In the past few decades, many people spent a lot of time and used many different ways, which includes analytic method, numerical method, and experimental observations for investigating the flow around circular cylinder problem. Eventually, the purpose of these investigations is to determinate the force acting on the cylinder and which is very useful and important for marine and hydraulic engineering. Essentially, it can be divided into three circumstances, (i) the flow around a fixed cylinder, (ii) the flow around a rotating cylinder, (iii) the flow around a moving cylinder. The first two conditions have already been will discussed. Consequently, besides analyzing the first two conditions and comparing with reference papers, the purpose of this present is discussing the variation of the flow field and the force acting on the cylinder by using finite difference method. Because of the considerable quantity of computation, using parallel computing for this model to speedup the numerical process is also one of the issues of the present.

Parallel-plate electrostatic actuators driven by a voltage difference between two electrodes suffer from an operation range limited to 30% of the gap that has significantly restrained their applications in Microelectromechanical systems (MEMS). In this thesis, the travel range of an electrostatic actuator made of a micro-cantilever beam electrode above a fixed electrode is extended quasi-statically to 90% of the capacitor gap by introducing a voltage regulator (controller) circuit designed for low frequency actuation. The developed large-stroke actuator is valuable contribution to applications in optical filters, optical modulators, digital micro-mirrors and micro-probe based memory disk drives. To implement the low-frequency large-stroke actuator, the beam tip velocity is measured by a vibrometer, the corresponding signal is integrated in the regulator circuit to obtain the displacement feedback, which is used to modify the input voltage of the actuator to reach a target location. The voltage regulator reduces the total voltage, and therefore the electrostatic force, once the beam approaches the fixed electrode so that the balance is maintained between the mechanical restoring force and the electrostatic force that enables the actuator to achieve the desired large stroke. A mathematical model is developed for the actuator based on the mode shapes of the cantilever beam using experimentally identified parameters that yields good accuracy in predicting both the open loop and the closed loop responses. The low-frequency actuator also yields superharmonic resonances that are observed here for the first time in electrostatic actuators. The actuator can also be configured either as a bi-stable actuator using a low-frequency controller or as a chaotic resonator using a high-frequency controller. The high-frequency controller yields large and bounded chaotic attractors for a wide range of excitation magnitudes and frequencies making it suitable for sensor applications. Bifurcation diagrams reveal periodic motions, softening behavior, period doubling cascades, one-well and two-well chaos, superharmonic resonances and a reverse period doubling cascade. To verify the observed chaotic oscillations, Lyapunov exponents are calculated and found to be positive. Furthermore, a chaotic resonator with a quadratic controller is designed that not only requires less voltage, but also produces more robust and larger motions. Another metric of chaos, information entropy, is used to verify the chaotic attractors in this case. It is found that the attractors have a common information entropy of 0.732 independent of the excitation amplitude and frequency.

Parallel-plate electrostatic actuators driven by a voltage difference between two electrodes suffer from an operation range limited to 30% of the gap that has significantly restrained their applications in Microelectromechanical systems (MEMS). In this thesis, the travel range of an electrostatic actuator made of a micro-cantilever beam electrode above a fixed electrode is extended quasi-statically to 90% of the capacitor gap by introducing a voltage regulator (controller) circuit designed for low frequency actuation. The developed large-stroke actuator is valuable contribution to applications in optical filters, optical modulators, digital micro-mirrors and micro-probe based memory disk drives. To implement the low-frequency large-stroke actuator, the beam tip velocity is measured by a vibrometer, the corresponding signal is integrated in the regulator circuit to obtain the displacement feedback, which is used to modify the input voltage of the actuator to reach a target location. The voltage regulator reduces the total voltage, and therefore the electrostatic force, once the beam approaches the fixed electrode so that the balance is maintained between the mechanical restoring force and the electrostatic force that enables the actuator to achieve the desired large stroke. A mathematical model is developed for the actuator based on the mode shapes of the cantilever beam using experimentally identified parameters that yields good accuracy in predicting both the open loop and the closed loop responses. The low-frequency actuator also yields superharmonic resonances that are observed here for the first time in electrostatic actuators. The actuator can also be configured either as a bi-stable actuator using a low-frequency controller or as a chaotic resonator using a high-frequency controller. The high-frequency controller yields large and bounded chaotic attractors for a wide range of excitation magnitudes and frequencies making it suitable for sensor applications. Bifurcation diagrams reveal periodic motions, softening behavior, period doubling cascades, one-well and two-well chaos, superharmonic resonances and a reverse period doubling cascade. To verify the observed chaotic oscillations, Lyapunov exponents are calculated and found to be positive. Furthermore, a chaotic resonator with a quadratic controller is designed that not only requires less voltage, but also produces more robust and larger motions. Another metric of chaos, information entropy, is used to verify the chaotic attractors in this case. It is found that the attractors have a common information entropy of 0.732 independent of the excitation amplitude and frequency.

