Only in recent times has the true potential of cellulose as a high-end functional and sustainable material been realised. As the world’s most abundant resource cellulose has been utilised by man throughout history for timber, paper and yarns. It is found in every plant as a hierarchical material and can be extracted and converted into fibres which are of great use, especially in the form of nanofibrous materials. This thesis has focused on the utilisation and ability of cellulose nanocrystals (CNCs) to generate structural colour in fabricated thin films. This has been achieved in two ways: Firstly, the natural morphology of CNCs and their ability to form a suspension have been applied to a layer-by-layer (LbL) regime to produce tunable Bragg reflecting thin films. Secondly, a novel technique combining profilometry and spectroscopy has been developed to estimate the distribution of CNCs within EISA thin films and correlate this with the optical properties of the film. This thesis reports the successful fabrication of synthetic CNC LbL Bragg reflecting thin films. The film was compiled using an additive layer-by-layer technique which allowed the construction of a multi-layered thin film and control over individual layer thicknesses and refractive index. Also reported is the discovered reflection of both left and right handed circularly polarised light (CPL) from CNC EISA thin films. These reflections were found to correlate with CNC distribution within the thin films. The distribution of CNCs was estimated using a novel technique which combined spectroscopically measured film absorbance as a function of the volume of the film area under investigation. The specific volumes were calculated using profilometry measurements and the beam spot size used in the spectroscopy measurements.

La croissance des unités de distribution des ressources énergétiques ainsi que l'intégration des nouvelles technologies dans la production et le stockage d'énergie, ont imposé un contrôle nouveau et de nouvelles stratégies opérationnelles. Grâce à sa capacité de stockage et étant considérée comme une énergie propre; la pile à combustible (Pac) est l'une des technologies les plus prometteuse en tant que source d'énergie stationnaire dans les réseaux micro et aussi dans les applications de transport. Par conséquent, deux sujets principaux sont abordés dans cet ouvrage, la conception et l'installation d'un banc d'essai complet instrumenté a membrane échangeuse de polymère (PEM) Pac et de conception et l'essai expérimental d'une nouvelle stratégie de contrôle dynamique d'échange de l'énergie pour les systèmes multi - source et multi - charge. Pour définir le test instruments banc exigences, un examen complet de modèle dynamique est donné dans la première partie. Dans la prochaine section seront inclues, les renseignements concernant la configuration de la conception et la mise en œuvre de banc d'essai de Pac, i.e., critères de spécification des instruments, acquisition, et affichage des données du système. Des résultats expérimentaux sont réalisés afin de démontrer les potentialités de l'installation. Dans le chapitre suivant, une nouvelle stratégie de contrôle dynamique de l'énergie d'échange (DSER) sera introduite et testée par simulation et expérimentalement dans un système à deux ports. Afin d'établir une comparaison et d'intégrer la DSER dans une application Pac, un système à trois ports - y compris un modèle statique de Pac - et deux différentes approches de contrôle, seront testés par simulation dans le cinquième chapitre. La thèse s’achèvera par quelques conclusions et quelques thèmes de recherche potentiels générés à partir de ce travail. / The growth of distributed energy resources together with the incorporation of new technologies in the generation and storage of energy are imposing new control and operational strategies. Due to its storage capability and that it is considered to be clean energy; fuel cell (FC) is one of the most promissory technologies as a stationary energy source in micro grids and also in transportation applications. Therefore, two main issues are addressed in this work; the conception, design, and setup of a fully instrumented test bench for proton exchange membrane (PEM) FC stacks and the design and experimental test of a new dynamic energy-exchange control strategy for multi source and multi load systems. To define the test bench instrument requirements, in the first part a complete dynamic model review is given. In the next section, relevant information regarding the setup of the FC test bench design and implementation is included, i.e., specification criteria of the instruments and acquisition and data display system. Some experimental results are performed in order to demonstrate the potentialities of the setup. In the following chapter, a new dynamic energy exchange control strategy (DSER) is introduced and tested in a two port system via simulation and experimentation. In order to establish a comparison and integrate the DSER in a FC application, in the fifth chapter a three port system – including a static model of FC – and two different control approaches, are tested via simulation. The thesis is closed with some concluding remarks and some potential research topics generated from this work.

