Electro-Magnetic Source Imaging (EMSI) is the estimation of the position, orientation and strength of active electrical sources within the brain from electrical and magnetic measurements. For an accurate source localization, the head model must correctly represent the electrical and geometrical properties of the head. To solve the forward problem using realistic head models numerical techniques must be used. This work uses the Boundary Element Method (BEM) for solving the forward problem. The accuracy of the existing BEM formulation is improved by using second order elements, recursive integration and the isolated problem approach (IPA). Two new formulations are developed to improve the solution speed by computing transfer matrices for EEG and MEG solutions. The IPA formulation is generalized and integrated into the accelerated BEM algorithm. Once the transfer matrices are computed, the forward solutions take about 300 ms for a 256 sensor EEG and MEG system. The head model used in the BEM solutions is constructed by segmenting three dimensional multimodal magnetic resonance images. For segmentation, a semi-automatic hybrid algorithm is developed that makes use of snakes, morphological operations, thresholding and region growing. The mesh generation algorithm allows intersecting tissue compartments. For the inverse problem solution genetic algorithm (GA) is used to search for a given number of dipoles. Source localization with simulated data show that the localization error is within 1.1 mm for EEG and 1.2 mm for MEG when SNR is 10 on a realistic model with 7 compartments. When a single-dipole source in a realistic model is explored using a best-fit spherical model, the localization errors increase up to 8.5 mm for EEG and 7 mm for MEG. Similar tests are also performed with multiple dipoles. It was observed that realistic models provide definitely more accurate results compared to spherical models. The EMSI approach is also tested using experimental EEG data to localize the sources of auditory evoked potentials. The reconstructed source locations are correctly found in the Heschl&#039 / s gyrus. In conclusion, this thesis presents a complete source localization framework for future brain research using the EMSI.

Human brain functions are based on the electrochemical activity and interaction of the neurons constituting the brain. Some brain diseases are characterized by abnormalities of this activity. Detection of the location and orientation of this electrical activity is called electro-magnetic source imaging (EMSI) and is of signicant importance since it promises to serve as a powerful tool for neuroscience. Boundary Element Method (BEM) is a method applicable for EMSI on realistic head geometries that generates large systems of linear equations with dense matrices. Generation and solution of these matrix equations are time and memory consuming due to the size of the matrices and high computational complexity of direct methods. This study presents a relatively cheap and eective solution the this problem and reduces the processing times to clinically acceptable values using parallel cluster of personal computers on a local area network. For this purpose, a cluster of 8 workstations is used. A parallel BEM solver is implemented that distributes the model eciently to the processors. The parallel solver for BEM is developed using the PETSc library. The performance of the iv solver is evaluated in terms of CPU and memory usage for dierent number of processors. For a 15011 node mesh, a speed-up eciency of 97.5% is observed when computing transfer matrices. Individual solutions can be obtained in 520 ms on 8 processors with 94.2% parallellization eciency. It was observed that workstation clusters is a cost eective tool for solving complex BEM models in clinically acceptable time. Eect of parallelization on inverse problem is also demonstrated by a genetic algorithm and very similar speed-up is obtained.

We propose an improved technique to calculate the linear response of a single and multiple plates models due to ocean waves. The single plate model is the basis for the multiple plates model which we take to be a periodic array of identical plates. For the single plate model we solve the plate displacement by the Finite Element Method (FEM) and the water potential by the Boundary Element Method (BEM). The displacement is expanded in terms of the basis functions of the FEM. The boundary integral equation representing the potential is approximated by these basis functions. The resulting integral operator involving the free-surface Green's function is solved using an elementary integration scheme. Results are presented for the single plate model. We then use the same technique to solve for the periodic array of plates problem because the single and the periodic array plates model differ only in the expression of the Green's function. For the periodic array plate model the boundary integral equation for the potential involves a periodic Green's function which can be obtained by taking an infinite sum of the free-surface Green's function for the single plate model. The solution for the periodic array plate is derived in the same way as the single plate model. From this solution we then calculate the waves scattered by this periodic array.

