La résolution de systèmes linéaires creux est un problème qui apparaît dans de nombreuses applications scientifiques, et les solveurs creux sont une étape coûteuse pour ces applications ainsi que pour des solveurs plus avancés comme les solveurs hybrides direct-itératif. Pour ces raisons, optimiser la performance de ces solveurs pour les architectures modernes est un problème critique. Cependant, les contraintes mémoire et le temps de résolution limitent l’utilisation de ce type de solveur pour des problèmes de très grande taille. Pour les approches concurrentes, par exemple les méthodes itératives, des préconditionneurs garantissant une bonne convergence pour un large ensemble de problèmes sont toujours inexistants. Dans la première partie de cette thèse, nous présentons deux approches exploitant la compression Block Low-Rank (BLR) pour réduire la consommation mémoire et/ou le temps de résolution d’un solveur creux. Ce format de compression à plat, sans hiérarchie, permet de tirer profit du caractère low-rank des blocs apparaissant dans la factorisation de systèmes linéaires creux. La solution proposée peut être utilisée soit en tant que solveur direct avec une précision réduite, soit comme un préconditionneur très robuste. La première approche, appelée Minimal Memory, illustre le meilleur gain mémoire atteignable avec la compression BLR, alors que la seconde approche, appelée Just-In-Time, est dédiée à la réduction du nombre d’opérations, et donc du temps de résolution. Dans la seconde partie, nous présentons une stratégie de reordering qui augmente la granularité des blocs pour tirer davantage profit de la localité dans l’utilisation d’architectures multi-coeurs et pour fournir de tâches plus volumineuses aux GPUs. Cette stratégie s’appuie sur la factorisation symbolique par blocs pour raffiner la numérotation produite par des outils de partitionnement comme Metis ou Scotch, et ne modifie pas le nombre d’opérations nécessaires à la résolution du problème. A partir de cette approche, nous proposons dans la troisième partie de ce manuscrit une technique de clustering low-rank qui a pour objectif de former des clusters d’inconnues au sein d’un séparateur. Nous démontrons notamment les intérêts d’une telle approche par rapport aux techniques de clustering classiquement utilisées. Ces deux stratégies ont été développées pour le format à plat BLR, mais sont également une première étape pour le passage à un format hiérarchique. Dans la dernière partie de cette thèse, nous nous intéressons à une modification de la technique de dissection emboîtée afin d’aligner les séparateurs par rapport à leur père pour obtenir des structures de données plus régulières. / Solving sparse linear systems is a problem that arises in many scientific applications, and sparse direct solvers are a time consuming and key kernel for those applications and for more advanced solvers such as hybrid direct-iterative solvers. For those reasons, optimizing their performance on modern architectures is critical. However, memory requirements and time-to-solution limit the use of direct methods for very large matrices. For other approaches, such as iterative methods, general black-box preconditioners that can ensure fast convergence for a wide range of problems are still missing. In the first part of this thesis, we present two approaches using a Block Low-Rank (BLR) compression technique to reduce the memory footprint and/or the time-to-solution of a supernodal sparse direct solver. This flat, non-hierarchical, compression method allows to take advantage of the low-rank property of the blocks appearing during the factorization of sparse linear systems. The proposed solver can be used either as a direct solver at a lower precision or as a very robust preconditioner. The first approach, called Minimal Memory, illustrates the maximum memory gain that can be obtained with the BLR compression method, while the second approach, called Just-In-Time, mainly focuses on reducing the computational complexity and thus the time-to-solution. In the second part, we present a reordering strategy that increases the block granularity to better take advantage of the locality for multicores and provide larger tasks to GPUs. This strategy relies on the block-symbolic factorization to refine the ordering produced by tools such as Metis or Scotch, but it does not impact the number of operations required to solve the problem. From this approach, we propose in the third part of this manuscript a new low-rank clustering technique that is designed to cluster unknowns within a separator to obtain the BLR partition, and demonstrate its assets with respect to widely used clustering strategies. Both reordering and clustering where designed for the flat BLR representation but are also a first step to move to hierarchical formats. We investigate in the last part of this thesis a modified nested dissection strategy that aligns separators with respect to their father to obtain more regular data structure.

