SpiNNaker is an unconventional supercomputer architecture designed to simulate up to one billion biologically realistic neurons in real-time. To achieve this goal, SpiNNaker employs a novel network architecture which poses a number of practical problems in scaling up from desktop prototypes to machine room filling installations. SpiNNaker's hexagonal torus network topology has received mostly theoretical treatment in the literature. This thesis tackles some of the challenges encountered when building `real-world' systems. Firstly, a scheme is devised for physically laying out hexagonal torus topologies in machine rooms which avoids long cables; this is demonstrated on a half-million core SpiNNaker prototype. Secondly, to improve the performance of existing routing algorithms, a more efficient process is proposed for finding (logically) short paths through hexagonal torus topologies. This is complemented by a formula which provides routing algorithms with greater flexibility when finding paths, potentially resulting in a more balanced network utilisation. The scale of SpiNNaker's network and the models intended for it also present their own challenges. Placement and routing algorithms are developed which assign processes to nodes and generate paths through SpiNNaker's network. These algorithms minimise congestion and tolerate network faults. The proposed placement algorithm is inspired by techniques used in chip design and is shown to enable larger applications to run on SpiNNaker than the previous state-of-the-art. Likewise the routing algorithm developed is able to tolerate network faults, inevitably present in large-scale systems, with little performance overhead.

Variable-speed Induction motor drives are nowadays used for various kinds of industrial processes, transportation systems, wind turbines and household appliances in the world. The majority of drives are for general purpose speed control applications where accurate speed control is not required for entire speed range. But for high dynamic drive application, very precise and fast control of induction motor drive is essential. For such applications, sophisticated and well-performing control design is a key issue. Precise and accurate torque control of the Induction Motor (IM) can only be accomplished by vector control and direct torque control. In terms of space vector theory, vector control implies that the instantaneous torque is controlled by way of the stator current vector that is orthogonal to the rotor flux vector. Precise knowledge of the rotor flux angle is therefore essential for a vector controlled IM. IMs do not allow the flux position to be easily measured, so most modern vector controlled IM drives rely on flux estimation. This means that the flux angle is derived from a flux estimator, using the dynamic model of the IM. Given that the rotor speed of the IM is measured by a mechanical shaft sensor. Flux estimation is a fairly easy task. However, vector control of IM without mechanical shaft speed sensor is of current interest in industrial environment. The driving motivations behind the development in sensorless control are lower cost, improved reliability and operating environment. In this thesis, a sensorless vector control scheme for rotor flux estimation using current error space phasor based hysteresis controller is proposed including the method for estimation of leakage inductance, Ls. For frequencies of operation less than 25 Hz, the rotor voltage and hence the rotor flux position is computed during the inverter zero voltage space vector using steady state model of IM. For above 25 Hz, active vector period and steady state model of IM is used. The whole rotor flux estimation scheme is dependent on current error space phasor and the steady state motor model, with rotor flux as a reference vector. Since no terminal voltage sensing is involved, dead time effects will not create problem in rotor flux sensing at low frequencies of operation. But appropriate device on-state drop are compensated at low frequencies (below 5 Hz) of operation to achieve a steady state operation up to less than 1 Hz. A constant switching frequency hysteresis current controller is used in inner current control loop for the PWM regulation, with smooth transition of operation to six-step mode operation. A simple Ls estimation based on current error space phasor is also proposed to nullify the deteriorating effect on rotor flux estimation. The parameter sensitivity of the control scheme to changes in the stator resistance Rs is also investigated. The drive scheme is tested up to a low frequency operation less than 1 Hz. The extensive simulation and experiment results are presented to show the proposed scheme’s good dynamic performance extending up to six-step operation. In contrast to vector control, direct torque control (DTC) method requires the knowledge of stator resistance only and thereby decreasing the associated sensitivity to parameters variation and the elimination of speed information. DTC as compared to vector control does not require co-ordinate transformation and PI controller. DTC is easy to implement because it needs only two hysteresis comparators and a lookup table for both flux and torque control. This thesis also investigates the possibilities in improvement of direct torque control scheme for high performance induction motor drive applications. Here, two schemes are proposed based on the direct torque control scheme for IM drive using 12-sided polygonal voltage space vectors for fast torque control. The torque control scheme based on DTC algorithm is proposed using 12-sided polygonal voltage space vector. The basic DTC scheme is used to control the torque. But the IM drive is open-end type. For torque control, the voltage space vectors orthogonal to stator flux vector in 12-sided polygonal space vector structure are used as hexagonal space vector based DTC scheme. The advantages achieved due to 12-sided polygonal space vector are mainly fast torque control and small torque ripple. The fast transient of torque with precise control is achieved using voltage space vector placed with a resolution of ±15. The torque ripple will be less as 6n±1 (n=odd) harmonic torque is totally eliminated from the whole range of PWM modulation. The comparative analysis of proposed 12-sided polygonal voltage space vector based DTC and conventional hexagonal space vector based DTC is also presented. Extensive simulation and experiment results are also presented to show the fast torque control at speeds of operation ranging from 5 Hz to the rated speed. The concept of 12-sided polygonal space vector based DTC is further extended for a variable speed control scheme using estimated fundamental stator voltage for sector identification. The conventional DTC scheme uses stator flux vector for identification of the sector and the switching vector are selected based on this sector information to control stator flux and torque. However, the proposed DTC scheme selects switching vectors based on the sector information of the estimated fundamental stator voltage vector and its relative position with respect to the stator flux vector. The fundamental stator voltage estimation is based on the steady state model of IM and information of synchronous frequency which is derived from computed stator flux using a low pass filter technique. The proposed DTC scheme utilizes the exact position of fundamental stator voltage vector and stator flux vector position to select optimal switching vector for fast control of torque with small variation of stator flux within hysteresis band. The present DTC scheme allows the full load torque control with fast transient response to very low speeds of operation below 5 Hz. The extensive simulation and experiment results are presented to show the fast torque control for speed of operation from zero speed to rated speed. However, the present scheme will have all the advantages of DTC scheme using stator flux vector for sector identification. All the above propositions are first simulated by MATLAB/Simulink and subsequently verified by an experimental laboratory prototype. The proposed control schemes are experimentally verified on a 3.7 kW IM drive. The control algorithms of the sensorless vector control using current error space phasor as well as DTC using 12-sided polygonal voltage space vector are completely implemented on a TI TMS320LF2812 DSP controller platform. These are some of the constituents for chapters 2, 3 and 4 in this thesis. Additionally, the first chapter also covers a brief survey on some of the recent progresses made in the field of sensorless vector control, direct torque control and current hysteresis controller. The thesis concludes with suggestion for further exploration.