Doctor of philosophy (Ph D) / This thesis deals with numerical investigations on some rheological problems of fibre suspensions: the fibre level simulation of non-dilute fibre suspensions in shear flow; the numerical simulation of complex fibre suspension flows and simulating the particle motion in viscoelastic flows. These are challenging problems in rheology. Two numerical approaches were developed for simulating non-dilute fibre suspensions from the fibre level. The first is based on a model that accounts for full hydrodynamic interactions between fibres, which are approximately calculated as a superposition of the long-range and short-range hydrodynamic interactions. The long-range one is approximated by using slender body theory and includes infinite particle interactions. The short-range one is approximated in terms of the normal lubrication forces between close neighbouring fibres. The second is based on a model that accounts only for short-range interactions, which comprise the lubrication forces and normal contact and friction forces. These two methods were applied to simulate the microstructure evolution and rheological properties of non-dilute fibre suspensions. The Brownian configuration method was combined with the highly stable finite element method to simulate the complex flow of fibre suspensions. The method is stable and robust, and can provide both micro and macro information. It does not require any closure approximations in calculating the fibre stress tensor and is more efficient and variance reduction, compared to CONNFFESSITT, for example. The flow of fibre suspensions past a sphere in a tube and the shear induced fibre migration were successfully simulated using this method The completed double layer boundary element method was extended to viscoelastic flow cases. A point-wise solver was developed to solve the constitutive equation point by point and the fixed least square method was employed to interpolate and differentiate data locally. The method avoids volume meshing and only requires the boundary mesh on particle surfaces and data points in the flow domain. A sphere settling in the Oldroyd-B fluid and a prolate spheroid rotating in shear flow of the Oldroyd-B fluid were simulated. Based on the simulated orbit of a prolate spheroid in shear flow, a constitutive model for the weakly viscoelastic fibre suspensions was proposed and its predictions were compared with some available experimental results. All simulated results are in general agreement with experimental and other numerical results reported in literature. This indicates that these numerical methods are useful tools in rheological research.

Cette thèse est consacrée à l'étude numérique des interactions non-linéaires entre des vagues et un corps rigide perçant la surface libre. La méthode développée repose sur un modèle d'éléments de frontière qui réduit la dimensionnalité du problème d'une dimension. Dans un premier temps, un modèle2D est appliqué à des géométries simples et permet de démontrer la pertinence de l'approche envisagée pour la prédiction des mouvements d'une structure flottante soumise à des vagues monochromatiques régulières. Dans un second temps, en nous inspirant d'un modèle potentiel non-linéaire 3D développé par Grilli textit{et al.}~cite{grilli2001fully}, nous proposons une généralisation de la méthode pour des maillages triangulaires non-structurés de surfaces 3D. Le modèle développé permet de traiter des configurations arbitraires de plusieurs cylindres verticaux en interaction avec les vagues. Nous présentons des cas de validation de nature académique qui permettent d'apprécier le comportement du modèle numérique. Puis nous nous tournons vers l'application visée par EDF R&D, qui concerne le dimensionnement d'éoliennes off-shore flottantes. Un flotteur de type semi-submersible est évalué à l'aide du modèle non-linéaire / This PhD work is devoted to the study of nonlinear interactions between waves and floating rigid structures. The developed model relies on a boundary element method which reduces the dimensionality of the problem by one. First, a 2D model is applied to basic geometries and allows us to demonstrate the validity of the method for predicting the motion of a floating structrure subject to incoming monochromatic regular waves. Secondly, getting inspired by the 3D fully nonlinear potential flow model of Grilli textit{et al.}~cite{grilli2001fully}, we propose a novel model which generalizes the method for unstructured triangular meshes of 3D surfaces. The proposed model is able to deal with arbitrary configurations of multiple vertical cylinders interacting with the waves. We present academic validation test cases which show how the model works and behaves. Finally, we study situations of interest for EDF R&D related to floating off-shore wind turbines. A semi-submersible platform is evaluated with the nonlinear model