Trata da problemática dos riscos socioambientais, relacionando-os às fragilidades do ambientes, à vulnerabilidade da sociedade e ao uso e ocupação da terra. Maior ênfase, no entanto, é dada às áreas urbanizadas, em especial na cidade de Fortaleza-CE. Referida cidade passou por um crescimento desordenado, que trouxe uma série de problemas socioambientais. O estudo da fragilidade ambiental tem bases teóricas, metodológicas e conceituais na análise ambiental integrada, que perpassa a funcionalidade dos ambientes, considerando, inclusive, os processos históricos de produção e construção do território. A vulnerabilidade social e os padrões de uso e ocupação da terra definem a exposição dos grupos sociais aos riscos. A associação dessas características possibilitou o estabelecimento de diferentes categorias de susceptibilidade aos riscos, considerando de um lado as fragilidades ambientais em face do desenvolvimento das atividades humanas, e, de outro lado, por meio do índice sintético da vulnerabilidade social à distribuição desigual da população e dos riscos no território. Deste modo, foram estabelecidas bases que possam conduzir a um adequado ordenamento do território, com vistas a minimizar a incidência dos riscos socioambientais. / This research is about the issue of socioenvironmental risks and its relations with environmental fragility, the societies vulnerability and land use and occupation. Its focused on urban areas, especially in Fortaleza (CE). The accelerated and disordered occupation in Fortaleza has bought several socioenvironmental problems. The environmental fragility issue has its theoretical, methodological and conceptual bases on integrated analysis that encompasses the environments functionality considering the historical production processes and the construction of territory. The social vulnerability and the land use and occupation patterns define the social groups exposure to risks. The association of such characteristics was used to establish different susceptibility categories to the risks, considering on one hand the environmental fragilities face to human activities development, and in the other, the population distribution and risks on the territory by the synthetic index of social vulnerability. In this sense, bases that lead to an adequate territorial ordering were established aiming to minimize the incidence of socioenvironmental risks.