Ziel der Diplomarbeit ist es, eine Dispersion aus Monomer und Silika-Partikeln auf ein festes Substrat aufzutragen, die so erhaltenen Versuchsträger mit dem Monomer/Partikel-Film (Modell Monomer/Partikel/festes Substrat) anschließend mit dem Rasterkraftmikroskop abzubilden und anschließend hinsichtlich der Oberflächenkrümmung und dem Kontaktwinkel zu untersuchen. Durch die Kontaktwinkelmessung am Dreiphasenpunkt soll eine erste Charakterisierung der kapillaren Oberfläche erfolgen, die jedoch nur ein Teilschritt bis zur vollständigen Optimierung von durch partikel-assistierte Benetzung hergestellte poröse Membranen ist.

В работе описаны процессы пластической деформации для наиболее распространённых металлов с ГПУ кристаллической решеткой, к числу которых относится титан, магний, цирконий. Рассматривается характеристика ГПУ металлов с точки зрения особенностей кристаллического строения и действующих механизмов пластической деформации, описана область применения. Описано поведение металлов с ГПУ решеткой при пластической деформации и при повышенных температурах, что накладывает требования к технологии производства изделий из данных металлов. Представлены основные условия текучести, применяемые для ортотропных металлов, отражающие различное поведение материала в разных направлениях. Методом конечных элементов выполнены расчеты и представлено сравнение параметров плоской прокатки циркониевого сплава Э125 при применении изотропного и ортотропного условий текучести. / The paper describes the processes of plastic deformation for the most common metals with an HCP crystal lattice, which include titanium, magnesium, and zirconium. The characteristics of HCP metals are considered from the point of view of the features of the crystal structure and the operating mechanisms of plastic deformation, and the scope is described. The behavior of metals with an HCP lattice under plastic deformation and at elevated temperatures is described, which imposes requirements on the technology for the production of products from these metals. The basic flow conditions used for orthotropic metals are presented, reflecting the different behavior of the material in different directions. The finite element method is used to perform calculations and compare the parameters of flat rolling of the E125 zirconium alloy using isotropic and orthotropic yield conditions.

No description available.

