Neste trabalho é desenvolvida uma metodologia baseada na aplicação do método dos momentos conjugado à aproximação por diferenças finitas, para análise no domínio do tempo da distribuição de correntes transitórias em meios condutores. O procedimento computacional permite que sejam considerados sistemas de condutores longos, com seção transversal genérica, discretizados em elementos filiformes. Assim, a equação integral que descreve o problema é substituída por uma representação matricial. No caso de meios não homogêneos, o método dos elementos finitos é usado para a obtenção da matriz que permite a solução do problema. Como o método proposto baseia-se na solução passo a passo no tempo, é interessante, para maior eficiência computacional, que as matrizes tenham sua dimensão reduzida, o que pode ser realizado por meio da Transformada Discreta Wavelet, como investigado no trabalho. Também é feita uma análise da estabilidade do procedimento computacional, para verificação das condições de aplicabilidade do mesmo. A validação do procedimento desenvolvido é feita a partir da comparação dos resultados obtidos para problemas com solução já conhecida por meio de medições ou por outros métodos de resolução. São apresentados resultados da aplicação do método proposto a alguns casos de interesse prático, na área de Compatibilidade Eletromagnética, destacando-se a análise do efeito de blindagem e o estudo de crosstalk em um sistema multicondutor. O procedimento desenvolvido ainda pode ser aprimorado com a utilização de elementos com formas mais gerais do que a filiforme proposta, permitindo sua extensão a problemas tridimensionais. Também pode ser estudado o pré-condicionamento da matriz a ser reduzida pela aplicação da Transformada Wavelet, a fim de permitir um melhor desempenho do processo computacional. / In this work, a methodology is developed, based on the application of the moment method associated to a finite difference approximation, for time domain analysis of transient current distribution in conducting media. The computational procedure allows that long conductors with any cross section can be considered, approximated by filamentary elements. Thus, the integral equation that describes the problem is substituted by a matrix representation. In the case of non homogeneous regions, the finite element method is used to obtain the matrix that allows the solution of the problem. As the proposed method works in a time step by step scheme, it is interesting, for increased computational efficiency, to reduce the size of the involved matrices. This can be done by application of the Discrete Wavelet Transform, as investigated in this work. An analysis of the numerical stability of the computational procedure is also done, aiming to establish its applicability conditions. Validation of the numerical procedure developed is done by comparing the results obtained for problems whose solution is known by measurement or by another method. Some problems of practical interest in Electromagnetic Compatibility were solved and the results are presented, concerning particularly shielding effects and multi conductor crosstalk. The developed procedure can be improved employing elements with more general shapes, besides the filamentary ones, allowing the consideration of three dimensional systems. Also, pre conditioning of the matrices before application of the Wavelet Transform can be studied, for a better performance of the method.

Cette thèse est consacrée à l'étude de la contrôlabilité approchée et à zéro des systèmes paraboliques linéaires sur un domaine non vide borné Ω de (), contrôlés par moins de forces que d'équations. Les contrôles seront localisés sur un ouvert de Ω ou sur son bord. Nous étudierons deux problèmes différents. Le premier consiste à contrôler une des équations indirectement à l'aide d'un opérateur de couplage d'ordre un. Nous obtenons alors des résultats pour plusieurs classes d'opérateurs et de systèmes. La deuxième question que nous étudierons est de savoir s'il est possible de contrôler seulement certaines composantes de la solution du système. Nous donnons une condition nécessaire et suffisante lorsque les coefficients de couplage sont constants ou dépendent du temps et étudions un système simplifié quand ils dépendent de l'espace. Nous terminerons en détaillant un schéma numérique avec lequel nous fournirons des perspectives quant à quelques problèmes qui restent ouverts en contrôlabilité partielle des systèmes paraboliques linéaires. / This thesis is devoted to the study of the approximate and null controllability of linear parabolic systems on a nonempty bounded domain Ω of(), controlled by less controls than equations. The controls will be localized in an open set of Ω or on its boundary. We will study two different problems. The first of them involves controlling one of the equations indirectly with a coupling operator of order one. We obtain some results for different class of operators and systems. The second question we will study is to know if it is possible to control only some components of the solution of the system. We give a necessary and sufficient condition when the coupling coefficients are constant or time depending and study a simplified system when they are space dependent. We will finish by giving details on a numerical scheme with which we provide perspectives concerning some open problems in partial controllability of linear parabolic systems.

