence : modélisations urans et hybride rans/les". Poitiers, 2006. http://www.theses.fr/2006POIT2342.
Texto completo da fonteThe aim of this work is to account for the unsteadiness effects on the turbulence in single point closure. The existence of large scale structures in statistically steady flows leads to reconsider some hypothesis. Much more than adding the time derivatives , the URANS equations needs to consider a new decomposition and an assiociated operator. Therefore, the applicability of usual closure methods has to be examined. For exemple, the periodicity of a synthetic jet leads to a non-equilibrium, which induces a permanent misalignment of anisotropy tensor and strain tensors. RSM are able to reproduce this misalignment, whereas k-ε. Model can't. A seamless hybrid RANS/LES method, based on the version of Schiestel's model, relies on transport equations for the subgrid stress (ij)SGS and dissipation. The decomposition operator is then assimilated as a filter with an adapatative cutoff frequency. The predictions obtained on a temporal mixing layer shows the ability of this model to capture the very large structure of the flow
Madec, Laurent. "Conception, mise en oeuvre et modélisation d'un procédé d'agglomération de particules submicroniques en vue de les rendre filtrables et redispersables". Vandoeuvre-les-Nancy, INPL, 2000. http://www.theses.fr/2000INPL064N.
Texto completo da fonteDerouineau, Stéphanie. "Etude expérimentale et numérique de l'écoulement de convection naturelle généré entre deux volumes de fluide superposés et communiquant au travers d'une ouverture horizontale". Poitiers, 2003. http://www.theses.fr/2003POIT2286.
Texto completo da fonteThis thesis presents an experimental and numerical (DNS 2D) study of natural convection flow induced by exchange between two superposed fluid compartments at different temperatures, through a thin horizontal aperture. Laser tomography visualisations reveal different flows : laminar, pseudo periodical for low Rayleigh number and chaotic for high Rayleigh number. An analysis of complex three-dimensional local exchange mechanisms is proposed. Direct numerical simulations give faithful mechanisms reproduction. Experimentally, net heat transfer quantification shows that characteristic length is independent of compartment's heights but is a complex function of thickness and diameter D, D remaining the main parameter. Numerical simulations confirm this result and give an exchange law proportional to RaD1/2
Gurtner, Gérald. "Géométrie, topologie et optimisation des réseaux et structures cellulaires". Paris 7, 2011. http://www.theses.fr/2011PA077165.
Texto completo da fonteSome particular networks of very different essences - electrical, thermal, fluidic, mecanic - exhibit, in a first approximation, some strong mathematical analogies, allowing us to conduct a common analysis of their emergent properties - electrical, thermal or fluidic conductivity, and elastic moduli. With a variationnal approach, we established absolute bounds on these quantifies as well as a set of geometrical necessary and sufficient conditions (NSC) to reach them. These conditions lead to new optimal structures, both in two and three dimensions. Thanks to a numerical program, which allowed us to verify these predictions, we then characterized the bending/streching transition which appears in fibrous networks. With the help of the NSC, we computed analytically some statistic, microscopic features of these networks, which might be of importance in the future to understand this phenomenon, as our analyze suggests it. Moreover, we used the programm to investigate the problem of the junctions' energy and showed the presence of several transitions, described by power laws. Finally, we calculated the macroscopic characteristics of some networks close to the optimality, and introduced a new average quantity based on the NSC which seemed to be of importance to quantify this deviation from optimality
Léon-Becerril, Elisabeth. "Analyse de stabilité et simulation numérique des colonnes à bulles". Toulouse, INSA, 2001. http://www.theses.fr/2001ISAT0010.
Texto completo da fonteThe objective of this thesis is the numerical simulation of hydrodynamics and mixing in bubble columns. Experimentally, two different flow behaviours for bubble columns can be observed. In the homogeneous flow, the gas bubbles move upward uniformly. If the gas flow rate increases, the system becomes heterogeneous. This flow is characterised by the presence of large structures. These structures have an unsteady behaviour, and it is possible that they enhance the mixing in bubble columns. In the first part, the existence of large structures in bubble columns is explained by a stability analysis. An uniform bubbly flow is perturbed and the analysis of stability determines the conditions at which this perturbation is amplified. This analysis is based on two main approaches presented in the literature. A particular objective of this research was the synthesis of such approaches, taking into account the importance of bubble deformation in stability. The proposed approach was validated with experimental data of gas velocity and kinematic velocity at several liquid flow rates. In addition this approach allows to analyse the stability and to determine the transition between flow regimes in bubble columns and in pipe flows. The second point is devoted to the numerical simulation of bubble columns in unsteady flows. Simulations were performed for a rectangular system without liquid flow and with a punctual gas flow rate (Sokolichin & Eigenberger, 1994). The importance of the added mass force which is modelled as a function of the gas fraction and bubble deformation is determined. Finally, a simplified CFD model of bubble column has been constructed with a rectangular surface and an uniformly distributed gas feed at the bottom. The aim of such simulations was to test different models for the drag and added mass forces. Simulated tracer tests were carried out and the results confirmed the importance of the large structures in terms of the concentration transport
Rakotoarivelo, Hoby. "Contributions au co-design de noyaux irréguliers sur architectures manycore : cas du remaillage anisotrope multi-échelle en mécanique des fluides numérique". Thesis, Université Paris-Saclay (ComUE), 2018. http://www.theses.fr/2018SACLE012/document.