Este trabalho apresenta o desenvolvimento e implementação computacional de técnicas de otimização de topologia para problemas governados pela equação de Poisson. O método numérico utilizado para solução numérica das equações foi o método dos elementos de contorno (MEC). Para tanto, três metodologias foram desenvolvidas. A primeira é direcionada à aplicação de algoritmos genéticos (AG) para investigar como um domínio inicialmente preenchido com cavidades aleatórias evolui durante um processo de otimização e verificar a possibilidade de se extrair topologias ótimas a partir da interpretação da solução encontrada. Os contornos externos permanecem fixos enquanto as posições e as dimensões das cavidades são otimizadas com o objetivo extremizar uma função custo especificada. O desempenho do algoritmo proposto é ilustrada com uma série de exemplos e os resultados são discutidos. A segunda metodologia apresenta um algoritmo numérico para otimização topológica baseado na avaliação da derivada topológica (DT), adotando a energia potencial total como função custo. Este procedimento é uma alternativa às tradicionais técnicas de otimização, evitando assim soluções de projeto com densidade de material intermediária. Sólidos com comportamento anisotrópico são estudados sob condições de contorno de Robin, Neumann e Dirichlet. Uma transformação linear de coordenadas é utilizada para mapear o problema original e suas condições de contorno para um novo domínio equivalente isotrópico, onde o procedimento de otimização é aplicado. A solução otimizada é então transformada de volta ao domínio original. A metodologia proposta mostrou-se particularmente atrativa para resolver esta classe de problemas já que o MEC dispensa o uso de malha no domínio, reduzindo significantemente o custo computacional. Na última parte deste trabalho foi implementada uma formulação de sensibilidade topológica para problemas de otimização de transferência de calor e massa simultâneos. Como as sensibilidades para cada equação diferencial são diferentes, utiliza-se um coeficiente de ponderação para compor a sensibilidade do problema acoplado. Isto permite a imposição de distintos fatores para cada problema, de acordo com uma prioridade especificada. Diversos exemplos são apresentados e seus resultados comparados com os da literatura, quando disponíveis, a fim de validar as formulações propostas. / This work presents the computational development and implementation of topology optimization techniques for problems governed by the Poisson equation. The boundary element method was the numerical technique chosen to solve the equations. Three different methodologies were developed aiming this objective. The first methodology is directed to the application of genetic algorithms to investigate how a domain previously populated with randomly placed cavities evolves during the optimization process, and to verify the resemblance of the final solution with a optimal design. The external boundaries remain fixed during the process, while the location and dimension of the cavities are optimized in order to extremize a given cost function. The performance of the proposed algorithm is verified with a number of examples and the results are discussed. The second methodology presents a numerical algorithm for topology optimization based on the evaluation of topological derivatives, using the total potential energy as the cost function. This procedure is an alternative to the traditional optimization techniques, avoiding design solutions containing intermediary material densities. Solids with anisotropic constitutive behavior are studied under Robin, Neumann and Dirichlet boundary conditions. A linear coordinate transformation approach is used to map the original problem into an isotropic one, where the optimization is carried out. The final solution is then mapped back to the original coordinate system. The proposed method was found to be an attractive way to solve this class of problems, since no interior mesh is necessary, which reduces significantly the computational cost of the analysis. In the last part of the present work the topological derivative approach was further developed to deal with the optimization of problems under simultaneous heat and mass transfer. Since the sensitivities for each differential equation are different, a weighting factor was used to evaluate the final sensitivities of the coupled problem. This allows the imposition of different priorities for each problem Several examples are presented and their results are compared with the literature, when available, in order to validate the proposed formulations.