Neste trabalho é desenvolvida uma metodologia baseada na aplicação do método dos momentos conjugado à aproximação por diferenças finitas, para análise no domínio do tempo da distribuição de correntes transitórias em meios condutores. O procedimento computacional permite que sejam considerados sistemas de condutores longos, com seção transversal genérica, discretizados em elementos filiformes. Assim, a equação integral que descreve o problema é substituída por uma representação matricial. No caso de meios não homogêneos, o método dos elementos finitos é usado para a obtenção da matriz que permite a solução do problema. Como o método proposto baseia-se na solução passo a passo no tempo, é interessante, para maior eficiência computacional, que as matrizes tenham sua dimensão reduzida, o que pode ser realizado por meio da Transformada Discreta Wavelet, como investigado no trabalho. Também é feita uma análise da estabilidade do procedimento computacional, para verificação das condições de aplicabilidade do mesmo. A validação do procedimento desenvolvido é feita a partir da comparação dos resultados obtidos para problemas com solução já conhecida por meio de medições ou por outros métodos de resolução. São apresentados resultados da aplicação do método proposto a alguns casos de interesse prático, na área de Compatibilidade Eletromagnética, destacando-se a análise do efeito de blindagem e o estudo de crosstalk em um sistema multicondutor. O procedimento desenvolvido ainda pode ser aprimorado com a utilização de elementos com formas mais gerais do que a filiforme proposta, permitindo sua extensão a problemas tridimensionais. Também pode ser estudado o pré-condicionamento da matriz a ser reduzida pela aplicação da Transformada Wavelet, a fim de permitir um melhor desempenho do processo computacional. / In this work, a methodology is developed, based on the application of the moment method associated to a finite difference approximation, for time domain analysis of transient current distribution in conducting media. The computational procedure allows that long conductors with any cross section can be considered, approximated by filamentary elements. Thus, the integral equation that describes the problem is substituted by a matrix representation. In the case of non homogeneous regions, the finite element method is used to obtain the matrix that allows the solution of the problem. As the proposed method works in a time step by step scheme, it is interesting, for increased computational efficiency, to reduce the size of the involved matrices. This can be done by application of the Discrete Wavelet Transform, as investigated in this work. An analysis of the numerical stability of the computational procedure is also done, aiming to establish its applicability conditions. Validation of the numerical procedure developed is done by comparing the results obtained for problems whose solution is known by measurement or by another method. Some problems of practical interest in Electromagnetic Compatibility were solved and the results are presented, concerning particularly shielding effects and multi conductor crosstalk. The developed procedure can be improved employing elements with more general shapes, besides the filamentary ones, allowing the consideration of three dimensional systems. Also, pre conditioning of the matrices before application of the Wavelet Transform can be studied, for a better performance of the method.

In this dissertation, a group of vehicle dynamics simulation tools is developed with two primary goals: to accurately represent vehicle behavior and to provide insight that improves the understanding of vehicle performance. Three tools are developed that focus on tire modeling, vehicle modeling and lap time simulation. Tire modeling is based on Nondimensional Tire Theory, which is extended to provide a flexible model structure that allows arbitrary inputs to be included. For example, rim width is incorporated as a continuous variable in addition to vertical load, inclination angle and inflation pressure. Model order is determined statistically and only significant effects are included. The fitting process is shown to provide satisfactory fits while fit parameters clearly demonstrate characteristic behavior of the tire. To represent the behavior of a complete vehicle, a Nondimensional Tire Model is used, along with a three degree of freedom vehicle model, to create Milliken Moment Diagrams (MMD) at different speeds, longitudinal accelerations, and under various yaw rate conditions. In addition to the normal utility of MMDs for understanding vehicle performance, they are used to develop Limit Acceleration Surfaces that represent the longitudinal, lateral and yaw acceleration limits of the vehicle. Quasi-transient lap time simulation is developed that simulates the performance of a vehicle on a predetermined path based on the Limit Acceleration Surfaces described above. The method improves on the quasi-static simulation method by representing yaw dynamics and indicating the vehicle's stability and controllability over the lap. These improvements are accomplished while maintaining the simplicity and computational efficiency of the two degree of freedom method. / Graduation date: 2013

We prove exponential decay for a fractional power of the Green's function for some correlated Anderson models using the fractional moment method.