The mechanisms that order cellular packing geometry are critical for the functioning of many tissues, but are poorly understood. Here we investigate this problem in the developing wing of Drosophila. The surface of the wing is decorated by hexagonally packed hairs that are uniformly oriented towards the distal wing tip. They are constructed by a hexagonal array of wing epithelial cells. We find that wing epithelial cells are irregularly arranged throughout most of development but become hexagonally packed shortly before hair formation. During the process, individual cell junctions grow and shrink, resulting in local neighbor exchanges. These dynamic changes mediate hexagonal packing and require the efficient delivery of E-cadherin to remodeling junctions; a process that depends on both the large GTPase Dynamin and the function of Rab11 recycling endosomes. We suggest that E-cadherin is actively internalized and recycled as wing epithelial cells pack into a regular hexagonal array. Hexagonal packing furthermore depends on the activity of the Planar Cell Polarity proteins. The Planar Cell Polarity group of proteins coordinates complex and polarized cell behavior in many contexts. No common cell biological mechanism has yet been identified to explain their functions in different tissues. A genetic interaction between Dynamin and the Planar Cell Polarity mutants suggests that the planar cell polarity proteins may modulate Dynamin-dependent trafficking of E-cadherin to enable the dynamic remodeling of junctions. We furthermore show that the Planar Cell Polarity protein Flamingo can recruit the exocyst component Sec5. Sec5 vesicles also co-localizes with E-cadherin and Flamingo. Based on these observations we propose that during the hexagonal repacking of the wing epithelium these proteins polarize the trafficking of E-cadherin-containing exocyst vesicles to remodeling junctions. The work presented in this thesis shows that one of the basic cellular functions of planar cell polarity signaling may be the regulation of dynamic cell adhesion. In doing so, the planar cell polarity pathway mediates the acquisition of a regular packing geometry of Drosophila wing epithelial cells. We identify polarized exocyst-dependent membrane traffic as the first basic cellular mechanism that can explain the role of PCP proteins in different developmental systems.

Une étude expérimentale sur la structure et la dynamique d'un complexe hydraté de fragments d'ADN (150 pb) et des phases lamellaires de lipides non-cationiques est présentée. Par la variation de d'hydratation, il est possible de contrôler le confinement imposé par cette matrice hôte sur les nucléotides insérés dans les couches aqueuses. L’organisation supramoléculaire du complexe est suivie par diffraction des rayons-X et des techniques de microscopie optique et électronique. Un riche polymorphisme de mésophases est observé en fonction du confinement. Dans le régime plus hydraté, les fragments se distribuent selon une orientation nématique. Dans la mesure où la quantité de l'eau diminue, le confinement des bicouches sur les nucléotides monte et des corrélations trans-membranaires donnent origine à des phases hautement organisées, avec de symétries rectangulaires et hexagonales (2D) d'ADN dans la phase lipidique. L'incorporation totale des nucléotides par la phase lamellaire est observée uniquement lorsque des grandes quantités d'ADN y sont présentes. Ce fait souligne une importance majeur des interactions de volume exclu. Une analyse du paramètre de Caillé montre que l'insertion des fragments diminue les fluctuations des membranes. À partir des ces observations, il est suggéré que la modification des interactions stériques entre des lamelles, associée à des effets interfaciaux ADN-membranes, est un mécanisme important dans le comportement de phases. Les propriétés dynamiques sont étudiés avec la technique de retour de fluorescence après photo-blanchiment (FRAP). Un modèle développé récemment pour l'analyse de diffusion anisotrope est testé avec succès, démontrant une corrélation proche entre structure et dynamique. / An experimental study on the structural and dynamical properties of a hydrated DNA-non-cationic complex is presented. By varying the water amount, it is possible to control the confinement imposed by this host matrix over the organization of the nucleotides inserted within the water layers. The supramolecular assembly is investigated by X-rays diffraction and techniques involving both optical and electron microscopy. A rich polymorphism of mesophases is observed in function of confinement. In the more hydrated regime, the fragments are distributed according to nematic orientation in-between lamellae. As the water amount decreases, the confinement of bilayers over the particles increases and transmembrane correlations appear, giving raise to highly-ordered phases, with 2D-rectangular and -hexagonal symmetries of DNA embodied in the lamellar phase. The full incorporation of nucleotides by the lamellar phase is observed only in the presence of large amounts of DNA. This finding points to major importance of excluded volume interactions. An analysis of the Caillé parameter shows that the insertion of DNA reduces the fluctuations of membranes. From these observations, it is suggested that changes in the interactions between bilayers, together with the appearance of interfacial effects between DNA and membranes, are a mechanism relevant for the phase behavior of these systems. The dynamical properties of nucleotides are investigated through the fluorescence recovery after photobleach (FRAP). A model recently developed for analyses of anisotropic diffusion is sucessfully tested, demonstrating a close relationship between structure and dynamics. / Um estudo experimental sobre aspectos estruturais e dinâmicos de um complexo hidratado de fragmentos de DNA (150 pb) e fases lamelares de lipídios não-catiônicos é apresentado. Variando-se a hidratação, é possível controlar o confinamento imposto por essa matriz hospedeira sobre os nucleotídeos inseridos na camada aquosa. O arranjo supramolecular do complexo é investigado por difração de raios X e técnicas de microscopia óptica e eletrônica. Um rico polimorfismo de mesofases é observado em função do confinamento. No regime mais hidratado, os fragmentos se distribuem segundo uma orientação nemática entre as membranas. À medida que a quantidade de água diminui, o confinamento das bicamadas sobre os nucleotídeos aumenta e correlações transmembranares aparecem, dando origem a fases altamente organizadas, com simetrias retangulares e hexagonais 2D de DNA entre as lamelas. A incorporação completa de nucleotídeos é observada apenas quando grandes quantidades de DNA estão presentes. Esse fato aponta para importância maior de interações de volume excluído. Uma análise do parâmetro de Caillé mostra que as flutuações das membranas diminuem com a inserção de DNA. A partir dessas observações, é sugerido que a alteração das interações entre membranas, aliada à aparição de efeitos interfaciais entre DNA e membranas, é um mecanismo relevante no comportamento de fase. As propriedades dinâmicas dos nucleotídeos são investigadas através da técnica de FRAP (fluorescence recovery after photobleaching). Um modelo recentemente desenvolvido para análise de difusão anisotrópica é testado com sucesso, demonstrando estreita correlação entre estrutura e dinâmica.