O desenvolvimento de novos materiais cerâmicos condutores de prótons é tecnologicamente importante devido às suas aplicações como eletrólitos em células a combustível de óxido sólido (SOFCs), dispositivos eletroquímicos fontes de energia limpa e renovável. Entre os desafios encontrados na aplicação nessas células de novos óxidos cerâmicos prótoncondutores está a alta resistividade de seus contornos de grão, que bloqueiam eletricamente a corrente de defeitos protônicos. Esse fato torna relevantes as pesquisas sobre a natureza desse fenômeno de bloqueio e sua relação com as características próton-condutivas do material. Nesta dissertação, investigamos as propriedades do sistema não estequiométrico Ba3Ca1,18Nb1,82O9-δ, juntamente com os compostos Ba3Ca1,18Nb1,52R0,3O9-δ (R = Y, Gd, Sm, Nd). Pós cristalinos dessas composições foram sintetizados via reação em estado sólido e utilizados na fabricação de cerâmicas. Os materiais foram caracterizados do ponto de vista estrutural, microestrutural, vibracional e elétrico, utilizando diversas técnicas físicas e correlacionado as características de cada composição às suas propriedades condutivas. Experimentos de difração de raios X e cálculos de fator de estrutura revelaram o aparecimento de ordenamento estrutural na estrutura perovskita de todos os sistemas sintetizados, sendo a intensidade de reflexões características utilizada como parâmetro de ordenamento. A microestrutura das cerâmicas foi otimizada em função do tempo de sinterização e as amostras finais obtidas apresentaram baixa porosidade. A caracterização vibracional das cerâmicas via espectroscopia Raman corrobora o resultado do ordenamento obtido via difração, além de indicar a presença ou preenchimento de vacâncias de oxigênio na rede cristalina dos materiais. Análises vibracionais também permitiram o estudo da acumulação de defeitos na borda das amostras cerâmicas e de sua estabilidade química, sendo tais diretamente correlacionadas à dopagem feita no material. O estudo de propriedades elétricas por espectroscopia de impedância aliado a modelagens por circuitos equivalentes permitiu separar propriedades elétricas de grão e contorno de grão. O efeito de bloqueio elétrico dos contornos pôde ser observado nas cerâmicas estudadas e está correlacionado à condutividade do interior dos grãos, sendo tais resultados interpretados à luz do modelo de cargas espaciais. / The development of novel proton conducting ceramic materials is technologically important due to their application as electrolytes in solid oxide fuel cells (SOFC), electrochemical devices that constitute clean and renewable energy sources. Among the challenges faced when applying new oxide materials to fuel cells is the high resistivity of grain boundaries, which causes an electrical blocking effect of proton transport. This issue stimulates research on the nature of the blocking phenomenon and its relation to proton-conducting properties of the materials. In the present study, we investigate physical properties of the non-stoichiometric system Ba3Ca1,18Nb1,82O9-δ together with the compounds Ba3Ca1,18Nb1,52R0,3O9-δ (R = Y, Nd, Sm, Gd). Crystalline powders of all the chemical compositions were synthetized by a solidstate reaction and used to prepare ceramic samples. The materials were characterized in terms of their structural, microstructural, vibrational and electrical properties by the usage of various physical techniques, correlating characteristics of each composition to its proton-conduction properties. X-ray diffraction experiments combined with structure factor calculations revealed the presence of perovskite structural ordering in all the compounds, and the intensity of characteristic reflections was used as ordering parameter. Ceramic microstructure was optimized with respect to sintering time and the final samples achieved low porosity. Vibrational characterization by Raman spectroscopy supported the ordering result obtained by diffraction and indicated the presence or filling of oxygen vacancies in the materials crystal structures. Vibrational analysis also allowed the study of defect accumulation near the ceramic samples edges and their chemical stability, which are directly related to material doping. Electrical studies by impedance spectroscopy together with equivalent circuit modeling allowed the separation of grain and grain boundary electrical properties. Blocking effect by the boundaries was observed on the considered samples and is correlated to conductivity on grain bulk. Blocking results are interpreted in the framework of a space charge model.

Preventive quality ordering is a provider intervention aimed at disease prevention through the ordering of industry-recommended health maintenance tests. This pilot study evaluated the effectiveness of provider mentoring/coaching to improve preventive quality ordering using the 2014 Agency for Healthcare Research and Quality best practice preventive clinical services guidelines. Literature indicates provider inconsistency in preventive and quality ordering as the primary cause of disparate health outcomes. Guided by theories of modeling and role-modeling, as well as the theory of cognitive continuum, this pilot study offered provider mentoring/coaching to encourage timely preventative quality ordering. Routinely monitored historic provider practice patterns in a proprietary database were analyzed; 10 providers with the lowest ordering patterns were identified for participation. Mentoring/coaching interventions were provided to improve preventive quality measure ordering. This process included a review of the 2014 Adult Healthcare Effectiveness Data and Information Set documentation criteria, a preventive measures clinical checklist, medical record preparation guidance, clinical shadowing, and post-training discussions. Following the pilot, a 5-person subject matter expert panel of key organizational leaders used on-site observations and standardized semi-structured interviews to evaluate the usefulness of mentoring/coaching and the developed documents to improve timely quality ordering. This small-scale pilot study (a) improved providers' awareness of quality ordering through peer mentoring, communication, and training; and (b) provided a platform for future initiatives. A larger follow-up study will allow healthcare leaders/providers to address disparate health outcomes, and patients will likely benefit from optimal delivery of preventive care.