Le développement des bioraffineries repose sur une conception optimisée d’installations industrielles en synergie comportant un grand nombre de flux de matière et d’opérations unitaires. Le recours à des simulateurs de procédés présente un intérêt certain dans la conception, l’analyse et l’optimisation de tels procédés. Souhaitant initier le développement d’un outil de simulation adapté à ce secteur, nous nous sommes appuyés sur l’exemple d’un procédé d’oxydation de biopolymères par extrusion réactive. Les procédés d’extrusion réactive sont caractérisés par un couplage intime entre écoulement, thermique et cinétiques réactionnelles. Les modalités de ce couplage dépendent des réactions visées. Souhaitant proposer un modèle flexible, intégrable dans un simulateur statique de procédés, et permettant d’atteindre un bon compromis entre la prédictivité et la quantité d’essais nécessaires pour ajuster les paramètres du modèle, nous avons opté pour une approche de modélisation mixte reposant à la fois sur une représentation de l’écoulement à l’aide de réacteurs idéaux et sur des lois de la mécanique des fluides. L’écoulement est modélisé par une cascade de réacteurs continus parfaitement agités (RCPA) avec reflux. Chaque RCPA est caractérisé par un taux de remplissage qui dépend des conditions opératoires. Le calcul du taux de remplissage des RCPA, de la pression matière et des débits circulant entre les RCPA en régime permanent est effectué en réalisant un bilan matière sur chaque RCPA. La température matière dans chacun des RCPA est calculée grâce à un bilan thermique. La modification chimique du matériau est décrite à l’aide de trois réactions : l’oxydation dépolymérisante, la formation de groupements fonctionnels (carbonyles et carboxyles) et la dégradation thermomécanique du biopolymère sous l’effet de la chaleur et des contraintes de cisaillement. L’établissement des équations de bilan de population auxquelles on applique la méthode des moments, permet de calculer simultanément les masses molaires moyennes en nombre et en poids du polymère ainsi que la teneur en agent oxydant dans chacun des RCPA. La viscosité est reliée à masse molaire moyenne. Un algorithme de calcul itératif permet de coupler le bilan matière, le bilan thermique et le calcul réactionnel. Les données expérimentales nécessaires à la validation du modèle ont été fournies par la plate-forme expérimentale mise au point au CVG (Centre de Valorisation des Glucides, Amiens) dans le cadre du programme Synthons. Une méthode d’ajustement des paramètres du modèle à partir d’un nombre minimal de données expérimentales a été proposée, permettant d’évaluer le caractère prédictif du modèle. Le modèle d’extrusion réactive ainsi ajusté a permis de reproduire les résultats expérimentaux obtenus pour différents matériaux, débits, vitesses de rotation, et sur deux extrudeuses detaille et de configuration différentes. L’intégration du modèle d’extrusion réactive dans un simulateur de procédés - le logiciel USIM PAC - a permis de simplifier sa mise en œuvre,offre des perspectives en optimisation et dimensionnement d’équipement et rend possible la simulation de l’opération d’extrusion réactive au sein d’une chaîne de transformation complète. / The development of biorefineries requires integrating and optimizing plants and handling a large number of material flows and unit operations. The development of a process simulator dedicated to this field would thus be of great interest. This is what we intended to initiate by relying on the example of the oxidation of biopolymers by reactive extrusion. Reactive extrusion is characterized by a strong coupling between flow, heat transfer and reaction kinetics. This coupling depends on the desired reactions. We here intended to elaborate aflexible model, being easily integrated into a static process simulator, and enabling to reach agood compromise between the predictive character of the model and the amount of experiments required to adjust model parameters. Therefore, we adopted a hybrid modelling approach combining a flow description based on ideal reactors and continuum mechanics laws. Flow is modeled as a cascade of continuous stirred tank reactors (CSTR) with possible backflow. Flow rates between CSTRs are calculated using physical laws taking into account the operating conditions and geometric parameters of the equipment. Each CSTR is characterized by a filling ratio, which depends on the operating conditions. The calculation of steady-state filling ratio, pressure and flow rates between the CSTRs is achieved by performing a material balance in each CSTR. Material temperature in each CSTR is calculated through a thermal balance. The chemical modification of the material is described using three reactions: the oxidative depolymerization, the formation of functional groups(carbonyl and carboxyl) and the thermomechanical degradation of the biopolymer induced by heating and shearing. The number-averaged and weight-averaged molecular weight of the biopolymer and the oxidant content in each CSTR are computed simultaneously by applying the moment operation to population balance equations. Viscosity is linked to the mean molecular weight. An iterative algorithm enables to couple material balance, thermal balance and reaction kinetics. The experimental data required for model validation were provided by the experimental platform developed at the CVG (Centre de Valorisation des Glucides,Amiens, France) in the frame of the Synthons program. A method was proposed in order to adjust model parameters with a minimal number of experimental data, enabling to assess the predictive character of the model. Once the parameters were adjusted, the reactive extrusion model enabled to reproduce the experimental results obtained with different raw materials,flow rates, screw rotation speeds, and using two extruders with different size and screw configuration. The integration of the reactive extrusion model into a process simulator - the USIM PAC software - enabled to simplify its implementation. This constitutes a promising step in a perspective of process optimization and scale-up, and enables to simulate a reactive extrusion operation within a global plant simulator.