Atualmente, pode-se observar que a Lei de Moore vem estagnando. A freqüência de operação já não cresce da mesma forma, e a potência consumida aumenta drasticamente em processadores de propósito geral. Ao mesmo tempo, sistemas embarcados vêm se tornando cada vez mais heterogêneos, caracterizados por uma grande quantidade de modelos computacionais diferentes, sendo executados em um mesmo dispositivo. Desta maneira, como novas tecnologias que irão substituir totalmente ou parcialmente o silício estão surgindo, novas soluções arquiteturais são necessárias. Apesar de sistemas reconfiguráveis já terem demonstrado serem candidatos em potencial para os problemas supracitados, ganhos significativos de desempenho são alcançados apenas em programas que manipulam dados massivamente, não representando a realidade dos sistemas atuais. Ademais, o seu uso em alta escala ainda está limitado à utilização de ferramentas ou compiladores que, claramente, não mantêm a compatibilidade de software e a reutilização do código binário já existente. Baseando-se nestes fatos, a presente tese propõe uma nova técnica para, utilizando um sistema reconfigurável, otimizar tanto programas orientados a dados como aqueles orientados a controle, sem a necessidade de modificação do código fonte ou binário. Para isto, um algoritmo de Tradução Binária, que trabalha em paralelo ao processador, foi desenvolvido. O mecanismo proposto é responsável pela transformação de seqüências de instruções, em tempo de execução, para serem executadas em uma unidade funcional reconfigurável de granularidade grossa, suportando execução especulativa. Desta maneira, é possível aproveitar as vantagens do uso da lógica combinacional para aumentar o desempenho e reduzir o gasto de energia, mantendo a compatibilidade binária em um processo totalmente transparente. Três diferentes estudos de caso foram feitos: os processadores Java e MIPS R3000 – representando o campo de sistemas embarcados – e o conjunto de ferramentas Simplescalar, que simula um processador superescalar baseado no MIPS R10000 – representando o mercado de processadores de propósito geral. / As Moore’s law is losing steam, one already sees the phenomenon of clock frequency reduction caused by the excessive power dissipation in general purpose processors. At the same time, embedded systems are getting more heterogeneous, characterized by a high diversity of computational models coexisting in a single device. Therefore, as innovative technologies that will completely or partially replace silicon are arising, new architectural alternatives are necessary. Although reconfigurable computing has already shown to be a potential solution for such problems, significant speedups are achieved just in very specific dataflow oriented software, not representing the reality of nowadays systems. Moreover, its wide spread use is still withheld by the need of special tools and compilers, which clearly preclude software portability and reuse of legacy code. Based on all these facts, this thesis presents a new technique using reconfigurable systems to optimize both control and dataflow oriented software without the need of any modification in the source or binary codes. For that, a Binary Translation algorithm has been developed, which works in parallel to the processor. The proposed mechanism is responsible for transforming sequences of instructions at runtime to be executed on a dynamic coarse-grain reconfigurable array, supporting speculative execution. This way, it is possible to take advantage of using pure combinational logic to speed up the execution, maintaining full binary compatibility in a totally transparent process. Three different case studies were evaluated: a Java Processor and a MIPS R3000 – representing the embedded systems field – and the Simplescalar Toolset, a widely used toolset that simulates a superscalar architecture based on the MIPS R10000 processor – representing the general-purpose market.

