Dans les applications de l’électronique de puissance, les inductances sont généralement utilisées avec des formes d’ondes triangulaire et rectangulaire. Dans ces conditions, les inductances ne peuvent pas être caractérisées avec des formes d’ondes sinusoïdales. Il existe dans la littérature plusieurs méthodes de caractérisation des inductances. Aucune de ces méthodes ne permet de caractériser les composants magnétiques intégrés dans des conditions compatibles avec les formes d’onde de l’électronique de puissance. Une nouvelle méthode de caractérisation en temps réel convenable aux composants magnétiques planaires intégrées est développée dans cette thèse. Le composant sous test est utilisé dans un convertisseur DC-DC. A partir du courant i(t) et de la tension u(t) relevés à ses bornes, il est possible de déterminer la caractéristique φ(i) du composant. Connaissant le cycle d’hystérésis φ(i), les principaux paramètres du composant tels que la valeur de l’inductance, le niveau de saturation du matériau magnétique, les pertes fer peuvent être déterminées. Il convient de préciser qu’après l’enregistrement de u(t) et de i(t), un traitement des données approprié est effectué pour obtenir des signaux exploitables. L’approche a été testée avec succès sur des inductances discrètes et planaires intégrées pour des fréquences de mesure allant jusqu’au MHz. Nous avons ainsi montré que cette méthode permet de relever des cycles mineurs jusqu’à la saturation du matériau magnétique et de tracer l’évolution de l’inductance en fonction de la composante continue du courant. Les résultats montrent que pour de très faibles courants, on a des fortes valeurs de l’inductance qui sont quasiment constantes, en revanche à des courants moyens, l’inductance diminue fortement et à des forts courants, le circuit magnétique se sature et la valeur de l’inductance devient donc très faible. Cette méthode permet également d’étudier l’évolution des pertes fer dans les mêmes conditions d’utilisation que celles rencontrées en électronique de puissance / In applications of power electronics, inductors are generally used with triangular and rectangular waveforms. In these conditions, inductors cannot be characterized with sinusoidal waveforms. Many methods of characterization are developed, but none of them allow the characterization of integrated magnetic components under conditions compatible with power electronics waveforms. A new real-time characterization method suitable for integrated magnetic components is developed in this thesis. The Device under test is inserted in a DC-DC converter. From u(t) voltage and i(t) current recordings, it is possible to determine the φ(i) hysteresis loop that allows main inductor parameters to be determined: inductance value, core losses, and saturation of the magnetic material. It should be made clear when u(t) and i(t) are recorded, suitable signal processing is completed in order to obtain usable signals. The method has been successfully tested on discrete and integrated inductors with measurement frequencies up to MHz. We have thus shown that this method makes it possible to determine minor loops up to the saturation of the magnetic material and to plot the evolution of the inductance versus the DC bias current.The results show that for very low current values, we have obtained strong values of inductor and these values are constant. However, for the mean current, the inductance value decreases drastically and for strong current values, the magnetic material is saturated which causes very small values of inductor. This method also makes it possible to study the evolution of iron losses under the same conditions of use as those encountered in power electronics

Notre stratégie de stockage d’électrons repose sur la formation et la rupture de liaisons chimiques fonctionnant comme des réservoirs d’électrons. Le processus de formation réversible de liaisons est rendu possible grâce à la pré-organisation de groupements électroactifs – des pyridiniums – autour de plateformes aromatiques rigides et semi-rigides. Le relargage des électrons stockés au sein des réservoirs d’électrons par rupture de liaison suggère de manipuler des liaisons de nature affaiblie, c’est-à-dire allongées et contraintes stériquement, comme celles résultantes d’interactions hors-du-plan des groupements électrophoriques. Les contraintes stériques permettent ainsi la formation et la rupture de ces liaisons dans une gamme d’énergie raisonnable. Les pyridiniums sont des unités électrophoriques de choix dû à la versatilité de leurs propriétés de réduction. Les assemblages obtenus sont des systèmes bistables qui montrent des propriétés électrochimiques hystérétiques