Neste trabalho foram utilizadas técnicas de difração de raios-x para estudar propriedades químicas e estruturais de ilhas de Ge:Si(001). Através de experimentos de difração por incidência rasante foi realizado um mapeamento estrutural da relaxação de strain dentro de pirâmides e domos de Ge. Alterando-se a energia dos raios-x próximo à borda K do Ge – em medidas de difração anômala – foi possível determinar a composição química dos dois tipos de ilhas. A energia elástica, obtida correlacionando-se estes dois resultados, provou ser um dos fatores responsáveis pelas transições morfológicas neste sistema. Uma extensão dos resultados, com o uso de um novo método de análise, permitiu um completo mapeamento tri-dimensional da estrutura e estequiometria dos domos de Ge. Por último, foi observada a existência de uma liga ordenada de SiGe dentro dos domos, indicando o importante papel da cinética de crescimento na incorporação de Si nas ilhas.

<p>Understanding lithium-ion insertion/extraction mechanisms in battery electrode materials is of crucial importance in developing new materials with better cycling performance. In this thesis, these mechanisms are probed for two different potential cathode materials by a combination of electrochemical and single-crystal X-ray diffraction studies. The materials investigated are V₆O₁₃and cubic LiMn₂O₄spinel.</p><p>Single-crystal X-ray diffraction studies of lithiated phases in the Li_xV₆O₁₃ system (x=2/3 and 1) exhibit superlattice phenomena and an underlying Li⁺ ion insertion mechanism which involves the stepwise addition of Li⁺ions into a two-dimensional array of chemically equivalent sites. Each successive stage in the insertion process is accompanied by a rearrangement of the Li⁺ ions together with an electron redistribution associated with the reduction of specific V-atoms in the structure. This results in the formation of electrochemically active sheets in the structure. A similar mechanism occurs in the LiMn₂O₄ delithiation process, whereby lithium is extracted in a layered arrangement, with the Mn atoms forming charge-ordered Mn³⁺/Mn⁴⁺ layers.</p><p>Lithium-ion insertion/extraction processes in transition-metal oxides would thus seem to occur through an ordered two-dimensional arrangement of lithium ions extending throughout the structure. The lithium ions and the host structure rearrange cooperatively to form superlattices through lithium and transition-metal ion charge-ordering. A picture begins to emerge of a universal two-dimensional lithium-ion insertion/extraction mechanism analogous to the familiar staging sequence in graphite.</p>

Understanding lithium-ion insertion/extraction mechanisms in battery electrode materials is of crucial importance in developing new materials with better cycling performance. In this thesis, these mechanisms are probed for two different potential cathode materials by a combination of electrochemical and single-crystal X-ray diffraction studies. The materials investigated are V6O13 and cubic LiMn2O4 spinel. Single-crystal X-ray diffraction studies of lithiated phases in the LixV6O13 system (x=2/3 and 1) exhibit superlattice phenomena and an underlying Li+ ion insertion mechanism which involves the stepwise addition of Li+ ions into a two-dimensional array of chemically equivalent sites. Each successive stage in the insertion process is accompanied by a rearrangement of the Li+ ions together with an electron redistribution associated with the reduction of specific V-atoms in the structure. This results in the formation of electrochemically active sheets in the structure. A similar mechanism occurs in the LiMn2O4 delithiation process, whereby lithium is extracted in a layered arrangement, with the Mn atoms forming charge-ordered Mn3+/Mn4+ layers. Lithium-ion insertion/extraction processes in transition-metal oxides would thus seem to occur through an ordered two-dimensional arrangement of lithium ions extending throughout the structure. The lithium ions and the host structure rearrange cooperatively to form superlattices through lithium and transition-metal ion charge-ordering. A picture begins to emerge of a universal two-dimensional lithium-ion insertion/extraction mechanism analogous to the familiar staging sequence in graphite.

