O presente trabalho tem como objetivo caracterizar e correlacionar formas de onda de sinais de emissÃo acÃstica, utilizando mÃtodos de anÃlise espectral e fractal, em corpos de prova previamente soldados com seus respectivos defeitos, avaliando assim a integridade estrutural do material aÃo carbono ASTM 516 Grau 60, utilizado em vasos de pressÃo, tubulaÃÃes, esferas de armazenamento, etc. Os sinais de emissÃo acÃstica foram gerados submetendo os corpos de prova a ensaios de traÃÃo. Para anÃlise dos sinais no domÃnio dos tempos, foram utilizados o mÃtodo de contagem de caixas, que fornece a dimensÃo fractal, e o de intervalo re-escalado de Hurst, e no domÃnio das freqÃÃncias o mÃtodo espectral de Fourier. AtravÃs das relaÃÃes previstas pelas leis de escala, concluiu-se que os expoentes calculados pelos mÃtodos espectral de Fourier e reescalado de Hurst, correlacionam-se entre si, bem como com a dimensÃo fractal obtida pelo mÃtodo de contagem de caixas. Mostrou-se tambÃm a existÃncia de duas regiÃes de escala distintas que caracterizam a mudanÃa do regime elÃstico para o regime plÃstico quando as amostras foram submetidas ao carregamento de traÃÃo. Verificou-se ainda que este comporta-mento foi caracterÃstico tanto para os materiais soldados como para o metal de base. / The present work aims to characterize and correlate waveforms of acoustic emission signals, using methods of spectral analysis and fractal in specimens previously soldiers with their defects, thus evaluating the structural integrity of the material carbon steel ASTM 516 Grade 60 used in pressure vessels, piping, storage spheres, etc.. The acoustic emission signals were generated by subjecting the specimens to tensile tests. For analysis of the signals in the time domain, we used the method of box counting, which provides the fractal dimension, and re-scaled range of Hurst, and in the frequency domain the Fourier spectral method. Through the relations laid down by law scale, it was concluded that the exponents calculated by Fourier transform and spectral methods reescalado Hurst, are correlated with each other as well as the fractal dimension obtained from a box counting. It was also shown existence of two regions of different scale characterizing the change of the elastic to the plastic regime when the samples were subjected to traction. It was further found that this behavior is characteristic for both the welded material as for the parent metal.