remarquables. En général, ce comportement repose sur la proche proximité des groupements électrophoriques qui permettent la formation d’une (ou plusieurs) orbitale(s) supramoléculaire(s) vacante(s) et liante(s) (SupLUMOs). Lors de la réduction, cette orbitale accueille deux électrons créant ainsi la liaison d’intérêt. Une famille de composés modèles a ainsi été conçue et synthétisée impliquant la formation réversible de liaisons homonucléaires (Csp3‒Csp3). Ces liaisons sont de nature localisée dans des systèmes à deux pyridiniums (bi-électroniques) et délocalisée dans des systèmes de plus de deux pyridiniums. / Our strategy to store electrons is based on chemical bond formation/cleavage being used as electron reservoirs. Reversible bond formation results from the preorganization of electrophoric moieties – pyridiniums – around aromatic rigid and semi-rigid scaffold. Using lower energy bonds is required to release stored electrons from the reservoirs. Such bonds can be obtained from sterically hindered out-of-plan bond formation between two electrophorical subunits. The resulting elongated bond can be cleaved in an available level of energy. Pyridiniums are prime candidates to achieve such properties based on their noteworthy redox behavior. The resulting molecular devices are bistable systems that display remarkable hysteretic electrochemical properties, based on close proximity of electrophoric subunits allowing one (or more) unoccupied supramolecular bonding orbital(s) (SupLUMOs). Upon reduction, two electrons are injected in this orbital resulting in the bond of interest. An entire family of compounds has been designed and synthesized to highlight reversible homonuclear bond formation (Csp3‒Csp3). Those bonds are localized in the case of a two-pyridiniums system, and delocalized when more pyridiniums are involved.

Ce travail de thèse met en évidence deux phénomènes microscopiques dissipatifs au voisinage de la ligne de contact dans les phénomènes de capillarité et d'adhésion. L'étude s'appuie sur des expériences dynamiques et des modélisations théoriques. La mesure expérimentale de quantités macroscopiques et la théorie hydrodynamique permettent d'extraire les informations dynamiques localisées au voisinage la ligne de contact. Les différents phénomènes dissipatifs localisés au voisinage de la ligne de contact ont pour origines les propriétés des substrats sur lesquels se déplace la ligne de contact. Pour une surface rigide hétérogène, nous avons développé un modèle rhéologique de la ligne de contact fondé sur l'hydrodynamique, permettant d'établir théoriquement l'évolution temporelle de la ligne de contact et de ses déformations. Une décomposition modale fondée sur la réduction de l'énergie par la théorie du chemin de réaction fournit une prédiction quantitative de la dynamique thermiquement activée de la ligne de contact, en accord avec l'expérience réalisée. Pour un substrat déformable, à partir de l'analyse de deux expériences différentes de mouillage dynamique et de l'estimation de la dissipation dans le substrat, fondée sur sa viscoélasticité, nous avons développé une compréhension générale du comportement dynamique d'une ligne de contact sur un substrat viscoélastique. Pour finir, ce modèle de dissipation viscoélastique est appliqué au cas de l'adhésion réversible, où expérimentalement nous mesurons la dynamique de pelage et de recollement sur un substrat viscoélastique. Cette extension à l'adhésion permet de relier les phénomènes interfaciaux en une compréhension générale / This thesis reveals two dissipative microscopic phenomena close to the contact line in the fields of capillarity and adhesion. The study is based on dynamic experiments and theoretical predictions. Experimental measurement of macroscopic quantities and the hydrodynamic theory give access to dynamic information located close to the contact line. The different dissipative phenomena, located close to the contact line, originate from the properties of the substrates on which the contact line moves. For a heterogeneous rigid surface, we have developed a rheological model of the contact line based on hydrodynamics, in order to theoretically establish the temporal evolution of the contact line and its deformations. A modal decomposition based on the reduction of the energy by the reaction path theory allows a quantitative prediction of the thermally activated dynamics of the contact line, in agreements with the experiment carried out. For a deformable substrate, based on the analysis of two different experiments of wetting dynamics and on the estimation of the dissipation in the substrate founded on its viscoelasticity, a general understanding of the dynamical behavior of contact lines on viscoelastic substrates is achieve. Finally, this viscoelastic model is applied to the case of weak adhesion, where experimentally we measured the peeling dynamics from a viscoelastic substrate. This extension to adhesion bridge the gap between different interfacial phenomena into a general understanding