The properties of dilute magnetic semiconductors have been studied by combined ab initio, Monte Carlo, and experimental techniques. This class of materials could be very important for future spintronic devices, that offer enriched functionality by making use of both the spin and the charge of the electrons. The main part of the thesis concerns the transition metal doped ZnO. The role of defects on the magnetic interactions in Mn-doped ZnO was investigated. In the presence of acceptor defects such as zinc vacancies and oxygen substitution by nitrogen, the magnetic interactions are ferromagnetic. For dilute concentrations of Mn (~ 5%) the ordering temperature of the system is low, due to the short ranged character of the exchange interactions and disorder effects. The clustering tendency of the Co atoms in a ZnO matrix was also studied. The electronic structure, and in turn the magnetic interactions among the Co atoms, is strongly dependent on the exchange-correlation functional used. It is found that Co impurities tend to form nanoclusters and that the interactions among these atoms are antiferromagnetic within the local spin density approximation + Hubbard U approach. The electronic structure, as well as the chemical and magnetic interactions in Co and (Co,Al)-doped ZnO, was investigated by joined experimental and theoretical techniques. For a good agreement between the two, approximations beyond the local density approximation must be used. It is found that the Co atoms prefer to cluster within the semiconducting matrix, a tendency which is increased with Al co-doping. We envision that it is best to describe the system as superparamagnetic due to the formation of  Co nanoclusters within which the interactions are antiferromagnetic. The magnetic anisotropy and evolution of magnetic domains in Fe81Ni19/Co(001) superlattices were investigated both experimentally, as well as using model spin dynamics. A magnetic reorientation transition was found.

Polymer electrolytes are the most obvious candidates for safe "all-solid" Li-ion batteries and other electrochemical devices. However, they still have relatively poor ionic conductivities, which limits their wider adoption in commercial applications. It has earlier been the conventional wisdom that only amorphous phases of polymer electrolytes show usefully high ionic conduction, while crystalline forms are insulators. However, this has been challenged in the last decade by the discovery of highly organized, low-dimensional ion-conducting materials. Specifically, the crystalline phases of LiXF6.PEO6 exhibit higher ionic conductivities than their amorphous counterparts, with the Li-ion conduction taking place along the PEO channels. Polymer chain-length and chain-end registry has emerged as potentially significant in determining ionic conduction in these materials. Molecular Dynamics simulations have therefore been made of short-chain, monodisperse (Mw~1000), methoxy end-capped LiPF6.PEO6 to examine relationships between ion conduction and mode of chain-ordering. Studies of smectic and nematic arrangements of PEO chains have revealed that ion-transport mechanisms within the smectic planes formed by cooperative chain-end registry appear to be more suppressed by ion-pairing than in-channel conduction. Disorder phenomena in the chain-end regions emerge as a critical factor in promoting Li-ion migration across chain-gaps, as does the structural continuity of the PEO channels. Simulations incorporating ~1% aliovalent SiF62- dopants further suggest an increase in Li-ion conduction when the extra Li-ions reside within the PEO channels, with the anion influencing charge-carrier concentration through enhanced ion-pair formation. XRD techniques alone are shown to be inadequate in ascertaining the significance of the various short-chain models proposed; atomistic modelling is clearly a helpful complement in distinguishing more or less favourable situations for ion conduction. Though providing valuable insights, it must be concluded that this work has hardly brought us significantly closer to breakthroughs in polymer electrolyte design; the critical factors which will make this possible remain as yet obscure.

The optimum ordering strategies for the coded V-BLAST system with capacity achieving temporal codes on each stream are studied in this thesis. Mathematical representations of the optimum detection ordering strategies for the coded V-BLAST under instantaneous rate allocation (IRA), uniform power/rate allocation (URA), instantaneous power allocation(IPA) and instantaneous power/rate allocation (IPRA) are derived. For two transmit antennas, it is shown that the optimum detection strategies are based on the per-stream before-processing channel gains. Based on approximations of the per-stream capacity equation, closed-form expressions of the optimal ordering strategy under the IRA at low and high signal to noise ratio (SNR) are derived. Necessary optimality conditions under the IRA are given. Thresholds for the low, intermediate and high SNR regimes in the 2-Tx-antenna system under the IPRA are determined, and the SNR gain of the ordering is studied for each regime. Performances of simple suboptimal ordering strategies are analysed, some of which perform very close to the optimum one.