Cette thèse comporte trois chapitres indépendants, consacrés à l’étude mathématique de trois problèmes physiques distincts, ayant pour modèles trois équations aux dérivées partielles différentes. Ces équations relèvent plus précisément de la méthode des surfaces de niveau, de la théorie de l’écoulement incompressible des matériaux non newtoniens et de la théorie cinétique des gaz raréfiés. Le premier chapitre de la thèse porte sur la dynamique des frontières en mouvement et contient une justification mathématique de la procédure numérique dite de ré-initialisation, dont les applications sont nombreuses dans le contexte de la célèbre méthode des surfaces de niveau. Nous appliquons ces résultats pour une classe d’équations issues de la méthode des surfaces de niveau de premier ordre. Nous écrivons la procédure de ré-initialisation comme un algorithme de décomposition et nous étudions la convergence de l’algorithme en utilisant des techniques d’homogénéisation dans la variable temporelle. Grâce à cette analyse rigoureuse nous introduisons également une nouvelle méthode pour l’approximation de la fonction de distance dans le contexte de la méthode des surfaces de niveau. Dans le cas où l’on cherche seulement une fonction de l’ensemble de niveau avec un gradient minoré proche du niveau zéro, nous proposons une approximation plus simple. Dans le cas général, où le niveau zéro pourrait présenter des changements de topologie, nous introduisons une nouvelle notion de limites relâchées. Dans le deuxième chapitre de la thèse, nous étudions un problème de frontière libre résultant de l’étude de l’écoulement incompressible d’un matériau non-newtonien, avec limite d’élasticité de type Drucker-Prager, sur un plan incliné et sous l’effet de la pesanteur. Nous obtenons une équation sous-différentielle, que nous formulons comme un problème variationnel avec un terme à croissance linéaire de type gradient, et nous étudions le problème dans un domaine non borné. Nous montrons que les équations sont bien posées et satisfont certaines propriétés de régularité. Nous sommes alors capables de relier les paramètres physiques avec le problème abstrait et de prouver des propriétés quantitatives de la solution. En particulier, nous montrons que la solution a un support compact, la limite de ce que nous appelons la frontière libre. Nous construisons également des solutions explicites d’une équation différentielle ordinaire qui peut estimer la frontière libre. Enfin, le troisième et dernier chapitre de la thèse est dédié aux solutions de l’équation de Boltzmann homogène avec molécules maxwelliennes et énergie infinie. Nous obtenons de nouveaux résultats d’existence de solutions éternelles pour cette équation dans un espace de mesures de probabilité d’énergie infinie (i.e. de moment d’ordre deux infini). Elles permettent de décrire le comportement asymptotique en temps d’autres solutions d’énergie infinie, mais elles apparaissent aussi comme des états asymptotiques intermédiaires dans l’étude des solutions d’énergie finie, mais arbitrairement grande. Les méthodes issues de l’analyse harmonique sont utilisées pour étudier l’équation de Boltzmann, où la variable de vitesse est exprimée en Fourier. Enfin, un changement d’échelle logarithmique en la variable temporelle permet de déterminer le bon comportement asymptotique à l’infini des solutions / This thesis consists of three different and independent chapters, concerning the mathematical study of three distinctive physical problems, which are modelled by three non- linear partial differential equations. These equations concern the level set method, the theory of incompressible flow of non-Newtonian materials and the kinetic theory of rare- fied gases. The first chapter of the thesis concerns the dynamics of moving interfaces and contains a rigorous justification of a numerical procedure called re-initialization, for which there are several applications in the context of the level set method. We apply these results for first order level set equations. We write the re-initialization procedure as a splitting algorithm and study the convergence of the algorithm using homogenization techniques in the time variable. As a result of the rigorous analysis, we are also able to introduce a new method for the approximation of the distance function in the context of the level set method. In the case where one only looks for a level set function with gradient bounded from below near the zero level, we propose a simpler approximation. In the general case where the zero level might present changes of topology we introduce a new notion of relaxed limits. In the second chapter of the thesis, we study a free boundary problem arising in the study of the flow of an incompressible non-Newtonian material with Drucker-Prager plasticity on an inclined plane. We derive a subdifferential equation, which we reformulate as a variational problem containing a term with linear growth in the gradient variable, and we study the problem in an unbounded domain. We show that the equations are well posed and satisfy some regularity properties. We are then able to connect the physical parameters with the abstract problem and prove some quantitative properties of the solution. In particular, we show that the solution has compact support and the support is