Alcançar altas taxas de dados em comunicações sem fio é difícil. Altas taxas de dados para redes locais sem fio tornou-se comercialmente um sucesso por volta do ano de 2000. Redes de longa distância sem fio ainda são projetados e utilizados principalmente para serviços de voz em baixas taxas. Apesar de muitas tecnologias promissoras, a realidade de uma rede de área ampla que atenda muitos usuários com altas taxas de dados e largura de banda e consumo de energia razoáveis, além de uma boa cobertura e qualidade no serviço ainda é um desafio. O objetivo do IEEE 802.16 é projetar um sistema de comunicação sem fio para obter uma internet de banda larga para usuários móveis em uma área metropolitana. É importante perceber que o sistema WIMAX tem que enfrentar desafios semelhantes aos existentes sistemas celulares e seu desempenho eventual será delimitado pelas mesmas leis da física e da teoria da informação. Em muitas áreas da engenharia elétrica, tem-se direcionado atenção à miniaturização de componentes e equipamentos. Em particular, antenas não são exceções. Desde que Wheeler iniciou estudos sobre os limites fundamentais de miniaturização de antenas, o assunto tem sido discutido por muitos estudiosos e várias contribuições nesse sentido foram feitas desde então. Os avanços das últimas décadas na área de microeletrônica permitiram a miniaturização dos demais componentes empregados no desenvolvimento de equipamentos eletrônicos e disponibilizaram o uso de aparelhos compactos, leves e com diversas funcionalidades e aplicações comerciais. No entanto, ainda que a integração de circuitos seja uma realidade, a integração completa de um sistema de comunicação sem fio, incluindo a antena, é ainda um dos grandes desafios tecnológicos. No caso de antenas impressas procura-se continuamente desenvolver antenas que, além de compactas, apresentem maior largura de banda, ou operação em múltiplas bandas dada sua inerente característica de banda estreita em projetos convencionais. Neste trabalho, o foco está na miniaturização de antenas impressas através da aplicação de fractais. São apresentadas comparações entre antenas fractais quadradas de Minkowski e fractais triangulares de Koch. Inicialmente, antenas 6 impressas com geometrias convencionais quadradas e triangulares foram projetadas para ter a mesma frequência de ressonância. Depois disso, as estruturas fractais de Minkowski Island e Koch Loop foram implementadas nas antenas quadrada e triangular, respectivamente, até a terceira iteração. As frequências escolhidas foram as de 2,4 GHz, 3,5 GHz, 5,0 GHz e 5,8 GHz. Diversos protótipos foram construídos em dois substratos de permissividade diferentes, o FR-4 e o DUROID 5870. Para validar os resultados foram construídas antenas na frequência de 3,5 GHz para as geometrias quadrada e triangular e suas iterações fractais. A contribuição deste trabalho está na análise sobre as vantagens e desvantagens de cada uma das estruturas propostas. Dependendo dos requisitos de um projeto, a opção pode ser por antenas miniaturizadas com maior largura de banda, como normalmente acontece em alguns projetos comerciais. Entretanto, o interesse por bandas estreitas muitas vezes pode ser um requisito, principalmente para emprego militar, onde por vezes a máxima discrição na transmissão é uma exigência. Além disso, também foi feita uma análise sobre as geometrias que atingiram maior miniaturização. / Achieving high data rates in wireless communication is difficult. High data rates for wireless local area networks became commercially successful only around 2000. Wide area wireless networks are still designed and used primarily for low rate voice services. Despite many promising technologies, the reality of a wide area network that services many users at high data rates with reasonable bandwidth and power consumption, while maintaining high coverage and quality of service has not been achieved. The goal of the IEEE 802.16 was to design a wireless communication system processing to achieve a broadband internet for mobile users over a wide or metropolitan area. It is important to realize that WIMAX system have to confront similar challenges as existing cellular systems and their eventual performance will be bounded by the same laws of physics and information theory. In many areas of electrical engineering, miniaturization has been an important issue. Antennas are not an exception. After Wheeler initiated studies on the fundamental limits for miniaturization of antennas, this subject has been extensively discussed by several scholars and many contributions have been made. The advances of recent decades in the field of microelectronics enabled the miniaturization of components and provided the use of compact, lightweight, equipments with many features in commercial applications. Although circuit integration is a reality, the integration of a complete system, including its antenna, is still one of the major technological challenges. In the case of patch antennas, the search is for compact structures with increased bandwidth, due to the inherent narrowband characteristic of this type of antenna. In this work the focus is on a comparison between the Minkowski and the Koch Fractal Patch Antennas. Initially, patch antennas with conventional square and triangular geometries were simulated to present the same resonance frequency. After that, fractal Minkowski and Koch Island Loop antennas were implemented in the square and triangular geometries, respectively, to the third iteration. A comparison was made for two substrates of different permittivities FR-4 and DUROID 5870 at the frequencies of 2,4 GHz; 3,5 GHz; 5,0 GHz and 5,8 GHz. 8 Prototype antennas were built using FR-4 and DUROID 5870 to resonate at a frequency of 3,5 GHz to validate simulation results. The contribution of this work is the analysis of the advantages and disadvantages of each proposed fractal structure. According to the project requirements, the best option can be use a miniaturized antenna with a wider band, as in commercial projects. Particularly in military applications, a narrow band antenna can be a requirement, as sometimes maximum discretion in transmission is a paramount. An additional analysis was performed to verify which of the geometries fulfilled the miniaturization criteria of Hansen.