Ce manuscrit est consacré à l'amélioration des deux méthodes numériques : MMDF et Méthode des Moments avec la technique de la RSA. Dans un premier temps, on présente les outils théoriques sur l'étude du problème de la diffraction par des réseaux lamellaires, éclairée sous incidence classique et les outils de simulation pour résoudre les équations de Maxwell. On choisit comme fonctions de base et de test, les fonctions triangles. On introduit par la suite, la technique de la RSA afin d'améliorer la vitesse de convergence de calcul. Les résultats obtenus sont comparés à ceux de la méthode MMFE et la méthode MMDF de "Lalanne et al". Une étude numérique de la stabilité et de la convergence de la méthode est effectuée. Enfin, on présente une extension de la Méthode des Moments basée sur l'hypothèse de Galerkin au cas de l'incidence conique. Le détail de la résolution des équations Maxwell est décrit. L'application numérique est traitée dans le cas du réseau diélectrique pour mettre au point les influences de l'état de la polarisation et les paramètres physiques. / This manuscript is devoted to improvement of the two numerical methods : MMDF and Method of the Moments with the technique of the RSA. In the first part, we present the theoretical tools on study of the problem of diffraction by binary grating,illuminated of classical incidence and the tools for simulation to solve the Maxwell's equations. We choose like basic functions and test, the functions triangles. In the second part, we introduce the technique of the RSA so to improve the speed of convergence of calculation. The results obtained are compared with those of method MMFE and method MMDF of"Lalanne and al". A numerical study of the stability and convergence of the method are carried. The last part, we present the Method of the Moments with triangle functions as expansion and triangle as test functions to the case of conical incidence. We describe the detail of the resolution of the Maxwell equations. Numerical application is treated in the case of the dielectric grating to develop the influences of state of polarization and the physical parameters.

Le travail effectué au cours de cette thèse s'inscrit au sein du projet ANR Mascaret, dont l'objectif est la compréhension du phénomène de mascaret, l'étude de ses conséquences sur l'environnement et sa sensibilité aux modifications de cet environnement. La contribution de cette thèse s'inscrit uniquement dans la partie numérique de ce projet. Seul l'aspect transport sédimentaire causé par le mascaret sera abordé. Le but est de construire un modèle numérique de transport sédimentaire général qui pourra notamment s'appliquer au cas du mascaret. Trois méthodes numériques sont explorées, une première permettant le suivi individuel des grains sédimentaires et deux autres permettant de suivre l'évolution de la concentration en grains au sein de l'écoulement. La première méthode considérera les plus petites échelles et sera appelée méthode tracker et consistera en un suivi individuel des grains sédimentaires. La seconde méthode, dite méthode particulaire, portera sur des échelles plus larges et le transport d'une concentration locale en grains sédimentaires. Enfin, la troisième méthode, que l'on appellera méthode des moments, s'intéressera aux échelles les plus larges en transportant un nuage de particules sédimentaires dans son ensemble grâce à une seule particule numérique caractérisée par les moments de sa distribution en concentration interne. Ceci permettra de caractériser le transport sédimentaire de manière locale qui se produit lors du passage d'un mascaret. Deux mascarets ondulés de nombre de Froude proches seront étudiés. Il sera notamment montré que le nombre de Froude n'est pas un critère permettant de caractériser le transport sédimentaire induit par les mascarets. / The work performed during this thesis is a part of the Mascaret ANR project, which aims to understand the phenomenon of tidal bore, the study of its impact on the environment and its sensitivity to changes in that environment. The contribution of this thesis lies solely in the numerical part of this project. Only the sediment transport caused by the tidal bore is discussed. The goal is to build a generic numerical model of sediment transport which can therefore be applied to the specific case of tidal bores. Three methods are explored, a first for individual tracking of sediment grains and two to model the concentration of grains in the flow. The first method considers the smallest scales and will be called tracking method and consists of individual tracking of sediment grains. The second method, called particle method, focuses on larger scales and the transport of local concentration of sedimentary grains. The third method, which we call moments method, will focus on the largest scales, carrying a cloud of sediment grains as a whole using a single numerical particle characterized by the moments of its internal concentration distribution. This will characterize the local sediment transport process occurring during the passage of a tidal bore. Two undulating bores will be studied whose Froude numbers are close. It will be shown in particular that the Froude number is not a criterion to deduce the intensity of the induced tidal bores sediment transport.