the free boundary. We also construct explicit solutions of an ordinary differential equation, which we use to estimate the free boundary. The last chapter of the thesis is dedicated to the study of infinite energy solutions of the homogeneous Boltzmann equation with Maxwellian molecules. We obtain new results concerning the existence of eternal solutions in the space of probability measure with infinite energy (i.e. the second order moment is infinite). These solutions describe the asymptotic behaviour of other infinite energy solutions but could also be useful in the study of intermediate asymptotic states of solutions with finite but arbitrarily large energy. We use harmonic analysis tools to study the equation, where the velocity variable is expressed in the Fourier space. Finally, a logarithmic scaling of the time variable allows to determine the correct asymptotic scaling of the solutions

En physique du transport, en particulier en physique du transfert radiatif, la méthode de Monte-Carlo a été développée à l'origine comme la simulation de l'histoire d'un grand nombre de particules, dont on déduit des observables moyennes. Cette méthode numérique doit son succès à plusieurs qualités : une gestion naturelle des espaces des phases aux nombreuses dimensions, une erreur systématique nulle par rapport au modèle physico-mathématique, les intervalles de confiance donnés avec les résultats, une capacité à prendre en compte simultanément de nombreux phénomènes physiques, la possibilité de calcul de sensibilités simultané, et une parallélisation aisée. En cinétique des gaz, les particules collisionnent entre elles et non pas avec un milieu extérieur ; on dit que leur transport est non-linéaire. Ces collisions mutuelles mettent en défaut l'approche évoquée ci-dessus de la méthode de Monte-Carlo ; car pour simuler des trajectoires indépendantes de multiples particules et ainsi estimer leur distribution, il faut connaître au préalable exactement cette même distribution...Cette thèse fait suite à celles de Jérémi DAUCHET (2012) et de Mathieu GALTIER (2014), consacrées au transfert radiatif. Entre autres travaux, ces auteurs montraient comment la méthode de Monte-Carlo peut s'accommoder de non-linéarités, en gardant son formalisme et ses spécificités habituelles. Les non-linéarités alors franchies étaient respectivement une loi de couplage chimie/luminance, et la dépendance de la luminance envers le coefficient d'absorption. On essaie dans ce manuscrit d'outrepasser la non-linéarité du transport. Pour cela, nos principaux outils sont un suivi des particules en remontant le temps, basé sur des formulations intégrales des équations de transport, formulations largement inspirées des algorithmes dits à collisions nulles. Nous montrons, sur plusieurs exemples académiques, que nous avons en effet étendu la méthode de Monte-Carlo à la résolution de l'équation de Boltzmann. Ces exemples sont aussi l'occasion de tester les limites de ce que nous avons mis en place. Les résultats les plus marquants sont certainement l'absence totale de maillage dans la méthode numérique, ainsi que sa capacité à calculer correctement les quantités de particules de haute énergie cinétique (toujours peu nombreuses par rapport au total, en cinétique des gaz). Au-delà des exemples fournis, ce manuscrit est voulu comme un essai de formalisme et une exploration des bases de la méthode développée. L'accent est mis sur les raisonnements menant à la mise au point de la méthode, plutôt que sur les implémentations particulières qui ont été abouties. La méthode est encore, aux yeux de l'auteur, largement susceptible d'être retravaillée. En particulier, les temps maximaux sur lesquels l'évolution des particules est calculable, qui constituent la faiblesse principale de la méthode numérique développée, peuvent sûrement être augmentés. / In transport physics, especially in radiative transfer physics, the Monte-Carlo method has been originally developed as the simulation of the history of numerous particles, from which are deduced mean observables. This numerical method owes its success to several qualities : a natural management of many-dimensional phase space, a null systematic error away from the mathematical and physical model, the confidence intervals given with the results, an ability to take into account simultaneously numerous physical phenomenons, the simultaneous sensitivities calculating possibility, and an easy parallelization. In gas kinetics, particles collide each other, not with an external fixed medium ; it is said that their transport is non-linear. These mutual collisions put out of action the aforesaid approach of the Monte-Carlo method ; because in order to simulate the independent trajectories of multiple particles and thus estimate their distribution, this distribution must beforehand be exactly known...This thesis follows on from those of Jérémy DAUCHET (2012) and of Mathieu GALTIER (2014), dedicated to radiative transfer physics. Between other works, these authors have shown how the Monte-Carlo method can bear non-linearities, while keeping its customary formalism and specificities. The then overcome non-linearities were respectively a chemistry/irradiance coupling law, and the dependence of the irradiance toward the absorption coefficient. We try in this manuscript