The development of interactive communication technologies, such as email and the Internet, has stimulated much discussion about their potential to assist in the renewal of democracy. Globalisation, intimately connected with technological advance, has likewise led to considerations of 'governance' that transcend the nation state. This thesis poses the general question: Can the efficiencies and economies which drive commercial applications of interactive technology be applied to democratic policy processes? What conditions will facilitate this? The perspective taken here views policy as a communication process between networks of actors who seek to use resources and decisions to promote their value systems (Considine 1994). Democracy is also seen as a communication process, with a set of criteria which ensure equal access to information and agenda setting among participants (Dahl 1989). These definitions and criteria have been chosen for their generality. They apply at all levels of analysis, and thus may be used for comparison across scales. These are aligned with concepts from complexity theory as applied to human nonlinear systems and organisational dynamics, drawing mostly on Kiel(1994) and Stacey (1996). The inherent unpredictability of these systems and their capacity for complex learning and adaptive self-organisation offers an alternative paradigm to linear, hierarchical management models. A combination of literature analysis and case study evidence leads to three extensions of complexity theory as applied to human systems: Firstly, interactive technology is proposed as a technique for the potential re-pluralisation of democratic processes in complex human systems. The possibility of adding many new non-elite voices, and of making these available to all others, relates to both chaotic forms of self-organisation and the benefits of electronic democracy. Secondly, evidence is presented for the existence of fractal, or self-similar patterns in the ways information technology is applied at different levels of governance. It is shown that instrumental approaches which emphasise efficiency dominate technology use at the global, national and organisational levels, and these are interdependent. Alternative patterns which encourage participation also exist at every level. Thirdly, it is argued that the values of the dominant actors are the main determinants of whether interactive technologies will be structured to favour democratic forms of communication at every level. Three theory chapters in Part I develop these arguments by extensive reviews of relevant literatures. On the global level, convergent media, telecommunications and technology conglomerates underpin a global 'nervous system' which discourages government intervention, promotes a global monoculture, inhibits pluralistic debate by minimising access to alternative forms of information, and emphasises individualism and consumption. Within nations, widespread uniformity of public sector reform is sympathetic and responsive to these globalising pressures. Deregulation, privatisation, retreat from public broadcasting, down-sizing and outsourcing have become standard approaches, and are reflected in Australian information technology policy and programs. Several exceptions demonstrate more participatory approaches. At the organisational level, instrumental approaches to management and computerisation also prevail. In each case, a shift towards globalising values corresponds to applications of information technology which dampen the complex interactivity required for democratic policy processes. Part II supplements this analysis with case study evidence. The organisational data were collected primarily during a two year qualitative study of interactive technology use in the Australian Department of Finance and Administration. The researcher found technology use was inseparable from other change processes, and these were found to have strong elements inhibiting participation in internal policy. An instrumental approach to interactive technology use reinforced hierarchical decision processes. Three minor case studies looked at an internal mailing list in a federal agency, a mostly national list on Internet and telecommunications policy, and an experiment in electronic democracy at the local level. These offered additional insights into the ways interactive technologies can contribute to complex but adaptive policy processes, if normative democratic values guide their design. The researcher proposes a set of communication protocols for the use of interactive technologies in democratic policy processes. These would enable the forms of communication necessary to reinvigorate democracy in an information age. It is also argued that these protocols, if applied at the organisational level, and particularly within the public sector, could become part of a reaffirmation of industrial democracy. This is necessary to ensure the integrity and accountability of the public sector, given the progressive intermingling of these institutions with private enterprise. Additional suggestions are made for research into government uses of information technology as an important focus for policy analysts.

In this thesis we develop and use a continuum random walk framework to solve problems that are usually studied using a discrete random walk on a discrete lattice. Problems studied include; the time it takes for a random walker to be absorbed at a trap on a fractal lattice, the calculation of the spectral dimension for several different classes of networks, the calculation of the density of states for a multi-layered Bethe lattice and the relationship between diffusion exponents and a resistivity exponent that occur in relevant power laws. The majority of the results are obtained by deriving an expression for a Laplace transformed Green’s function or first passage time, and then using Tauberian theorems to find the relevant asymptotic behaviour. The continuum framework is established by studying the diffusion equation on a 1-d bar with non-homogeneous boundary conditions. The result is extended to model diffusion on networks through linear algebra. We derive the transformation linking the Green’s functions and first passage time results in the continuum and discrete settings. The continuum method is used in conjunction with renormalization techniques to calculate the time taken for a random walker to be absorbed at a trap on a fractal lattice and also to find the spectral dimension of new classes of networks. Although these networks can be embedded in the d- dimensional Euclidean plane, they do not have a spectral dimension equal to twice the ratio of the fractal dimension and the random walk dimension when the random walk on the network is transient. The networks therefore violate the Alexander-Orbach law. The fractal Einstein relationship (a relationship relating a diffusion exponent to a resistivity exponent) also does not hold on these networks. Through a suitable scaling argument, we derive a generalised fractal Einstein relationship which holds for our lattices and explains anomalous results concerning transport on diffusion limited aggregates and Eden trees.