La première thématique développée dans ce mémoire vise à développer de nouveaux algorithmes pour la commande des éoliennes reposant sur une machine asynchrone à double alimentation et des convertisseurs multi niveaux. Deux stratégies de contrôle direct du courant, basées sur l'Hystérésis à Zones Carrées et l'Hystérésis à Zones Circulaires (HZCi) ont été proposées. Celles-ci consistent à appliquer des vecteurs de tension appropriés pour contrôler les puissances actives et réactive générées et permettent également d'équilibrer les tensions du bus continu interne des convertisseurs. Des résultats de simulation et d'expérimentation montrent que la stratégie basée sur l'HZCi est meilleure en termes de forme d'onde et de contenu harmonique des tensions de sortie.La seconde concerne la supervision et la gestion des puissances active et réactive dans une ferme éolienne au vu de son intégration dans un réseau électrique. Cette supervision centralisée est assurée par un algorithme qui distribue les consignes de puissance aux éoliennes de la ferme de manière proportionnelle. Ces références sont fonction de la capacité maximale de production de l'éolienne. Pour cela, une analyse des transits de puissance dans le système éolien à base de la machine asynchrone à double alimentation a été effectuée. Elle a permis de déterminer la caractéristique (P, Q) du générateur et de calculer ses limites de compensation en termes de puissance réactive. Une gestion locale des puissances de chaque éolienne a été développée permettant ainsi une répartition des puissances entre le stator de la machine et le convertisseur coté réseau en considérant plusieurs modes de fonctionnement du système éolien.

The link suspension is the most prevailing suspension system for freight wagons in Central and Western Europe. The system design is simple and has existed for more than 100 years. However, still its characteristics are not fully understood. This thesis investigates the dynamic performance of freight wagons and comprises five parts: In the first part a review of freight wagon running gear is made. The different suspension systems are described and their advantages and disadvantages are discussed. The second part focuses on the lateral force-displacement characteristics of the link suspension. Results from stationary measurements on freight wagons and laboratory tests of the link suspension characteristics are presented. To improve the understanding of various mechanisms and phenomena in link suspension systems, a simulation model is developed. In the third part the multibody dynamic simulation model is discussed. The previous freight wagon model developed at KTH is able to explain many of the phenomena observed in tests. In some cases, however, simulated and measured running behaviour differ. Therefore, a new simulation model is presented and validated against on-track test results. The performance of standard two-axle freight wagons is investigated. The most important parameters for the running behaviour of the vehicle are the suspension characteristics. The variation in characteristics between different wagons is large due to geometrical tolerances of the components, wear, corrosion, moisture or other lubrication. The influence of the variation in suspension characteristics and other parameters on the behaviour of the wagon, on tangent track and in curves, is discussed. Finally, suggestions for improvements of the system are made. A majority of the traffic related track deterioration cost originates from freight traffic. With heavier and faster freight trains the maintenance cost is likely to increase. In the fourth part the possibility to improve ride comfort and reduce track forces on standard freight wagons with link suspension is discussed. The variation of characteristics in link suspension running gear is considerable and unfavourable conditions leading to hunting are likely to occur. Supported by on-track tests and multibody dynamic simulations, it is concluded that the running behaviour of two-axled wagons with UIC double-link suspension as well as wagons with link suspension bogies (G-type) can be improved when the running gear are equipped with supplementary hydraulic dampers. Finally in the fifth part the effects of different types of running gear and operational conditions on the track deterioration marginal cost — in terms of settlement in the ballast, component fatigue, wear and RCF — is investigated. Considerable differences in track deterioration cost per produced ton-km for the different types of running gear are observed. Axle load is an important parameter for settlement and component fatigue. Also the height of centre of gravity has significant influence on track deterioration, especially on track sections with high cant deficiency or cant excess. / QC 20100802