The mechanisms that order cellular packing geometry are critical for the functioning of many tissues, but are poorly understood. Here we investigate this problem in the developing wing of Drosophila. The surface of the wing is decorated by hexagonally packed hairs that are uniformly oriented towards the distal wing tip. They are constructed by a hexagonal array of wing epithelial cells. We find that wing epithelial cells are irregularly arranged throughout most of development but become hexagonally packed shortly before hair formation. During the process, individual cell junctions grow and shrink, resulting in local neighbor exchanges. These dynamic changes mediate hexagonal packing and require the efficient delivery of E-cadherin to remodeling junctions; a process that depends on both the large GTPase Dynamin and the function of Rab11 recycling endosomes. We suggest that E-cadherin is actively internalized and recycled as wing epithelial cells pack into a regular hexagonal array. Hexagonal packing furthermore depends on the activity of the Planar Cell Polarity proteins. The Planar Cell Polarity group of proteins coordinates complex and polarized cell behavior in many contexts. No common cell biological mechanism has yet been identified to explain their functions in different tissues. A genetic interaction between Dynamin and the Planar Cell Polarity mutants suggests that the planar cell polarity proteins may modulate Dynamin-dependent trafficking of E-cadherin to enable the dynamic remodeling of junctions. We furthermore show that the Planar Cell Polarity protein Flamingo can recruit the exocyst component Sec5. Sec5 vesicles also co-localizes with E-cadherin and Flamingo. Based on these observations we propose that during the hexagonal repacking of the wing epithelium these proteins polarize the trafficking of E-cadherin-containing exocyst vesicles to remodeling junctions. The work presented in this thesis shows that one of the basic cellular functions of planar cell polarity signaling may be the regulation of dynamic cell adhesion. In doing so, the planar cell polarity pathway mediates the acquisition of a regular packing geometry of Drosophila wing epithelial cells. We identify polarized exocyst-dependent membrane traffic as the first basic cellular mechanism that can explain the role of PCP proteins in different developmental systems.

A wearable body temperature sensor would allow for early detection of fever or infection, as well as frequent and accurate hassle-free recording. This thesis explores the design of a body-temperature-sensing device inkjet-printed on a flexible substrate. All structures were first modeled by first-principles, theoretical calculations, and then simulated in HFSS. A variety of planar square inductor geometries were studied before selecting an optimal design. The designs were fabricated using multiple techniques and compared to the simulation results. It was determined that inductance must be carefully measured and documented to ensure good functionality. The same is true for parallel-plate and interdigitated capacitors. While inductance remains relatively constant with temperature, the capacitance of the device with a temperature-sensitive dielectric layer will result in a shift in the resonant frequency as environmental or ambient temperature changes. This resonant frequency can be wirelessly detected, with no battery required for the sensing device, from which the temperature can be deduced. From this work, the optimized version of the design comprises of conductive silver in with a temperature-sensitive graphene oxide layer, intended for inkjet-printing on flexible polyimide substrates. Graphene oxide demonstrates a high dielectric permittivity with good sensing capabilities and high accuracy. This work pushes the state-of-the-art in applying these novel materials and techniques to enable flexible body temperature sensors for future biomedical applications.