Student Number : 0100677A - PhD thesis - School of Civil and Environmental Engineering - Faculty of Engineering and the Built Environment / This thesis presents the findings on empirical study of dry-stack masonry. Dry-stack masonry refers to a method of building masonry walls, where most of the masonry units are laid without mortar in the joints. Of late (since mid eighties) in modern construction, dry-stacking or mortarless technology is increasingly becoming popular because of its advantages. The construction industry is acknowledging the need to accelerate the masonry construction process, as the traditional method is labour intensive and hence slower due to the presence of a large number of mortar joints. Early attempts were made to increase the size of masonry units (block instead of brick), thereby reducing the number of mortar joints, wherein the use of bedding mortar imposed constraints on the number of courses to be constructed in a day. Elimination of bedding mortar accelerates construction; thereby reducing cost, variation due to workmanship and generally small pool of skilled labour is required in dry stacking. Dry-stack masonry is a relatively new technology not yet regulated in the code of practice and therefore very limited information on the structural behaviour of the masonry is available. This project is based on the investigation of the HYDRAFORM dry-stack system, which utilises compressed soil-cement interlocking, blocks. The system is now widely used in Africa, Asia and South America. The main objective of the project was to establish through physical testing the capacity of the system to resist lateral load (e.g. wind load), vertical load and dynamic load such as earthquake loading. In the first phase of the project investigations were conducted under static loading where series of full-scale wall panels were constructed in the laboratory and tested under lateral loading, and others were tested under vertical loading to establish the mode of failure and load capacity of the system. Series of control tests were also conducted by testing series of wallettes to establish failure mechanism of the units and to establish the flexural strength of the system. Finally the test results were used for modelling, where load prediction models for the system under vertical loading and under lateral loading were developed. The theoretical load prediction models were tested against the test results and show good agreement. After the load capacity was established the next step in the study was to further improve the system for increased capacity particularly under dynamic loading. The normal Hydraform system was modified by introducing a conduit, which allows introduction of reinforcements. Series of dry-stack seismic systems were constructed and initially tested under static lateral loading to establish the lateral load capacity. The second Phase of the project was to investigate the structural behaviour and performance of the Hydraform system under seismic loading. A shaking table of 20 tonnes payload, (4m x 4m) in plan was designed and fabricated. A full-scale plain dry-stack masonry house was constructed on the shaking table and subjected to seismic base motions. The shaking table test was performed using sine wave signals excitations starting from low to very severe intensity. A conventional masonry test structure of similar parameters was also constructed on the table and tested in a similar manner for comparison. The tests were conducted using a frequency range of 1Hz to 12Hz and the specimens were monitored for peak accelerations and displacements. For both specimens the initial base motion was 0.05g. The study established the mode of failure of the system; the structural weak points of unreinforced dry-stack masonry, the general structural response of the system under seismic condition and the failure load. The plain dry-stack masonry failed at 0.3g and the conventional masonry failed at 0.6g. Finally recommendations for further strengthening of system to improve its lateral capacity were proposed.

Interlocking compressed earth blocks (ICEBs) are cement stabilized soil blocks that allow for dry stacked construction. The incomplete understanding of the inelastic performance of ICEB building systems limits widespread acceptance of this structural system in earthquake prone areas. This thesis presents results from an experimental program designed to explore the behavior of ICEB walls, built according to current design practice in Indonesia and Thailand, and subjected to out-of-plane loading. A total of five reinforced and grouted ICEB walls were constructed and tested. Results from experimentation show the current masonry design code, ACI 530, adequately predicts the yield strength of these walls. However, ACI 530 grossly over-predicts the ICEB wall stiffness. All tests showed flexural behavior and failure, except for one wall. A brittle failure was observed in one wall before reaching the predicted flexural strength, prompting a suggested maximum shear tie spacing. The testing results provide useful data for developing analytical models that predicts the seismic behavior of ICEB walls under out-of-plane loading. A moment-curvature relationship was developed that accurately predicts the behavior of these walls in the elastic range as well as the inelastic range. By comparing the data provided by two walls of similar sizes, one including a pilaster and one without a pilaster, insight into stiffener elements was gained. Analysis of these two walls provides a limit on the length and height of ICEB walls without stiffener elements to prevent significant structural damage during a seismic event. In all, conclusions based on experimental data from ICEB out-of-plane loading tests are aimed to provide suggestions for ICEB construction in areas of high-seismicity.

Communication complexity is an area of complexity theory that studies an abstract model of computation called a communication protocol. In a $k$-player communication protocol, an input to a known function is partitioned into $k$ pieces of $n$ bits each, and each piece is assigned to one of the players in the protocol. The goal of the players is to evaluate the function on the distributed input by using as little communication as possible. In a Number-On-Forehead (NOF) protocol, the input piece assigned to each player is metaphorically placed on that player's forehead, so that each player sees everyone else's input but its own. In a Number-In-Hand (NIH) protocol, the piece assigned to each player is seen only by that player. Overall, the study of communication protocols has been used to obtain lower bounds and impossibility results for a wide variety of other models of computation. Two of the main contributions presented in this thesis are negative results on the NOF model of communication, identifying limitations of NOF protocols. Together, these results consitute stepping stones towards a better fundamental understanding of this model. As the first contribution, we show that randomized NOF protocols are exponentially more powerful than deterministic NOF protocols, as long as $k \le n^c$ for some constant $c$. As the second contribution, we show that nondeterministic NOF protocols are exponentially more powerful than randomized NOF protocols, as long as $k \le \delta \cdot \log n$ for some constant $\delta < 1$. For the third major contribution, we turn to the NIH model and we present a positive result. Informally, we show that a NIH communication protocol for a function $f$ can simulate a Stack Machine (a Turing Machine augmented with a stack) for a related function $F$, consisting of several instances of $f$ bundled together. Using this simulation and known communication complexity lower bounds, we obtain the first known (space vs. number of passes) trade-off lower bounds for Stack Machines.