Afin de répondre au besoin croissant de puissance de calcul qui intéresse les applications scientifiques, des architectures de calculateurs de plus en plus parallèles se répandent, de type cluster, MPP, ou autre, avec de plus en plus d’unités de calcul. Les codes de calcul, en s’adaptant à ce parallélisme, produisent de fait des résultats de plus en plus volumineux. Dans le domaine plus spécifique de la dynamique moléculaire appliquée à la physique des matériaux, le nombre de particules d’un système étudié est aujourd’hui de l’ordre de quelques millions à quelques centaines de millions, ce qui pose d’importants problèmes d’exploitation des données produites. Les solutions de visualisation utilisées auparavant ne sont plus capables de traiter une telle quantité d’information de façon interactive. Le but de cette thèse est de concevoir et d’implémenter une solution de manipulation de données composées d’un très grand nombre de particules, formant un ensemble .dense. ( par exemple, une simulation en dynamique moléculaire de quelques dizaines de millions d’atomes ). Cette solution devra être adaptative pour pouvoir fonctionner non seulement sur un poste de bureau de type .desktop., mais aussi dans des environnements collaboratifs et immersifs simples, par exemple de type mur d’image, avec éventuellement des éléments de réalité virtuelle (écrans stéréo, périphériques 3D). De plus, elle devra augmenter l’interactivité, c’est-à-dire rendre le système plus apte à réagir avec l’utilisateur, afin d’effectuer des phases d’exploration des données plus efficaces. Nous proposons une solution répondant à ces besoins de visualisation particulaire dense, en apportant des améliorations : – à l’interaction, par un système permettant une exploration simple et efficace d’une simulation scientifique. A cette fin, nous avons conçu, développé, et testé le .FlowMenu3D., une extension tridimensionnelle du .FlowMenu., un menu octogonal utilisé en environnement 2D. – à l’interactivité, par une architecture optimisée pour le rendu de systèmes particulaires denses. Cette dernière, intégrée au framework de visualisation VTK, est conçue sur le principe d’un rendu parallèle hybride sort-first/sort-last, dans lequel le système sort-last Ice-T est couplé à des extensions de partitionnement spatial des données, d’occlusion statistique par une méthode de Monte-Carlo, mais aussi de programmation GPU à base de shaders pour améliorer la performance et la qualité du rendu des sphères représentant les données. / The need for more computing power for scientific applications leads supercomputer manufacturers to rely on more and more parallel architectures, such as cluster or MPP systems, increasing the number of processing units. The correlative adjustment of computing software - simulation codes - to massive parallelism leads to larger and larger datasets. More specifically, in molecular dynamics simulations applied to material physics, particle sets are typically composed of millions to hundreds of millions of particles, which raises important postprocessing issues. Most current visualization solutions do not scale up to interactively handle such an amount of information. The goal of this thesis is to design and implement a post-processing solution which is able to handle .dense. sets with a very large number of particles (for example, a molecular dynamics simulation result with several millions of atoms). This solution will have to be adaptive enough to run either on a desktop environment, or on a simple collaborative and immersive configuration, such as a tiled display, possibly with some virtual reality devices (such as stereo displays and 3D interaction peripherals). Moreover, this solution will have to maximize the interactivity, i.e. to increase the reactivity of the system against the user commands, to improve data exploration stages. We propose a solution optimized for dense particle systems visualization, improving both interaction and interactivity : – regarding interaction, a new simple and efficient way of exploring a scientific simulation is proposed. We developed and tested .FlowMenu3D., a 3D extension of the .FlowMenu. concept, an octagonal menu previously designed for 2D environments. – regarding interactivity, we propose an optimized architecture for dense particle systems rendering. The system relies on VTK visualization framework and a parallel hydrid sort-first/sort-last renderer. The existing Ice-T component is augmented with spatial data partitioning as well as statistical (Monte Carlo) occlusion mechanisms. GPU shaders eventually improve the performance and quality of rendering, using spheres as basic primitives.