Capacitance-voltage-frequency measurements on n+-GaN/AlxGa1−xN UV/IR dual-band detectors are reported. The presence of shallow Si-donor, deep Si-donor, and C-donor/N-vacancy defect states were found to significantly alter the electrical characteristics of the detectors. The barrier Al fraction was found to change the position of the interface defect states relative to the Fermi level. The sample with Al fraction of 0.1 shows a distinct capacitance-step and hysteresis, which is attributed to C-donor/N-vacancy electron trap states located above the Fermi level (200 meV) at the heterointerface; whereas, the sample with Al fraction of 0.026 shows negative capacitance and dispersion, indicating C-donor/N-vacancy and deep Si-donor defect states located below the Fermi level (88 meV). When an i-GaN buffer layer was added to the structure, an anomalous high-frequency capacitance peak was observed and attributed to resonance scattering due to hybridization of localized Si-donor states in the band gap with conduction band states at the i-GaN/n+-GaN interface.

The purpose of this master thesis has been to investigate the possibility to increase the damping capacity of shear walls with the use of glue in combination with nails. It has been done with experiment and with FE-simulations.   The first experimental part is tensile tests of four different glue types which have been analyzed to get material data. The second experimental part is analysis of glued-nail joint in shear test. FE-simulations of glued-nail joints are finally done in Abaqus/CAE.   The experiments in shear tests of glued-nail joints showed that two glues failed and could not be used for further analysis. The two other glues had an increase in damping capacity compared with results from nail joint.   Further FE-simulations of glued-nailed joints have to be done in order to calibrate the FE-model with the results of glued-nailed joints in shear tests. The uncertainty is found in the material properties of the glue, experimental test data was not adequate. The conclusion is that glued-nail joints can increase the damping capacity in shear walls. / Damping capacity, shear wall, hysteresis, shear test, finite element method, Abaqus/CAE

Severe seismic events in urban regions during the last two decades revealed that the structures constructed before the development of modern seismic codes are the most vulnerable to earthquakes. Sub-standard reinforced concrete buildings constitute an important part of this highly vulnerable urban building stock. There is urgent need for the development and improvement of methods for seismic performance assessment of existing reinforced concrete structures. As an alternative to current conventional force-based assessment methods, a performance evaluation procedure for structural members, mainly reinforced concrete columns is proposed in this study, by using an energy-based approach combined with the low cycle fatigue concept. An energy-based hysteresis model is further introduced for representing the inelastic response of column members under severe seismic excitations. The shape of the hysteresis loops are controlled by the dissipated cumulative energy whereas the ultimate strength is governed by the low cycle fatigue behavior. These two basic characteristics are obtained experimentally from full scale specimens tested under constant and variable amplitude displacement cycles. The first phase of the experimental program presented in the study constitutes of testing sub-standard non-conforming column specimens. The second phase of testing was conducted on standard, code compliant reinforced concrete columns. A total number of 13 specimens were tested. The behavior of these specimens was observed individually and comparatively according to the performance based objectives. The results obtained from the experiments were employed for developing relations between the energy dissipation capacity of specimens, the specimen properties as well as the imposed displacement history. Moreover, the measured rotation capacities at the plastic regions are evaluated comparatively with the limits proposed by modern displacement-based seismic design and assessment provisions.