Relatively few studies have been carried out on the turbulent wake structure of a finite circular cylinder and a stack partially immersed in a flat-plate turbulent boundary layer. There is a need to develop a better understanding of the wakes of these structures, since they have many important engineering applications. This thesis investigates the influence of the aspect ratio on the wake of a finite circular cylinder and the effects of the ratio of jet flow velocity to crossflow velocity (velocity ratio, R) on the wake of a stack in a cross-flow. <p>The wake characteristics of flows over a finite circular cylinder at four different aspect ratios (AR = 3, 5, 7 and 9) were investigated experimentally at a Reynolds number of ReD = 6104 using two-component thermal anemometry. Each cylinder was mounted normal to a ground plane and was either completely or partially immersed in a flat-plate turbulent boundary layer. The ratio of boundary layer thickness to the cylinder diameter was 3. <p>A similar turbulent wake structure (time-averaged velocity, turbulence intensity, and Reynolds shear stress distributions) was found for the cylinders with AR = 5, 7, and 9, while a distinctly different turbulent wake structure was found for the cylinder with AR = 3. This was consistent with the results of a previous study that focused on the time-averaged streamwise vortex structures in the wake. In addition, irrespective of the value of AR, high values were observed for the skewness and flatness factors around the free end of the cylinders, which may be attributed to the interaction of the tip vortex structures and downwash flow that dominates this region of the cylinder.<p>The wake characteristics of a stack of aspect ratio AR = 9 were investigated using both the seven-hole pressure probe and thermal anemometry. The seven-hole probe was used to measure the three components of the time-averaged velocity field, while the thermal anemometry was used to measure two components of the turbulent velocity field at various downstream locations from the stack. The stack was mounted normal to the ground plane and was partially immersed in a flat-plate turbulent boundary layer, for which the ratio of boundary layer thickness to the stack diameter was 4.5. In addition, measurements of the vortex shedding frequency were made with a single-component hot-wire probe. The cross-flow Reynolds number was ReD = 2.3 x 104, the jet Reynolds number ranged from Red = 7.6 x 103 to 4.7 x 104, and R was varied from 0 to 3. <p>In the stack study, three flow regimes were identified depending on the value of R: the downwash (R < 0.7), cross-wind-dominated (0.7 < R < 1.5), and jet-dominated (R ≥ 1.5) flow regimes. Each flow regime had a distinct structure for the time-averaged velocity and streamwise vorticity fields, and turbulence characteristics, as well as the variation of the Strouhal number and the power spectrum of the streamwise velocity fluctuations along the stack height. The turbulence structure is complex and changes in the streamwise and wall-normal directions within the near and intermediate stack and jet wakes. In the downwash and crosswind-dominated flow regimes, two pairs of counter-rotating streamwise vortex structures were identified within the stack wake. The tip-vortex pair and base-vortex pair were similar to those found in the wake of a finite circular cylinder, located close to the free end and the base of the stack (ground plane), respectively. In the jet-dominated flow regime, a third pair of streamwise vortex structures was observed, referred to as the jet-wake vortex pair, which occurred within the jet-wake region above the free end of the stack. The jet-wake vortex pair has the same orientation as the base vortex pair and is associated with the jet rise.