Notre proposition consiste à présenter un modèle et une infrastructure d'exécution orienté composant pour le domaine des applications multi-tâches à contraintes temps réel souples et embarquées (nommé Tinap).<br /><br />Nous définissons un modèle de composant reposant sur plusieurs vues : une vue structurelle, placée au centre du cycle de conception, reposant initialement sur le modèle Fractal, une vue dynamique permettant au concepteur, dans une démarche descriptive, de personnaliser l'architecture métier pour définir les aspects de son applicatif liés à la concurrence, et enfin une vue implantation et une vue comportement fournissant respectivement une abstraction de l'implantation interne des composants et de leur comportement à l'égard de leur environnement. De plus, notre proposition est intégrée à un cadre méthodologique dirigé par les modèles.<br /><br />Nous avons également expérimenté le paradigme composant à différents niveaux d'abstraction : pour l'applicatif et pour celui de l'infrastructure d'exécution qui les implante. Enfin, au niveau du système d'exploitation fournissant les services élémentaires nécessaires. Cette démarche est motivée par la volonté d'exploiter notre modèle multi-vues canonique et de l'adapter en fonction des besoins de chaque niveau d'abstraction. Cette expérimentation est menée avec Think, une implantation en C des spécifications Fractal.<br /><br />Nous expérimentons Tinap par deux cas d'étude que nous avons prototypés. En premier lieu, pour concevoir une application d'analyse et de contrôle de flux multimédias par l'intermédiaire de disques vinyles. En second lieu pour expérimenter la mise en oeuvre du modèle d'exécution Accord (une méthodologie de conception pour applications temps-réel).

L'administration des systèmes informatiques modernes est une tâche de plus en plus complexe et coûteuse. En effet, l'architecture de ces systèmes n'est pas en géneral suffisamment modulaire pour permettre l'adaptation aux événements extérieurs (pannes, variation des performances, attaques de sécurité, etc.). Dans cette thèse, nous nous intéressons à l'administration des intergiciels qui a fait l'objet de moins de travail que l'administration des couches physiques et des systèmes d'exploitation.<br /><br />Nous proposons, d'abord, une démarche pour rendre les intergiciels patrimoniaux administrables, c'est à dire facilement configurables, déployables et adaptables à l'exécution. Cette approche consiste à transformer l'architetcure des intergiciels en une architecture explicite à base de composants. Ensuite, nous proposons un système d'administration automatisant certaines fonctions de l'administration, telles que le déploiement, la mise à jour de code et la réparation des pannes.<br /><br />Nous avons appliqué notre approche à l'administration d'un serveur d'applications J2EE patrimonial, JOnAS. JonasALaCarte est le nom du serveur J2EE administrable obtenu. Nous montrons que la configuration et le déploiement du serveur dans des environnements distribués sont simplifiés. De plus, de nouvelles propriétés de reconfiguration dynamique sont introduites comme la mise à jour dynamique de code et l'auto-réparation des pannes.

La théorie de la Relativité d'échelle peut être utilisée pour contraindre des modèles dynamiques précis de systèmes gravitationnels. Dans un premier temps, l'objet de ce travail est d'explorer les possibilités théoriques liées aux diverses solutions de l'équation de Schrödinger généralisée (qui n'est autre que la forme prise par l'équation du mouvement dans un espace-temps fractal). Ces solutions théoriques sont alors utilisées pour modéliser la dynamique de plusieurs systèmes gravitationnels stationnaires. Nous analysons en détail : (i) la morphogenèse des vents stellaires, (ii) la dynamique du Groupe Local de galaxies et (iii) la dynamique des paires de galaxies dans l'univers proche. Une attention particulière a été portée sur l'élaboration d'un catalogue de paires de galaxies à partir d'un ensemble représentatif de sous-catalogues existant. L'ensemble des résultats montre l'existence de structures dans l'espace des phases en accord avec les diverses prédictions permises par la Relativité d'échelle. Nous concluons que l'utilisation des principes de la Relativité d'échelle permet de dévoiler et ainsi de mieux comprendre la dynamique des systèmes gravitationnels.