Zweidimensionale (2D) Materialien bieten eine Vielzahl von neuartigen Eigenschaften und sind aussichtsreich Kandidaten für ein breites Spektrum an Anwendungen. Da hexagonales Bornitrid (h-BN) für eine Integration in Heterostrukturen mit anderen 2D Materialien geeignet ist, erweckte dieses in letzter Zeit großes Interesse. Insbesondere van-der-Waals-Heterostrukturen, welche h-BN und Graphen verbinden, weisen viele potenzielle Vorteile auf, verbleiben in ihrer großflächigen Herstellung von kontinuierlichen Filmen allerdings problematisch. Diese Dissertation stellt eine Untersuchung betreffend des Wachstums von h-BN und vertikalen Heterostrukturen von Graphen und h-BN auf Ni-Substraten durch Molekularstrahlepitaxie (MBE) vor. Zuerst wurde das Wachstum von h-BN mittels elementarer B- und N-Quellen auf Ni als Wachstumssubstrat untersucht. Kristalline h-BN-Schichten konnten durch Raman-spektroskopie nachgewiesen werden. Wachstumsparameter für kontinuierliche und atomar dünne Schichten wurden erlangt. Das Keimbildungs- und Wachstumsverhalten so wie die strukturelle Güte von h-BN wurden mittels einer systemischen Veränderung der Wachstumstemperatur und -dauer untersucht. Die entsprechenden Beobachtungen wie der Änderungen der bevorzugten Keimbildungszentren, der Kristallgröße und der Bedeckung des h-BN wurden diskutiert. Ein Wachstum von großflächigen vertikalen h-BN/Graphen Heterostrukturen (h-BN auf Graphen) konnte mittels einem neuartigen, MBE-basierenden Verfahren demonstriert werden, welche es h-BN und Graphen jeweils erlaubt sich in der vorteilhaften Wachstumsumgebung, welche von Ni bereitgestellt wird, zu formen. In diesem Verfahren formt sich Graphen an der Schnittstelle von h-BN und Ni durch Präzipitation von zuvor in der Ni-Schicht eingebrachten C-Atomen. Schließlich konnte noch ein großflächiges Wachstum von Graphen/h-BN-Heterostrukturen (Graphen auf h-BN) durch das direkte abscheiden von C auf MBE-gewachsenen h-BN gezeigt werden. / Two-dimensional (2D) materials offer a variety of novel properties and have shown great promise to be used in a wide range of applications. Recently, hexagonal boron nitride (h-BN) has attracted significant attention due to its suitability for integration into heterostructures with other 2D materials. In particular, van der Waals heterostructures combining h-BN and graphene offer many potential advantages, but remain difficult to produce as continuous films over large areas. This thesis presents an investigation regarding the growth of h-BN and vertical heterostructures of graphene and h-BN on Ni substrates using molecular beam epitaxy (MBE). The growth of h-BN from elemental sources of B and N was investigated initially by using Ni as the growth substrate. The presence of crystalline h-BN was confirmed using Raman spectroscopy. Growth parameters resulting in continuous and atomically thin h-BN films were obtained. By systematically varying the growth temperature and time the structural quality as well as the nucleation and growth behavior of h-BN was studied. Corresponding observations such as changes in preferred nucleation site, crystallite size, and coverage of h-BN were discussed. Growth of h-BN/graphene vertical heterostructures (h-BN on graphene) over large areas was demonstrated by employing a novel MBE-based technique, which allows both h-BN and graphene to form in the favorable growth environment provided by Ni. In this technique, graphene forms at the interface of h-BN/Ni via the precipitation of C atoms previously dissolved in the thin Ni film. No evidence for the formation of BCN alloy could be found. Additionally, the suitability of ultraviolet Raman spectroscopy for characterization of h-BN/graphene heterostructures was demonstrated. Finally, growth of large-area graphene/h-BN heterostructures (graphene on h-BN) was demonstrated via the direct deposition of C on top of MBE-grown h-BN.