Dans l'objectif d'une augmentation en puissance des piles à combustible pour satisfaire les besoins énergétiques des applications embarquées, une solution consiste à augmenter la taille des assemblages. Dès lors, des problèmes de disparités fluidique, thermique et électrique peuvent survenir dans le cœur des piles et conduire ainsi à l'apparition de défaut. La pile à combustible, de par sa nature de source électrique basse tension - fort courant, requiert d'être couplée au réseau électrique embarqué par l'intermédiaire d'un convertisseur statique. Ce dernier peut alors être employé pour agir de façon corrective sur la pile et aussi de corriger les défaillances qui en sont liées. Dans cette perspective, le convertisseur doit avoir en permanence un retour sur l'état de santé de la pile. Pour cela, une méthode de détection et d'identification de défaut de type noyage et d'assèchement pour une pile du type PEMFC a été approfondie. Cette méthode simple, économique en capteurs, se base sur la mesure de 3 tensions de cellule judicieusement sélectionnées et localisées sur la pile. Ainsi, l'utilisation de l'information " spatiale ", qui correspond à la position de la mesure de tension dans la pile permet d'identifier les défauts. Le principe de la détection localisée nous amène alors à considérer le concept de pile segmentée qui consiste à séparer électriquement la pile en 3 parties de façon à ce que des convertisseurs associés puissent agir électriquement sur chaque segment. L'action peut être du type tout ou rien, ou contrôlée. Cette dernière offre davantage de degrés de liberté, et est moins contraignante pour la pile d'un point de vue électrique. Pour choisir comment réaliser cette action, une étude comparative de plusieurs topologies de convertisseur est effectuée. Les structures alimentées en courant répondent au mieux aux contraintes électriques d'une PEMFC et sont donc privilégiées, de même que la nécessité d'une isolation galvanique imposée par la segmentation de la pile. Au final, une topologie de BOOST isolé résonant est apparue comme étant la topologie répondant au mieux à l'ensemble des critères (plage de fonctionnement, performances énergétiques, nombre de composants). L'ensemble convertisseur global intègre ainsi trois structures unitaires qui permettent d'offrir une modularité, une action indépendante sur chaque segment et de garantir une disponibilité du système grâce à un fonctionnement dégradé. Pour cela, la stratégie de commande de l'ensemble convertisseur intègre les informations issues de la méthode de détection. La thèse se termine avec le dimensionnement complet d'un pré-prototype du convertisseur avec le choix des composants actif et passifs, et du système de refroidissement associé.

Communication complexity is an area of complexity theory that studies an abstract model of computation called a communication protocol. In a $k$-player communication protocol, an input to a known function is partitioned into $k$ pieces of $n$ bits each, and each piece is assigned to one of the players in the protocol. The goal of the players is to evaluate the function on the distributed input by using as little communication as possible. In a Number-On-Forehead (NOF) protocol, the input piece assigned to each player is metaphorically placed on that player's forehead, so that each player sees everyone else's input but its own. In a Number-In-Hand (NIH) protocol, the piece assigned to each player is seen only by that player. Overall, the study of communication protocols has been used to obtain lower bounds and impossibility results for a wide variety of other models of computation. Two of the main contributions presented in this thesis are negative results on the NOF model of communication, identifying limitations of NOF protocols. Together, these results consitute stepping stones towards a better fundamental understanding of this model. As the first contribution, we show that randomized NOF protocols are exponentially more powerful than deterministic NOF protocols, as long as $k \le n^c$ for some constant $c$. As the second contribution, we show that nondeterministic NOF protocols are exponentially more powerful than randomized NOF protocols, as long as $k \le \delta \cdot \log n$ for some constant $\delta < 1$. For the third major contribution, we turn to the NIH model and we present a positive result. Informally, we show that a NIH communication protocol for a function $f$ can simulate a Stack Machine (a Turing Machine augmented with a stack) for a related function $F$, consisting of several instances of $f$ bundled together. Using this simulation and known communication complexity lower bounds, we obtain the first known (space vs. number of passes) trade-off lower bounds for Stack Machines.

Optical absorption spectroscopy has been applied to study properties such as the fundamental absorption edge and defect absorption centers of group III-nitride compound semiconductor epilayers. The investigation in this thesis focused on analyzing the band gap of indium-rich In1-xGaxN epilayers, which where grown by the high-pressure chemical vapor deposition (HPCVD) technique. Our results - together with literature data for gallium-rich In1-xGaxN alloys indicate that the shift of the fundamental band gap of In1-xGaxN with composition x can be described with a bowing parameter of b = 2.2eV. Temperature dependent transmission measurements show that the band gap variation with temperature follows a S-shape behavior for small gallium concentration and shifts towards a Varshni type behavior for a higher gallium concentrations. The S-shape behavior is attributed to nanoscale compositional fluctuations/clustering in the ternary alloy system. The thicknesses of the measured In1-xGaxN epilayers have been analyzed through multilayer stack model calculations of the transmission spectra. The free electron concentration in the In1-xGaxN epilayers has been obtained from simulations of infrared reflectance spectra.