The aim of the current study is to model (non-wetting) liquid flow in a packed bed under the influence of gas flow. It has been observed experimentally that non-wetting liquid flows in a packed bed in form of small droplets and rivulets falling through the void regions. Continuum models have not been successful in predicting liquid flow paths when the liquid is injected through a point source in the packed bed. In the current study, we present a discrete deterministic model for modeling the liquid flow in a packed bed, under the influence of gas flow. When a high velocity gas blast in injected into a dry packed bed, a cavity or a void is formed in front of the nozzle. The cavity size increases with increasing gas velocity and exhibits hystersis in size upon increasing and decreasing gas flow rate. The cavity size is very important in determining the gas penetration into the packed bed. A proper gas flow profile prediction is necessary for determining it’s effect on the liquid flow behavior. Attempts at modeling cavity sizes have mostly been confined to experimental studies and development of correlations. Different correlations show different dependence on operating as well as bed parameters and a fundamental understanding of the cavity formation and hystersis phenomena is missing. We adopt a combined Eulerean-Lagrangian approach to study the above mentioned phenomena mathematically. Gas is modeled as a continua and solid as discrete (soft sphere D.E.M. approach). Hystersis and cavity formation studies are carried out in a 2D-slot rectangular packed bed. A discrete deterministic liquid flow model (developed and validated under structured packing conditions using x-ray radiography flow visualization technique), is used to study the effect of presence of liquid on the dry bed void size, when liquid is injected in a packed bed through a point source. It is found that the gas pushes the liquid away from the nozzle side wall. Also, the cavity sizes during gas velocity decreasing case are found to be larger in size than the void size obtained during velocity increasing case for the same inlet gas flow rate. This difference is void size leads to more gas penetration into the bed and thus more liquid shift away from the nozzle side wall. Presence of liquid is found to affect the void size (compared to dry bed size) negligibly.

In this thesis I present studies of some transport and spectroscopic properties for two di erent materials exhibiting large magnetoresistance. Both of these materials are oxides of transition metals, showing exotic magnetic and transport properties. Despite these similarities, they are very different in many other aspects. One of them is an oxide of Manganese, along with a rare-earth metal, and exhibits large magnetoresistance under certain conditions, when doped by an alkaline earth metal. They are known as doped rare-earth manganites. The other material, Sr2FeMoO6, exhibits large magnetoresistance in the parent compound, without any doping, but only in the polycrystalline state. The manganites, on the other hand, show magnetoresistance under appropriate conditions in both single crystal and in polycrystalline state. Moreover, manganites exhibit several Metal-Insulator Transitions (MIT) as a function of doping, temperature and magnetic field. Sr2FeMoO6, on the other hand, is usually always metallic. In the first chapter, a brief introduction is provided regarding different types of magnetoresistance (MR) phenomena observed in different materials, namely Anisotropic MR (AMR), Giant MR (GMR), Collosal MR (CMR), Tunneling MR (TMR), Powder MR (PMR) etc. Out of these, CMR and PMR are found in doped manganites, while Sr2FeMoO6 exhibits PMR only. Next, a brief overview of the structure, properties and theories for both of these materials is provided. For the case of doped manganites, a short introduction is given for a novel two-fluid hamiltonian (called l - b model) which was proposed recently by Ramakrishnan et. al.. This model reproduces several exotic transport and magnetic properties of manganites which were inexplicible by earlier theories. The model was solved within the Dynamical Mean Field Theory (DMFT) framework by Hassan et. al.. A brief description of this DMFT solution is given. Many of the DMFT results for this model have been used in the subsequent chapters. In the second chapter, the hysteresis behaviour of the magnetoresistance and the magnetization (M ) of powdered Sr2FeMoO6 is considered in detail. In a recent experi- ment by Sarma et. al., it was found that this material, when powdered exhibits an exotic variety of PMR. In ordinary PMR, the hysteresis behaviour of the MR is supposed to follow that of M, in the sense that the coercive fields should be identical in both cases. Also, the MR is supposed to be roughly proportional to the square of the magnetization. However, in the experiments by Sarma et. al. on cold-pressed Sr2FeMoO6 powder, it was observed that the M R did not appear to be determined purely by the