to overcome the non-linearity of the transport. In this aim, our main tools are a reverse following of particles, based on integral formulations of the transport equations, formulations largely inspired from the so-called null collisions algorithms. We show, on several academic examples, that we have indeed extended the Monte Carlo method to the resolution of the Boltzmann equation. These examples are also occasions to test the limits of what we have built. The most noteworthy results are certainly the absence of any mesh in the numerical method, and its capacity to calculate correctly the high-speed particles quantities (always rare compared to the total, in gas kinetics). Beyond the given examples, this manuscript is wanted as a formalism attempt and an exploration of the developed method basics. The focus is made on the reasoning leading to the method, rather than on particular implementations which have been realized. In the eyes of the author, the method is still largely reworkable. In particular, the maximal times on which the evolution of particles is computable, which constitute the main weakness of the developed numerical method, can surely be increased.

No description available.

No description available.

Dans des milieux tels que les gaz, les plasmas et les milieux granulaires, un objet se déplaçant à des vitessessupersoniques, compresse et chauffe le fluide devant lui, formant ainsi une onde de choc. La zone hors-équilibreappelée front d’onde, où ont lieu de brusques variations de température, pression et densité, présente unestructure particulière, avec notamment des distributions des vitesses des particules fortement non-gaussienneset difficiles à visualiser. Dans une avancée importante en 1951, Mott-Smith décrit le front d’onde comme lasuperposition des deux états que sont le gaz supersonique initial et le gaz subsonique compressé et chauffé,impliquant ainsi l’existence de distributions des vitesses bimodales. Des expériences à grands nombres de Machont confirmé cette structure globalement bimodale. Ce modèle n’explique cependant pas la présence d’un surplusde particules à des vitesses intermédiaires, entre le gaz supersonique et le gaz subsonique.Ce travail de thèse porte sur l’étude des ondes de choc dans les gaz granulaires, où les particules interagissentuniquement par des collisions binaires inélastiques. Dans ces gaz dissipatifs, la température granulaire, traduisantl’agitation des particules, permet de définir l’équivalent d’une vitesse du son par analogie aux gaz moléculaires.Les basses valeurs de ces vitesses du son dans les gaz granulaires, permettent de générer facilement des ondes dechoc dans lesquelles chaque particule peut être suivie, contrairement aux gaz moléculaires. La première partie decette étude porte sur l’effet de la dissipation d’énergie, due aux collisions inélastiques, sur la structure des ondesde choc dans les gaz granulaires. Les modifications induites sur la température, la densité et la vitesse moyennemesurées, sont interprétées à l’aide d’un modèle basé sur l’hypothèse bimodale de Mott-Smith et intégrant ladissipation d’énergie. La deuxième partie est consacrée à l’interprétation des distributions des vitesses dans lefront d’onde. À partir des expériences réalisées dans les gaz granulaires, une description trimodale, incluant unétat intermédiaire supplémentaire, est proposée et étendue avec succès aux distributions des vitesses dans lesgaz moléculaires. / In different materials such as gases, plasmas and granular material, an object, moving at supersonic speed,compresses and heats the fluid ahead. The shock front is the out-of-equilibrium area, where violent changesin temperature, pressure and density occur. It has a particular structure with notably strongly non-Gaussianparticle velocity distributions, which are difficult to observe. In an important breakthrough in 1951, Mott-Smithdescribes the shock front as a superposition of two states: the initial supersonic gas and the compressed andheated subsonic gas, implying existence of bimodal velocity distributions. Several experiences at high Machnumbers show this overall bimodal structure. However this model does not explain the existence of a surplusof particles with intermediate velocities, between the supersonic and the subsonic gas.This thesis focuses on shock waves in granular gases, where particles undergo only inelastic binary collisions.In these dissipative gases, the granular temperature, reflecting the particle random motion, allows to definethe equivalent to the speed of sound by analogy with molecular gases. The low values of this speed of soundpermit to generate easily shock waves in which each particle can be tracked, unlike molecular gases. The firstpart of this work focuses on the effect of the energy dissipation, due to inelastic collisions, on the shock frontstructure in granular gases. Modifications induced on temperature, density and mean velocity, are captured bya model based on the bimodal hypothesis of Mott-Smith and including energy dissipation. The second part isdevoted to the study of velocity distributions in the shock front. From experiences in granular gases, a trimodaldescription, including an additional intermediate state, is proposed and successfully extended to the velocitydistributions in molecular gases.