L'approximation des objets naturels (courbes, surfaces ou images) par des formes fractales constitue aujourd'hui un important centre d'intérêt pour la recherche. Sous le nom générique de problème inverse, cette problématique intéresse en effet des domaines d'applications aussi divers que la représentation synthétique de l'information pour la transmission et la compression d'images, la reconstruction 3D pour la visualisation ou la CAO. Diverses études ont été menées pour répondre à ce problème inverse fractal. La plus connue est certainement la méthode dite de compression fractale d'images introduite par Jacquin. D'une manière générale, ces techniques existentes souffrent aujourd'hui d'un manque de souplesse en terme de contrôle sur la forme utilisée pour l'approximation. De plus, l'espace d'itération utilisé se limite à l'espace de visualisation, c'est à dire R². Des travaux antérieurs ont permis de dégager un modèle générique fractal : les formes fractales à pôles. Ce modèle, par l'intermédiaire d'une projection, permet de définir les auto-similarités d'un objet dans un espace de dimension supérieure. Nous proposons une résolution du problème inverse basée sur ce modèle et sur un algorithme de régression non-linéaire. Cette méthode générale permet d'approximer des courbes et des surfaces, aussi bien lisses que rugueuses, ainsi que des images en niveaux de gris. Une extension hiérarchique du modèle est introduite pour la modélisation d'objets hétérogènes dont les caractéristiques varient dans l'espace. Deux algorithmes sont proposés pour résoudre le problème d'approximation associé. Le premier calcule, à partir d'une image et d'un critère de distorsion, le modèle qui uniformise cette distorsion. Le deuxième décrit une méthode complète d'optimisation du codage du modèle en vue de comprimer les images à partir d'un critère de débit. Les résultats montrent l'avantage de ce type de modèle dans le domaine de la compression à faible débit.

Les réservoirs fracturés occupent une part importante de la production du pétrole dans le monde (Moyen Orient, golfe du Mexique, etc.). Un facteur clé de la récupération du pétrole dans un réservoir fracturé est la compréhension de la géométrie et de la conductivité hydraulique du réseau formé par les fractures. Cette compréhension nécessite la construction d'un modèle de réservoir intégrant l'ensemble des connaissances conceptuelles et des données disponibles sur le terrain. La thèse de Sandra Jenni soutenue en janvier 2005, a permis de poser les bases d'une méthodologie permettant d'intégrer à la fois les données statiques et les données dynamiques dans un modèle de fractures. Le présent sujet de thèse constitue donc une suite naturelle des travaux de thèse de Sandra Jenni. Il a permis d'établir une nouvelle approche de génération des réseaux de failles subsismiques se basant sur une caractérisation fractale de la géométrie des réseaux de failles identifiés à partir de données sismiques ou des affleurements. Le réseau de failles obtenu peut alors être déformé de manière à caractériser l'influence des propriétés géométriques et hydrauliques des failles sur le comportement hydrodynamique du modèle de réservoir fracturé. Cette propriété permet d'effectuer le calage à l'historique de production: les positions et les propriétés hydrauliques des failles incertaines sont modifiées par un algorithme d'optimisation, permettant de réduire l'écart avec les données dynamiques observées. La cohérence géologique du modèle de failles est préservée. La mise en œuvre des différentes étapes de l'approche proposée est illustrée par différentes applications sur réservoirs synthétiques inspirés de réservoirs exploités.

Polymer flocculants, traditionally used with primary coagulant to enhance flocculation and sedimentation, are used in the coagulation-/microfiltration process as well assuming they can improve membrane performance similarly. However, there are several uncertainties concerning the use of polymer flocculants in the coagulation-microfiltration process. First, polymer flocculants may not have measurable effect on turbidity removal, because microfiltration membranes can remove significantly smaller particles than those removed by the conventional treatment process. Second, the effect of using polymer flocculants on NOM removal has been controversial. Although a number of studies reported improved NOM removal when polymers were used, others reported no or negative impact of polymers on NOM removal. Third, polymer flocculants are high molecular weight organic compounds. When carried over to membrane residual polymers can potentially foul the membranes. Finally, the use of polymer flocculants will change floc properties (i.e. size, fractal dimension, and stickiness) and subsequently bring uncertain effect on cake layer resistance. Therefore, the role of polymer flocculants in coagulation-microfiltration system needs to be carefully assessed for system optimization. In the reported research, three types of polymer flocculants with different charge and molecular weights were tested for comprehensively evaluating the impact of polymer flocculants on the performance of coagulation-microfiltration of surface water. Operation conditions such as inline filtration, direct filtration, and filtration with sedimentation were included. Two membrane reactors were designed to study the mechanism through which polymer flocculants affect the performance of coagulation-microfiltration systems. The result demonstrated that the use of polymer flocculants provides little to no benefit to turbidity and NOM removal in most cases, but pDADMACs can enhance NOM removal if applied properly; All polymer flocculants significantly increased membrane fouling except for pDADMACs when sedimentation proceeds MF; Polymer flocculants increase deposition/attachment of floc particles on the membrane surface through both adsorption of residual polymer on the membrane surface and polymer molecules on the floc particle surface; Even though polymers form larger and more fractal floc particles, they did not have notable impact on cake layer structure.