Texto completo da fonteNumerical simulations of complex flows such as turbulence or shockwave propagation often require a huge computational time to achieve an industrial accuracy level. To speedup these simulations, two alternatives may be combined : mesh adaptation to reduce the number of required points on one hand, and parallel processing to absorb the computation workload on the other hand. However efficiently porting adaptive kernels on massively parallel architectures is far from being trivial. Indeed each task related to a local vicintiy need to be propagated, and it may induce new conflictual tasks though. Furthermore, these tasks are characterized by a low arithmetic intensity and a low reuse rate of already cached data. Besides, new kind of accelerators have arised in high performance computing landscape, involving a number of algorithmic constraints. In a context of electrical power consumption reduction, they are characterized by numerous underclocked cores and a deep hierarchy memory involving asymmetric expensive memory accesses. Therefore, kernels must expose a high degree of concurrency and high cached-data reuse rate to maintain an optimal core efficiency. The real issue is how to structure these data-driven and data-intensive kernels to match these constraints ?In this work, we provide an approach which conciliates both locality constraints and convergence in terms of mesh error and quality. More than a parallelization, it relies on redesign of kernels guided by hardware constraints while preserving accuracy. In fact, we devise a set of locality-aware kernels for anisotropic adaptation of triangulated differential manifold, as well as a lock-free and massively multithread parallelization of irregular kernels. Although being complementary, those axes come from distinct research themes mixing informatics and applied mathematics. Here, we aim to show that our devised schemes are as efficient as the state-of-the-art for both axes
Maurel, Philippe. "Analyse et modélisation des courants et de la turbulence sous les vagues de vent". Toulouse, INPT, 1997. http://www.theses.fr/1997INPT007H.
Texto completo da fonteOussama, Chikhaoui. "Simulations des grandes échelles et modélisations hybrides RANSE/LES pour le calcul d'écoulements turbulents de grande complexité". Nantes, 2007. http://www.theses.fr/2007NANT2143.
Texto completo da fonteThis PhD thesis deals with the Large Eddy Simulation and the hybrid RANSE/LES modelization. The abilityof ISIS-CFD flow solver, originally designed for solving Reynolds-Averaged Navier-Stokes equations, was extended to these different methods by integrating different turbulence models in order to provide a computing code able to deal with a wide range of academic and industrial complex problems. The models and numerical methods chosen and implemented were justified and validated over several academic test cases and comparisons of LES and hybrid simulations using different flow solvers have been successful. Hybrid methods studied showed a good ability to find expected solutions allowing a large saving in mesh resolution. A criterion based on the assessment of the residual of the momentum budget of the averaged flow is also proposed as an indicator of the numerical errors committed in LES and hybrid simulations. Finally, the simulation of complex turbulent flows using hybrid models have been undertaken. The use of RANS/LES models for the computation of the flow around a skidding simplified car model can reproduce faithfully and cheaply, the experimental observations of a turbulent flow of great complexity
Daboussy, David. "Calculs 3D multi-fluides appliqués à la co-injection et à l'injection assistée gaz". Paris, ENSMP, 2000. http://www.theses.fr/2000ENMP1049.
Texto completo da fonteGianfrancesco, Alessandro. "Séchage par atomisation : prorpiétés des collages des particules en relation avec l'agglomération". Paris, AgroParisTech, 2009. http://pastel.paristech.org/5602/01/2009AGPT0029.pdf.
Texto completo da fonteMoet, Henri. "Simulation numérique du comportement des tourbillons de sillage dans l'atmosphère". Toulouse, INPT, 2003. http://www.theses.fr/2003INPT007H.
Texto completo da fonteChetboun, Jonathan. "Conception de formes aérodynamiques en présence d'écoulements décollés : contrôle et optimisation". Phd thesis, Ecole Polytechnique X, 2010. http://pastel.archives-ouvertes.fr/pastel-00550726.
Texto completo da fonteProust, Sébastien. "Ecoulements non-uniformes en lits composés : effets de variations de la largeur du lit majeur". Lyon, INSA, 2005. http://theses.insa-lyon.fr/publication/2005ISAL0083/these.pdf.