Desde os trabalhos de Clausius, Boltzmann e Gibbs, sabe-se que partículas que interagem através de potenciais de curto alcance alcançam, após um processo de relaxação, o estado final estacionário que corresponde ao equilíbrio termodinâmico [I]. Embora nenhuma prova exata exista para isso, na prática, verifica-se que os sistemas não-integráveis com uma energia fixa e um número finito de partículas (ensemble microcanônico, por exemplo) sempre relaxam para um estado estacionário que só depende de quantidades globais conservadas pela dinâmica: energia, momentum e momentum angular. Este estado estacionário corresponde ao estado de equilíbrio termodinâmico e não depende das especificidades da distribuição inicial de partículas. Este cenário muda drasticamente quando a interação entre as partículas passa a ser de longo alcance [2]. A descrição estatística e termodinâmica desses sistemas ainda é objeto de estudo. Contudo, o que se sabe é que esses sistemas têm como propriedade fundamental o fato de que, no limite termodinâmico o tempo de colisão diverge e o equilíbrio termodinâmico nunca é atingido [3]. Nesse trabalho analisamos do ponto de vista teórico e por simulação de dinâmica molecular o estado estacionário atingido por sistemas auto-gravitantes em uma, duas e três dimensões e plasmas não-neutros na dinâmica de um feixe de partículas carregadas. Analisamos ainda um modelo com transição de fases para o estado fora do equilíbrio (HMF). Em todos os casos a teoria proposta na tese mostrou-se consistente com os simulações numéricas empregadas. / Since the work of Clausius, Boltzmann and Gibbs, it is known that particles interacting by a short-range potential, after a relaxation process, reach a final stationary state that corresponds to thermodynamic equilibrium. Although no exact proof exists, in practice non-integrable systems with fixed energy and a finite number of particles (i.e., microcanonical ensemble) always relax to a stationary state that depends only on global quantities conserved by the dynamics: energy, momentum and angular momentum. This stationary state corresponds to the state of thermodynamic equilibrium and does not depend on the specifics of the initial particle distribution. This scenario changes drastically when the interaction between particles is longranged [2] The statistical and thermodynamic description of these systems is still an object of study. However, a fundamental property of these systems is the fact that, in the thermodynamic limit, the collision time diverges and thermodynamic equilibrium is never achieved [3].. In this thesis we analyse, from a theoretical point of view and using molecular dynamics simulations, the stationary state achieved by self-gravitating systems in one, two and three dimensions and non-neutral plasmas in the dynamics of charged particle beams. We also analyse a model with out-of-equilibrium phase transitions (HMF). In all these cases, the theory proposed in this thesis is shown to be consistent with the numerical simulations applied.