The design, synthesis and evaluation of chiral selectors that allow the determination of enantiomeric composition using electrospray ionization mass spectrometry are detailed herein. The enantiomers of the chiral selector were mass labeled at a distant site from the chiral recognition sites of the molecules. The mass-labeled enantiomers were mixed in a one-to-one ratio to form a quasi-racemate. Chiral recognition can be observed by comparing relative abundances of the pseudo-diastereomeric selector-analyte complexes in the mass spectrum. The observed sense of chiral recognition with mass spectrometry was consistent with that observed chromatographically using a corresponding chiral stationary phase in every case. The complex intensity fraction (CIF, intensity of one selector-analyte complex divided by the sum of the intensities for both selector-analyte complexes) is linear with the enantiomeric composition. The slope of this line is an indication of the extent of the enantioselectivity: the larger the slope, the more significant the enantioselectivity. In addition, this line can be used as a calibration curve for the quantitative determination of enantiomeric composition of the same analyte with unkown enantiomeric composition. Amide derivatives of DNB-amino acids were first used as pseudo-enantiomeric chiral selectors in the presence of added lithium chloride. The enantioselectivity values were smaller than those observed on chiral HPLC using the corresponding chiral stationary phase. The use of deprotonated DNB-amino acids as chiral selectors provides higher enantioselectivities, but with low ion abundances. Tertiary amine appended analogues of the chiral stationary phase DNB-Leucine were prepared. The amine was appended to provide a site for ready ionization (through protonation). The performance of chiral selectors of this type was compared to the original chiral selectors that lack this functional group. Chiral recognition was also observed in a reciprocal sense using proline-derived pseudo-enantiomeric chiral selectors and analytes similar to DNB-amino acid esters or amides. Optimization of the electrospray ionization conditions provided similar enantioselectivities to those from chiral HPLC. Libraries of tertiary amine appended derivatives of DNB-dipeptides, which were prepared through combinatorial peptide synthesis, were screened using electrospray ionization mass spectrometry. The use of electrospray ionization mass spectrometry as a discovery tool for new chiral selectors is discussed.

Field-Programmable Gate Arrays (FPGAs) systems now comprise many processing elements that are processors running software and hardware engines used to accelerate specific functions. To make the programming of such a system simpler, it is easiest to think of a shared-memory environment, much like in current multi-core processor systems. This thesis introduces a novel, shared-memory, cache-coherent infrastructure for heterogeneous systems implemented on FPGAs that can then form the basis of a shared-memory programming model for heterogeneous systems. With simulation results, it is shown that the cache-coherent infrastructure outperforms the infrastructure of Woods [1] with a speedup of 1.10. The thesis explores the various configurations of the cache interconnection network and the benefit of the cache-to-cache cache line data transfer with its impact on main memory access. Finally, the thesis shows the cache-coherent infrastructure has very little overhead when using its cache coherence implementation.

Field-Programmable Gate Arrays (FPGAs) systems now comprise many processing elements that are processors running software and hardware engines used to accelerate specific functions. To make the programming of such a system simpler, it is easiest to think of a shared-memory environment, much like in current multi-core processor systems. This thesis introduces a novel, shared-memory, cache-coherent infrastructure for heterogeneous systems implemented on FPGAs that can then form the basis of a shared-memory programming model for heterogeneous systems. With simulation results, it is shown that the cache-coherent infrastructure outperforms the infrastructure of Woods [1] with a speedup of 1.10. The thesis explores the various configurations of the cache interconnection network and the benefit of the cache-to-cache cache line data transfer with its impact on main memory access. Finally, the thesis shows the cache-coherent infrastructure has very little overhead when using its cache coherence implementation.

This dissertation introduces FESM (Flood Event Simulation Model), a Geographic Information System (GIS) tool designed for use on gaged river systems that can be used to guide logistic support during disaster events. FESM rapidly generates flood predictions using elevation data from real-world sensors or generated by other models. Verification and validation data for FESM are provided. In order to construct a visualization system for interacting with FESM outputs, single buffer and dual buffer techniques for moving massive datasets to the GPU for processing using OpenCL were rigorously tested and timed, and an analysis of the costs/benefits of using buffers or images was conducted. Finally, DRO (Dynamic Raster Overlay), a visualization system for analysis of datasets composed of multiple overlapping flood maps is introduced, and expert feedback is provided on the effectiveness of DRO with selected case studies.