Este trabalho apresenta o desenvolvimento e implementação computacional de técnicas de otimização de topologia para problemas governados pela equação de Poisson. O método numérico utilizado para solução numérica das equações foi o método dos elementos de contorno (MEC). Para tanto, três metodologias foram desenvolvidas. A primeira é direcionada à aplicação de algoritmos genéticos (AG) para investigar como um domínio inicialmente preenchido com cavidades aleatórias evolui durante um processo de otimização e verificar a possibilidade de se extrair topologias ótimas a partir da interpretação da solução encontrada. Os contornos externos permanecem fixos enquanto as posições e as dimensões das cavidades são otimizadas com o objetivo extremizar uma função custo especificada. O desempenho do algoritmo proposto é ilustrada com uma série de exemplos e os resultados são discutidos. A segunda metodologia apresenta um algoritmo numérico para otimização topológica baseado na avaliação da derivada topológica (DT), adotando a energia potencial total como função custo. Este procedimento é uma alternativa às tradicionais técnicas de otimização, evitando assim soluções de projeto com densidade de material intermediária. Sólidos com comportamento anisotrópico são estudados sob condições de contorno de Robin, Neumann e Dirichlet. Uma transformação linear de coordenadas é utilizada para mapear o problema original e suas condições de contorno para um novo domínio equivalente isotrópico, onde o procedimento de otimização é aplicado. A solução otimizada é então transformada de volta ao domínio original. A metodologia proposta mostrou-se particularmente atrativa para resolver esta classe de problemas já que o MEC dispensa o uso de malha no domínio, reduzindo significantemente o custo computacional. Na última parte deste trabalho foi implementada uma formulação de sensibilidade topológica para problemas de otimização de transferência de calor e massa simultâneos. Como as sensibilidades para cada equação diferencial são diferentes, utiliza-se um coeficiente de ponderação para compor a sensibilidade do problema acoplado. Isto permite a imposição de distintos fatores para cada problema, de acordo com uma prioridade especificada. Diversos exemplos são apresentados e seus resultados comparados com os da literatura, quando disponíveis, a fim de validar as formulações propostas. / This work presents the computational development and implementation of topology optimization techniques for problems governed by the Poisson equation. The boundary element method was the numerical technique chosen to solve the equations. Three different methodologies were developed aiming this objective. The first methodology is directed to the application of genetic algorithms to investigate how a domain previously populated with randomly placed cavities evolves during the optimization process, and to verify the resemblance of the final solution with a optimal design. The external boundaries remain fixed during the process, while the location and dimension of the cavities are optimized in order to extremize a given cost function. The performance of the proposed algorithm is verified with a number of examples and the results are discussed. The second methodology presents a numerical algorithm for topology optimization based on the evaluation of topological derivatives, using the total potential energy as the cost function. This procedure is an alternative to the traditional optimization techniques, avoiding design solutions containing intermediary material densities. Solids with anisotropic constitutive behavior are studied under Robin, Neumann and Dirichlet boundary conditions. A linear coordinate transformation approach is used to map the original problem into an isotropic one, where the optimization is carried out. The final solution is then mapped back to the original coordinate system. The proposed method was found to be an attractive way to solve this class of problems, since no interior mesh is necessary, which reduces significantly the computational cost of the analysis. In the last part of the present work the topological derivative approach was further developed to deal with the optimization of problems under simultaneous heat and mass transfer. Since the sensitivities for each differential equation are different, a weighting factor was used to evaluate the final sensitivities of the coupled problem. This allows the imposition of different priorities for each problem Several examples are presented and their results are compared with the literature, when available, in order to validate the proposed formulations.

Neste trabalho os problemas de determinação dos Fatores de Intensidade de Tensão KI e KII estáticos e dinâmicos são tratados numericamente utilizando uma formulação alternativa do Método dos Elementos de Contorno (MEC) com solução fundamental de Kelvin e matriz de massa para os problemas dinâmicos. A trinca é suposta retangular inicialmente, com suas faces não-coincidentes. Tanto as faces da trinca, quanto o contorno externo são discretizados em elementos de contorno reto com variação de forças de deslocamentos quadráticas, não havendo, portanto distinção entre elementos de trinca e de contorno externo. Integrais analíticas também são obtidas para o elemento linear isoparamétrico. As células de domínio apresentam formato triangular e suas integrais são solucionadas semi-analiticamente. Quanto às integrais de contorno, essas são obtidas analiticamente segundo eixos de referência locais, procedendo-se em seguida a rotação pra eixos globais. O algoritmo de Houbolt é empregado como integrador temporal. Exemplos numéricos da determinação desses Fatores de Intensidade de Tensão são mostrados e comparados com resultados analíticos e resultados numéricos disponíveis na literatura. / In this work the stress intensity factors KI and KII for static and dynamic two-dimensional problem are obtained numerically by an alternative mass matrix boundary element formulation. The crack is considered a rectangular hole inside the domain and its faces are not coincident. Both crack faces and boundary are discretized by straight boundary elements with quadratic approximation. Domain cells are triangular with linear approximation and their integrals are developed semi-analytically. Boundary integrals are analytically performed, for linear and quadratic approximations. They are performed at local co-ordinate axes and transformed to global co-ordinate axes. The Houbolt algorithm is used to integrate the matrix time differential equation along time. Numerical examples are shown in order to compare the results obtained by the proposed formulation and the ones presents in literature.