Desde os trabalhos de Clausius, Boltzmann e Gibbs, sabe-se que partículas que interagem através de potenciais de curto alcance alcançam, após um processo de relaxação, o estado final estacionário que corresponde ao equilíbrio termodinâmico [I]. Embora nenhuma prova exata exista para isso, na prática, verifica-se que os sistemas não-integráveis com uma energia fixa e um número finito de partículas (ensemble microcanônico, por exemplo) sempre relaxam para um estado estacionário que só depende de quantidades globais conservadas pela dinâmica: energia, momentum e momentum angular. Este estado estacionário corresponde ao estado de equilíbrio termodinâmico e não depende das especificidades da distribuição inicial de partículas. Este cenário muda drasticamente quando a interação entre as partículas passa a ser de longo alcance [2]. A descrição estatística e termodinâmica desses sistemas ainda é objeto de estudo. Contudo, o que se sabe é que esses sistemas têm como propriedade fundamental o fato de que, no limite termodinâmico o tempo de colisão diverge e o equilíbrio termodinâmico nunca é atingido [3]. Nesse trabalho analisamos do ponto de vista teórico e por simulação de dinâmica molecular o estado estacionário atingido por sistemas auto-gravitantes em uma, duas e três dimensões e plasmas não-neutros na dinâmica de um feixe de partículas carregadas. Analisamos ainda um modelo com transição de fases para o estado fora do equilíbrio (HMF). Em todos os casos a teoria proposta na tese mostrou-se consistente com os simulações numéricas empregadas. / Since the work of Clausius, Boltzmann and Gibbs, it is known that particles interacting by a short-range potential, after a relaxation process, reach a final stationary state that corresponds to thermodynamic equilibrium. Although no exact proof exists, in practice non-integrable systems with fixed energy and a finite number of particles (i.e., microcanonical ensemble) always relax to a stationary state that depends only on global quantities conserved by the dynamics: energy, momentum and angular momentum. This stationary state corresponds to the state of thermodynamic equilibrium and does not depend on the specifics of the initial particle distribution. This scenario changes drastically when the interaction between particles is longranged [2] The statistical and thermodynamic description of these systems is still an object of study. However, a fundamental property of these systems is the fact that, in the thermodynamic limit, the collision time diverges and thermodynamic equilibrium is never achieved [3].. In this thesis we analyse, from a theoretical point of view and using molecular dynamics simulations, the stationary state achieved by self-gravitating systems in one, two and three dimensions and non-neutral plasmas in the dynamics of charged particle beams. We also analyse a model with out-of-equilibrium phase transitions (HMF). In all these cases, the theory proposed in this thesis is shown to be consistent with the numerical simulations applied.

Desde os trabalhos de Clausius, Boltzmann e Gibbs, sabe-se que partículas que interagem através de potenciais de curto alcance alcançam, após um processo de relaxação, o estado final estacionário que corresponde ao equilíbrio termodinâmico [I]. Embora nenhuma prova exata exista para isso, na prática, verifica-se que os sistemas não-integráveis com uma energia fixa e um número finito de partículas (ensemble microcanônico, por exemplo) sempre relaxam para um estado estacionário que só depende de quantidades globais conservadas pela dinâmica: energia, momentum e momentum angular. Este estado estacionário corresponde ao estado de equilíbrio termodinâmico e não depende das especificidades da distribuição inicial de partículas. Este cenário muda drasticamente quando a interação entre as partículas passa a ser de longo alcance [2]. A descrição estatística e termodinâmica desses sistemas ainda é objeto de estudo. Contudo, o que se sabe é que esses sistemas têm como propriedade fundamental o fato de que, no limite termodinâmico o tempo de colisão diverge e o equilíbrio termodinâmico nunca é atingido [3]. Nesse trabalho analisamos do ponto de vista teórico e por simulação de dinâmica molecular o estado estacionário atingido por sistemas auto-gravitantes em uma, duas e três dimensões e plasmas não-neutros na dinâmica de um feixe de partículas carregadas. Analisamos ainda um modelo com transição de fases para o estado fora do equilíbrio (HMF). Em todos os casos a teoria proposta na tese mostrou-se consistente com os simulações numéricas empregadas. / Since the work of Clausius, Boltzmann and Gibbs, it is known that particles interacting by a short-range potential, after a relaxation process, reach a final stationary state that corresponds to thermodynamic equilibrium. Although no exact proof exists, in practice non-integrable systems with fixed energy and a finite number of particles (i.e., microcanonical ensemble) always relax to a stationary state that depends only on global quantities conserved by the dynamics: energy, momentum and angular momentum. This stationary state corresponds to the state of thermodynamic equilibrium and does not depend on the specifics of the initial particle distribution. This scenario changes drastically when the interaction between particles is longranged [2] The statistical and thermodynamic description of these systems is still an object of study. However, a fundamental property of these systems is the fact that, in the thermodynamic limit, the collision time diverges and thermodynamic equilibrium is never achieved [3].. In this thesis we analyse, from a theoretical point of view and using molecular dynamics simulations, the stationary state achieved by self-gravitating systems in one, two and three dimensions and non-neutral plasmas in the dynamics of charged particle beams. We also analyse a model with out-of-equilibrium phase transitions (HMF). In all these cases, the theory proposed in this thesis is shown to be consistent with the numerical simulations applied.