Le phénomène du commerce parallèle repose sur un constat de base à la fois simple et évident: le niveau des prix d’un même produit « créé et fabriqué » par le même titulaire peut souvent varier de manière importante d’un pays à l’autre, en raison notamment des disparités dans le revenu moyen des consommateurs, des règlementations nationales entravant les échanges, ou encore du niveau de concurrence qui peut être affecté par des restrictions d’ordre privé, telles que des conventions de partage de marchés ou des contrats de distribution. Les commerçants que l’on appelle « importateurs parallèles » achèteront le produit là où il est le moins cher, qu’il provienne directement du titulaire ou non, en vue de le revendre là où ils pourront faire le meilleur profit. Dans ce contexte, des problèmes apparaîtront dans le pays d’importation dans la mesure où les commerçants parallèles vendront, bien évidemment, moins cher que les distributeurs désignés par le titulaire. Le titulaire en sa qualité de « créateur » de produits, est souvent titulaire de droits de propriété intellectuelle qu’il désirera, non sans une certaine légitimité, « monnayer », tentant ainsi de cloisonner les marchés en vue d’obtenir le prix le plus élevé qu’un consommateur serait prêt à payer pour un produit « x » sur un marché « y ». Cet essai a pour mérite de mettre en tension, d'une part, la liberté des échanges et de la concurrence qui favorisent, notamment, l’obtention des produits au meilleur prix pour les acheteurs et les consommateurs et, d’autre part, la nécessité de protéger les efforts de création dans l’industrie afin que celle-ci continue à investir dans la recherche et le développement de nouveaux produits. / The phenomenon of parallel trade rests on a simple and obvious basic premise: the price levels of the same product “created and made” by the same manufacturer can often vary considerably from one country to another. This variation is notably due to disparities in the mean income of consumers, trade-prohibiting national regulations or the level of competition which may itself be affected by private restrictions such as market sharing agreements or distribution contracts. The merchants, termed “parallel importers” will buy the product where it is cheapest, whether or not it stems directly from the manufacturer, and resell it where they will obtain the best profits. Some problems will arise in the country of importation, insofar as parallel importers will sell at a lower price than the distributors designated by the manufacturer. The manufacturer, in his quality of “creator” of goods, is often the owner of intellectual property rights which it will understandably try to monetize, in an attempt to partition the markets in order to obtain the highest price that the consumer is willing to pay for product “x” on market “y”. This article, mainly jurisprudential, seeks to underline the tension between the free exchange of goods in open markets which notably allows consumers and buyers to obtain products at the best price, and the necessity of protecting the creation efforts of the industry in order to encourage investment in the research and development of new products.