O presente trabalho trata da análise do comportamento mecânico da madeira, material anisotrópico, submetida ao crescimento de fissuras e aos processos mecânicos que causam sua ruptura, utilizando os conceitos de Mecânica da Fratura. A avaliação do colapso de peças de madeira, pela da mecânica da fratura, foi realizada por meio de análises experimentais. Para a determinação das propriedades de fratura da madeira foram determinadas propriedades mecânicas da madeira, tais com: módulo de ruptura a flexão estática, módulo de elasticidade à flexão estática, resistência à tração paralela às fibras, resistência ao cisalhamento, resistência ao fendilhamento paralela às fibras e resistência à tração normal às fibras. Foram efetuados ainda ensaios experimentais referentes à mecânica da fratura em corpos de prova tipo SENB (Single Edge Notched Beam). A modelagem numérica foi realizada utilizando-se um código computacional baseado no método de elementos de contorno. Para as simulações foram utilizados os valores das propriedades físicas, de resistência e de elasticidade obtidas na caracterização experimental dos elementos de madeira. Os resultados experimentais foram comparados com as análises numéricas para determinação de uma formulação específica para situações de colapso de elementos de madeira. / The present work deals with the analysis of the mechanical behavior of wood, anisotropic material, submitted to the growth of cracks and the mechanical processes that cause its rupture, using the concepts of Fracture Mechanics. The evaluation of the collapse of pieces of wood, by the fracture mechanics, was carried out by means of experimental analyzes. To determine the fracture properties of the wood, mechanical properties of the wood were determined, such as: modulus of static bending, modulus of elasticity at static bending, tension resistance parallel to grain, shearing resistance, cleavage strength parallel to grain and tension resistance normal to grain. Experimental tests were performed on fracture mechanic in Single Edge Notched beams (SENB). Numerical modeling was performed using a computational code based on the boundary element method. For the simulations, the values of the physical properties, strength and elasticity obtained in the experimental characterization of the wood elements were used. The experimental results were compared with the numerical analyzes to determine a specific formulation for situations of collapse of wood elements.

A aplicação da análise inversa é objeto de estudo nos mais diversos campos da ciência e da engenharia. A motivação para o tratamento de tais problemas se deve ao fato de que em muitas aplicações dessas áreas do conhecimento, há a necessidade da identificação de parâmetros físicos e geométricos a partir de dados do domínio medidos experimentalmente, já que tais parâmetros de entrada são desconhecidos para uma análise direta do problema. Neste tipo de análise o problema principal está na quantidade e qualidade dos dados experimentais obtidos, que são na maioria das vezes insuficientes para garantir que o sistema gerado apresente solução única, gerando com isto um problema essencialmente mal-posto. Assim, de forma geral o emprego confiável da análise inversa implica na utilização de ferramentas eficientes de aquisição de dados experimentais aliada a técnicas numéricas de regularização que buscam a minimização da função objetiva gerada por algum método numérico, como por exemplo, o método dos elementos de contorno (MEC). Sendo assim, o presente trabalho tem por objetivo apresentar uma formulação para resolução de problemas inversos de valor de contorno e estimativa dos parâmetros do modelo coesivo, através de medidas de campos de deslocamentos, em sólidos bidimensionais com domínio formado por multi-regiões via (MEC), utilizando-se de técnicas tais como: mínimos quadrados, regularização de Tikhonov, decomposição em valor singular (SVD) e filtro de Tikhonov, para regularização do problema. Além disto, são apresentados alguns exemplos de aplicação da formulação desenvolvida. / The application of inverse analysis is nowadays subject of research of many fields in engineering and science. The motivation to consider this problem is due to the fact that in many applications of these knowledge areas, physical and geometric parameters, that are not directly known, can be identified using domain data measured experimentally. In this kind of analysis the main problem is the quantity and the quality of the obtained experimental data, which, many times, are not sufficient to guarantee that the generated system of equations has only one solution, leading therefore to an ill-posed problem. Thus, in general the reliable use of the inverse analysis requires using efficient tools for experimental data acquisition together with the numerical techniques of regularization needed to impose the minimization of the objective function written by using any numerical method, as the boundary element method (BEM) for instance. In this context, the objective of the present work is to derive a formulation for resolution of boundary-value inverse problems and to estimate the material parameters of the cohesive model, by using measured displacements fields, in multi-region two-dimensional solid by BEM, using techniques such as: least squares, Tikhonov regularization, singular value decomposition (SVD) and Tikhonov filtering, for the problem regularization. Some application examples are presented using the developed formulation to illustrate its performance.