Lattice Boltzmann Method has been quite successful for incompressible flows. Its extension for compressible (especially supersonic and hypersonic) flows has attracted lot of attention in recent time. There have been some successful attempts but nearly all of them have either resulted in complex or expensive equilibrium function distributions or in extra energy levels. Thus, an efficient Lattice Boltzmann Method for compressible fluid flows is still a research idea worth pursuing for. In this thesis, a new Lattice Boltzmann Method has been developed for compressible flows, by using the concept of a relaxation system, which is traditionally used as semilinear alternative for non-linear hypebolic systems in CFD. In the relaxation system originally introduced by Jin and Xin (1995), the non-linear flux in a hyperbolic conservation law is replaced by a new variable, together with a relaxation equation for this new variable augmented by a relaxation term in which it relaxes to the original nonlinear flux, in the limit of a vanishing relaxation parameter. The advantage is that instead of one non-linear hyperbolic equation, two linear hyperbolic equations need to be solved, together with a non-linear relaxation term. Based on the interpretation of Natalini (1998) of a relaxation system as a discrete velocity Boltzmann equation, with a new isotropic relaxation system as the basic building block, a Lattice Boltzmann Method is introduced for solving the equations of inviscid compressible flows. Since the associated equilibrium distribution functions of the relaxation system are not based on a low Mach number expansion, this method is not restricted to the incompressible limit. Free slip boundary condition is introduced with this new relaxation system based Lattice Boltzmann method framework. The same scheme is then extended for curved boundaries using the ghost cell method. This new Lattice Boltzmann Relaxation Scheme is successfully tested on various bench-mark test cases for solving the equations of compressible flows such as shock tube problem in 1-D and in 2-D the test cases involving supersonic flow over a forward-facing step, supersonic oblique shock reflection from a flat plate, supersonic and hypersonic flows past half-cylinder.

Lattice-Boltzmann method(LBM), a promising new particle-based simulation technique for complex and multiscale fluid flows, has seen tremendous adoption in recent years in computational fluid dynamics. Even with a state-of-the-art LBM solver such as Palabos, a user still has to manually write his program using the library-supplied primitives. We propose an automated code generator for a class of LBM computations with the objective to achieve high performance on modern architectures. Tiling is a very important loop transformation used to improve the performance of stencil computations by exploiting locality and parallelism. In the first part of the work, we explore diamond tiling, a new tiling technique to exploit the inherent ability of most stencils to allow tile-wise concurrent start. This enables perfect load-balance during execution and reduces the frequency of synchronization required. Few studies have looked at time tiling for LBM codes. We exploit a key similarity between stencils and LBM to enable polyhedral optimizations and in turn time tiling for LBM. Besides polyhedral transformations, we also describe a number of other complementary transformations and post processing necessary to obtain good parallel and SIMD performance on modern architectures. We also characterize the performance of LBM with the Roofline performance model. Experimental results for standard LBM simulations like Lid Driven Cavity, Flow Past Cylinder, and Poiseuille Flow show that our scheme consistently outperforms Palabos–on average by3 x while running on 16 cores of a n Intel Xeon Sandy bridge system. We also obtain a very significant improvement of 2.47 x over the native production compiler on the SPECLBM benchmark.