An der TU Dresden wurde eine Forschungsarbeit zur experimentell-theoretischen Analyse des Tragverhaltens von Verankerungen durchgeführt, die hauptsächlich zur oberflächenparallelen Einleitung großer Schubkräfte in einen Betongrund ausgelegt sind. Diese Verankerungen ermöglichen die Übertragung einer großen Querkraftbe¬anspruchung mit geringer Exzentrizität zur Oberfläche, die mit einer Normalbean¬spruchung in Zug- oder Druckrichtung kombiniert werden kann. Die wesentlichen Grundelemente sind zumindest ein Schubdübel, ein Zuganker und eine Grundplatte, mehrere Schubdübel oder Anker können vorhanden sein. Der quaderförmige Schub¬dübel überträgt die Kraft in den Beton. Die angreifenden Lastmomente und das Moment aus der exzentrischen Einleitung der Schublast werden in ein vor und hinter dem Dübel wirkendes Kräftepaar zerlegt. Die Zugkomponente wird dem Zuganker in Form eines Kopfbolzendübels zugewiesen. Die Grundplatte verbindet die lastseitige Befestigungskonstruktion mit den lastübertragenden Bauteilen Schubdübel und Zug¬anker. Gleichzeitig kann die Grundplatte in den Beton eingelassen sein und überträgt dann ebenfalls Schublasten. Die Lastabtragung und das Verhalten der Einbauteile im Versuch wurden analysiert. Von besonderer Bedeutung ist hierbei die Last-Verschiebungslinie, die die Lastüber¬tragung charakterisiert. Ziel war es, ein Bemessungskonzept zur Dimensionierung der Schubdübel zu erarbeiten. Um die im Versuch erkennbare maßgebende Beein¬flussung der Lastübertragungskapazität durch den sich einstellenden Verformungs¬zustand eingehender untersuchen zu können, wurde der Versuch mit finiten Elemen¬ten abgebildet. So war es möglich, in einem ersten Schritt Lastübertragungskapazitä¬ten für unverformte Systeme zu ermitteln, die praktisch im Versuch nicht oder nur extrem aufwändig realisiert werden könnten. In einem zweiten Schritt kann dann der Einfluss der Verformung berücksichtigt werden. Hierbei können die Parameter ge¬nauer und vielschichtiger variiert werden als dies bei den relativ aufwändigen Versu¬chen rein quantitativ möglich war. Das in dieser Arbeit vorgestellte Berechnungsver¬fahren liefert neben der maximal aufnehmbaren Schubbeanspruchung geometrisch beliebiger Einbauteile auch die Beanspruchbarkeit einzelner Baugruppen und erlaubt es zu beurteilen, inwieweit die einzelnen Lastübertragungsmechanismen additiv oder alternativ wirksam sind. Das Verfahren gibt Auskunft über das zu erwartende lastabhängige Verhalten der Einbauteile inklusive der Vorhersage von Teilversagens¬zuständen. Das Potential der Befestigungssysteme zur Übertragung großer Schublasten ist durch die Aufsplittung der Lastkomponenten und Übertragung durch spezialisierte Bauteile begründet. Diese können in ihrer Beschaffenheit den Erfordernissen optimal angepasst werden. Einsatzgebiete für diese Befestigungen sind alle konzentrierten Lasteinleitungen oder Montagestöße zwischen Betonbauteilen oder Stahl-Holz-Kon¬struktionen auf der einen und Betonelementen auf der anderen Seite. Neben der Auflagerung von Trägern oder Fassadenelementen ist vor allem auch an die Ertüchti¬gung und Verstärkung bestehender Betonbauteile durch zusätzliche Stahlkonstrukti¬onen zu denken. / An experimental and theoretical analysis of the behaviour of complex shear loaded fastenings was carried out at Dresden University of Technology. The main focus was on applications which introduce a great amount of shear load value parallel to the surface into a concrete base using a stell shear lug. The behaviour of these special fixings is presented as the result of the finished research program. Structures, as shown in the opposite figure, are able to transmit high values of shear loads to the anchor ground. An additional loading of normal force und bending moment is suitable, but will only be supposed to cover the necessary tolerance and off centre condition of the fixture parts. The advantage of fastening with shear lugs is based on the splitting of the load transfer into different components. The shear lug/lugs caries/carry the shear load, and the tie bar itself only balances the system by tension load due to moment loading and normal loading. Thusall loads are transferred by highly specialized components. If the base plate is embedded, there will be a shear load transferred in front of the base plate, too. The load carrying behaviour of the fastenings in the experiments is analyzed. Here, the load-displacement relationship is of special importance . This thesis aims at elaborating a dimensioning concept for the design of shear lugs. The tests show a decisive influence of the movements – displacements as well as rotations – of the fastenings to the load carrying capacity. In order to to be able to examine this behaviour, the experiment is redesigned by finite elements. In a first step it is therefore possible to examine fastenings without movements based on the anchor bolt stiffness. In a second step, the influence of the deformation can be explored. Here, the parameters can be varied more exactly and in a more complex way than this is possible in terms of quantity with the comparetively extensive test setup. The calculation method being introduced here does not only supply the maximum shear load transfer capacity of the hole fastening system, a system with free geometries, but also shows the partial shear load transfer capacity of the individual parts of the unit. This method is a means of predicting the load-related behaviour of the fastenings to be expected, including the forecast of partial failure states. As a result, a suggestion for the design of the fittings is shown. This design can be applied to all fittings by splitting the different load components to especially provided anchor elements. The design resistance and behaviour of the fastenings is estimated, depending on different geometries and stiffness conditions of the lug. Different strength classes of the concrete as well as different load combinations are taken into consideration. First examples from building sites are also available.