Texto completo da fonteFlooding rivers usually present transition reaches where the floodplain width can significantly vary. The present work focuses on both physical and numerical modeling of over bank flows in such configurations. A particular attention is paid to flows in the flood plain. These flows are characterized by turbulent exchanges due to the velocity gradient between flows in the main channel and the floodplain, and by severe mass transfer and associated momentum exchange between the subsections. New experiments are carried out in non-prismatic compound channel flumes: flows in abrupt contraction of the flood plain, enlarging flood plains, flow in the vicinity of groynes. In addition to conventional 1D and 2D simulations, a new modeling is presented: it is called “1D Per Subsection Method” and separates the dynamic equations in each bed. The simulations of both water depth and discharge rate in the floodplain are in good agreement with experimental data, for eight tested geometries
Berton, Gäelle. "Modélisations multi-échelles de structures et d'écoulements engendrés à petits nombres de Reynolds en amont et en aval de milieux poreux fibreux". Poitiers, 1999. http://www.theses.fr/1999POIT2305.
Texto completo da fonteAl, Isber Aziz. "Modélisation de l'écoulement instationnaire décollé à la traversée d'un diaphragme par la méthode des vortex aléatoires". Toulouse, ENSAE, 1991. http://www.theses.fr/1991ESAE0002.
Texto completo da fonteBouhadji, Abdelatif. "Analyse physique par simulation numérique de phénomènes de transition bi- et tridimensionnels dans l'écoulement compressible, visqueux autour d'une aile d'avion". Toulouse, INPT, 1998. http://www.theses.fr/1998INPT026H.
Texto completo da fonteComas-Cardona, Sébastien. "Modélisation, simulation et contrôle du couplage hydro-mécanique pour le moulage de composites". Lille 1, 2005. https://ori-nuxeo.univ-lille1.fr/nuxeo/site/esupversions/3efa882a-9b13-4d8c-9093-5fb158ec4af4.
Texto completo da fonteVial, Christophe. "Apport des méthodes de la mécanique des fluides à l'étude des contacteurs gaz/liquide : expérience et simulation numérique". Vandoeuvre-les-Nancy, INPL, 2000. http://docnum.univ-lorraine.fr/public/INPL_T_2000_VIAL_C.pdf.
Texto completo da fonteThis work is devoted to the experimental study, the modelling and the numerical simulation of the hydrodynamics of a bubble column and an external loop airlift reactor. It includes three parts. First, measuring techniques have been developed to study the local and global hydrodynamic parameters of the bubbly flow. In this section, several treatments of the fluctuating wall pressure signal have been developed to identify the prevailing flow regime; three measuring techniques have been adapted to bubbly flows in order to measure the local velocity of the continuous phase: the "Pavlov" tube, an electrochemical method and Laser Doppler Velocimetry. These techniques have completed the methods already available to study the dynamics of the gas phase: local optical probes and an ultrasound Doppler technique. Then, this set of measuring techniques has been used to characterise the hydrodynamics in both reactors. Regime transitions have been deterrnined. The local and global parameters which have been measured are: the gas hold-up; the velocity, the size and the morphology of the bubbles; the liquid velocity and its local fluctuations. The Reynolds shear stress, which could not be measured, has been estimated using a simple model. The evolution of these parameters has been related to the gas flow rate, the gas distribution and the hydrodynamic regime. The whole experimental data are finally compared to the predicted values obtained using a commercial CFD code. Several models of drag, additional forces and turbulence have been used to obtain the best agreement between calculations and experiments. The ability of this commercial code to predict correctly the flow and its current limits have been highlighted
Zaïdi, Hanane. "Contribution à l’étude des écoulements tourbillonnaires en biomécanique du geste sportif". Reims, 2008. http://theses.univ-reims.fr/exl-doc/GED00000785.pdf.
Texto completo da fonteThis work focusses on the study of the turbulent hydrodynamic flows in human swimming during the submerged phases: after the start dive and after a turn. Its objective is to test various swimming positions in order to improve the hydrodynamic performance of high-level swimmers. More precisely, the study of the effect of the head position on the resisting drag force met by the swimmers was analyzed in a fine way. 2D and 3D CFD simulations were carried out. Three positions of the swimmer’s head were studied in the 2D case: raised, aligned and lowered with respect to the presumed rectilinear axis of the swimmer’s body. The results show that a position of the head "aligned" offers less resistance by comparison to the "raised" and "lowered" positions. The effect of the choice of the turbulence model was also studied in the 2D and 3D cases. The comparison of the numerical results obtained with results from experiments carried out at INSEP with high-level swimmers of the French swimming team shows that the standard k-ω turbulence model is the most appropriate tested model to find the vortical structures which develop around the body of the swimmer and to predict the drag forces opposed by the fluid. This work is supplemented by a thermal study. An experimental part, aiming at drawing a parallel between the energy expenditure of the swimmers and their cutaneous temperature after effort by using infrared thermography, made it possible to show significant variations of the distribution of the cutaneous temperature according to the swimming style. Finally, a numerical study of the influence of the thermal gradient between the water of the pool and the cutaneous temperature of the swimmer shows that a reduction of about 1. 2 % on the total drag force is possible by increasing the temperature of the water from 20°C to 30°C
Mustapha, Hussein. "Simulation numérique de l'écoulement dans des milieux fracturés tridimensionnels". Rennes 1, 2005. http://www.theses.fr/2005REN1S166.
Texto completo da fonte