[pt] Neste trabalho é apresentada a análise e síntese de redes refletoras para aplicações espaciais. Redes refletoras são abordadas para expor de forma sucinta suas aplicações, geometrias, métodos de análise e de síntese de diagramas de irradiação. O método do circuito equivalente é utilizado para a formulação da função diádica de Green, possibilitando estabelecer relações entre os campos eletromagnéticos e as densidades superficiais de corrente elétrica em meios estratificados e constituídos por diversas metalizações. Validações numéricas são realizadas através da determinação das constantes de propagação em estruturas slab e guia de onda parcialmente preenchido, formadas por paredes metálicas com condutividade infinita. Análises de redes refletoras com poucos elementos são realizadas empregando-se o método dos momentos para calcular numericamente as densidades superficiais de corrente elétrica que fluem sobre os patches. Adicionalmente, propõe-se a aplicação de uma função base de domínio completo com condição de borda segmentada para modelar o comportamento impulsivo da densidade de corrente nas bordas dos espalhadores. Diferentes curvas de fase são obtidas através da técnica de variação das dimensões físicas dos espalhadores metálicos, considerando o campo elétrico total composto pelos campos espalhado, refletido e difratado, devido às posições espaciais dos patches. Softwares comerciais são utilizados nas verificações dos resultados obtidos. Por fim, o conceito de defasagem progressiva e o método de enxame de partículas foram aplicados para se determinar a fase desejada em cada célula. Assim, as curvas de fase são interpoladas e utilizadas para projetar as dimensões dos elementos impressos e garantir a distribuição de fase calculada sobre a superfície da rede refletora, de forma a reproduzir os diagramas desejados. / [en] This thesis presents the analysis and synthesis of reflectarrays for space applications. A review of the state-of-the-art of reflectarrays is presented and the applications, geometries, methods of analysis and methods of synthesis are discussed. Dyadic Green s function is formulated using the equivalent circuit method to describe the relation between electromagnetic fields and electrical current densities in structures composed of stratified media and several metallizations. Slab and partially-filled waveguide formed by metallic walls with infinite conductivity are analyzed to validate the formulation. Analyses of reflectarrays with few elements are performed using the method of moments to numerically calculate the densities of electric current that flow on the patches. Additionally, entire-domain basis function with segmented edge condition to model the impulsive behavior of current density at the edges of the scatterers is proposed. Different phase curves are obtained due to the patches positions in the reflectarray using the variable size technique. Phase curves are calculated considering a total electric field composed of scattered, reflected and diffracted fields. Commercial software is used to verify the obtained results. Finally, the desired phase in each cell is determined using the progressive phase concept and the particle swarm optimisation method. The phase curves are used to design the dimensions of the printed elements, and, thus, ensure the desired phase distribution on the surface of the reflectarrays, and, consequently, the specified radiation pattern.

A procedure for computational dosimetry to verify safety standards compliance of mobile communications base stations is presented. Compared with the traditional power density method, a procedure based on more rigorous physics was devised, requiring computation or measurement of the specific absorption rate (SAR) within the biological tissue of a person at an arbitrary distance. This uses a hybrid methd of moments/finite difference time domain (MoM/FDTD) numerical method in order to determine the field or SAR distribution in complex penetrable media, without the computational penalties that would result from a wholly FDTD simulation. It is shown that the transmitted power allowed by the more precise SAR method is, in many cases, between two and five times greater than that allowed by standards implementing the power flux density method.