magnetization. Rather, the coercive fields for the hysteresis of the MR was almost 6 times that of M . Moreover, the quantity M R/M2, instead of remaining constant with changing magnetic field, itself has a hysteresis loop. Apart from establishing the exotic nature of the PMR, the experiment also tries to determine whether the MR originates from intra-grain or inter-grain tunneling. In the second chapter we present a simple toy model to reproduce the experimental results, and provide theoretical explanations. A combination of Monte Carlo and transfer matrix methods are used to simulate the hysteresis behaviour of the M R as well as of M . We show that the observed data can be understood if it is as- sumed firstly that the MR arises predominantly from inter-grain rather than intra-grain tunneling, and that the inter-grain boundaries are themselves magnetic with a coercive field higher than that of the grains. In order to motivate the use of Monte Carlo method for studying hysteresis, a brief survey of main results obtained for some simple models using this technique is also provided. In the third chapter, we study the doping and temperature dependence of core-level photoemission spectra in doped rare-earth manganites. In some recent experiments on Strontium doped (LSMO) and Barium doped (LBMO) samples, it has been observed that the M n2p3/2 core-level spectra shows an intriguing spectral weight transfer over a range of several eV , as a function of doping (x) and temperature (T ), in the ferromagnetic metallic phase. Specifically, there appears a shoulder adjacent to the main peak on the side of lower binding energy, which increases in weight and intensity as the doping increases or the temperature decreases. In LSMO samples, another shoulder was noticed on the higher binding energy side also. Moreover, in data obtained from LBMO samples, the spectra at different temperatures was subtracted from the spectra at/above Tc, and then this difference spectrum was integrated. The integrated weight, when normalized by the weight at the lowest temperature, appears to follow the square of the measured magnetization almost exactly. In order to understand the experimental data, we extended the aforementioned l - b model to include a core-level, and the attractive interaction due to a core-hole on the local valence levels. The impurity problem arising in DMFT, consisting of a single impurity site coupled to a bath, was tailored for the photoemission problem, by including this extra core-level at the impurity site. The hybridization parameters for the bath were determined self-consistently from the DMFT, and then the single particle spectral function for the core-hole was determined. This spectral function is proportional to the photo emission intensity. We found that our calculations reproduced the observed spectral weight transfer as a function of x and T both in trends and in magnitude. The integrated difference spectra weight was found to follow the square of the DMFT magnetization, just as in the experiment. Linear discretization of the conduction bath was used for all the above-mentioned cases. In one particular case, a logarithmic discretization was also undertaken for comparison, and also to obtain the exponents of the edge singularities in the theoretical spectra. In the fourth chapter, the possibility of Anderson Localization in manganites is in- vestigated, using the l - b model. According to this model, a large fraction of the valence electrons are polaronically self-trapped even in the ferromagnetic metallic phase. Due to strong on-site Coulomb interaction, these polarons provide a strongly scattering background, which can localize the mobile-electron band states close to the band edges. Since the fraction of valence electrons which are truly mobile is small, hence the Fermi energy lies close to the lower band edge. Hence, there is a possibility of an Anderson Insulator phase where all charge carriers are localized. To investigate this, we studied the behaviour of the mobility edges as a function of doping. DMFT alone does not include the physics of localization. Hence, in order to obtain the mobility edges, we combined the DMFT results with the Self-consistent Theory of Localization (STL), using a simplified prescription called Potential Well Analogy (PWA) due to Economou et. al.. We found that there is indeed an Anderson Insulator phase in a certain region of doping, which would otherwise have been supposed to be metallic based on purely DMFT results. Finally, we have compared this result, obtained using effective field theories, with an actual real space simulation of the l - b model at T=0. In this case, the mobility edge trajectories were obtained by studying the Inverse Participation Ratio (IPR), as a function of band energy and doping. In the concluding chapter, the principal results presented in this thesis are summa- rized. The limitations of the approach or approximations used are discussed, and future possibilities for overcoming these limitations outlined.