This thesis deals with the subject of fibre optic magnetic field sensors utilizing iron garnet materials. Such materials exhibit a large Faraday rotation which make them advantageous for application in compact mag­netic field sensors. After an introduction, in which fibre optic sensors and optical methods to measure electric current are reviewed, the original research work is summarized. A system for the measurement of the magneto-optic properties of trans­parent materials is described. Measurement results, showing the influence of temperature, magnetic field direction and sample treatment on the magneto-optical proper­ties of YIG-crystals, are presented. The proper­ties of thin magneto-optical waveguiding films have also been studied using different light coupling methods. Measurement results obtained for holo­graphic grating, prism and edge (end-fire) light coupling to different substituted YIG films are presented. It is shown that the launching method may affect the properties to be measured. The design and performance of several versions of extrinsic guided wave fibre optic magnetic field sensors are then reported. The sensors employ substi­tuted YIG (Yttrium Iron Garnet, Y3Fe5O12) thin film waveguides as sensing elements. Polari­zation maintaining fibres were used as feed and return to provide two signal channels. The signals were combined in a balanced measure­ment system, providing insensitivity to both fluctuations in optical power and loss. Sensors have been made both with separate fibres to guide the light to and from the sensing element and with a single fibre for both functions. The two fibre version, although less ”elegant”, is found to have a better performance. This version also makes it possible to determine both the magnitude and sign of the magnetic field. Measurement results indicate a usable measurement range of at least several mT with a noise equivalent magnetic field level of less than 8 nT/root(Hz). The design and performance of multimode fibre optic magnetic field sensors utilizing the Faraday effect in an epitaxially grown thick (YbTbBi)IG film is also described. This type of sensor is found to be linear over a range from 27 mT to less than 270 nT. Sensor prototypes suitable for current monitoring in high voltage transmission lines have also been developed. / QC 20111209 / YIG

La Magnéto-Impédance Géante (GMI pour Giant Magneto-Impedance) présente un certain nombre d’avantages, tels la haute sensibilité, la haute résolution de détection, la large bande passante et la flexibilité de l’élément sensible qui rendent cette technologie très prometteuse pour la réalisation de capteurs de courant flexibles, sans contact, capables de mesurer à la fois les courants continus (DC) et alternatifs (AC).Ce travail de thèse vise à explorer la faisabilité d’un capteur de courant flexible à base de GMI, en portant une attention particulière sur l’impact des paramètres d’influence qui conditionnent largement les solutions de mise en œuvre du capteur.Les effets de la température et des contraintes mécaniques de flexion et de torsion, qui s’appliquent dans un environnement de mesure réel, sont caractérisés en prenant en compte les grandeurs intrinsèques du fil nécessaires à la réalisation d’un capteur industriel. L’impact de la mise en œuvre et du conditionnement électronique vis-à-vis de ces grandeurs d’influence est aussi étudié. Les effets des perturbations magnétiques externes et de l’excentration du conducteur sous test dans la boucle de mesure sont quantifiés et une solution de blindage est proposée. Enfin, le prototype de capteur obtenu à l’issue de ces travaux est présenté, ainsi que ses performances, en dégageant les pistes d’optimisation et d’amélioration. / The GMI effect displays several advantages, such as high sensitivity, high detection resolution and bandwidth, and mechanical flexibility. These advantages predispose this technology to the implementation of flexible contactless current sensors measuring both DC and AC currents.This thesis work aims at exploring the feasibility of a flexible GMI current sensor. A particular attention to the impact of influence parameters which largely condition the design solutions of the sensor has been paid.The effects of temperature and mechanical constraints such as bending and torsion, which apply in a real measuring environment, are characterized by taking into account the intrinsic features which are necessary to the design of the sensor. The impact of the general measuring configuration and electronics are also studied. The effects of magnetic disturbances and of the position of the current-carrying conductor in the measuring loop are quantified and an adequate shielding method is proposed. Finally, the sensor prototype obtained at the end of this work is described with its performances and the possible optimization and enhancement ways.

Ce travail porte sur les techniques de mesure des champs en cavité électromagnétique et plus précisément en chambre réverbérante. En raison de la perturbation induite sur la distribution du champ au sein d'une cavité résonante par la présence d'un objet, les techniques de mesure de champ classiques utilisant une antenne souffrent d'une précision limitée. Par conséquent, nous proposons une nouvelle technique de mesure de la distribution du champ électrique basée sur la théorie des perturbations. Elle consiste à mesurer les variations de la fréquence de résonance de la cavité pour chaque position de l'élément perturbateur introduit dans la cavité, puis à en déduire la variation de l'amplitude du champ électrique. Le choix de la forme de l'objet perturbateur, de ses dimensions et de son matériau constitutif est effectué à partir des résultats des simulations et des mesures dans un cas canonique, de façon à adapter le banc de mesure au cas étudié. Cette technique de mesure est ensuite appliquée avec succès au cas d'une chambre réverbérante équipée d'un brasseur de modes, ainsi qu'à des mesures de champ à l'intérieur d'un boitier inséré dans la cavité. Cette approche a permis, via un post-traitement basé sur l'utilisation de l'algorithme MUSIC, de déterminer avec une grande précision les directions d'arrivée des champs dans la chambre réverbérante / This work deals with field measurement techniques in large electromagnetic enclosures namely reverberation chambers. Due to the perturbation of the field distribution within a resonant cavity due to the presence of an introduced object, conventional field measurement techniques employing an antenna suffer from a limited accuracy. Therefore we propose a new measurement technique of the electric field distribution based on the perturbation theory; it consists of a measure of the cavity resonant frequency variation when displacing a small perturbing object within the cavity, and leads to the electric field distribution. The choice of the perturbing object shape, dimension and material is discussed with the help of simulation and measurement results in a canonical case in order to adapt the measurement setup to the studied case. This technique is then successfully employed in a reverberation chamber equipped with a mode stirrer, as well as to measure the field within a metallic box placed in the cavity. Using a post-processing based on MUSIC algorithm, this approach has permitted to determine accurately the field directions-of-arrival in the reverberation chamber

Neste trabalho é apresentado um novo método de medição de descargas parciais, através de transformadores de corrente de alta frequência, instalados no condutor de aterramento dos equipamentos do sistema elétrico de potência e acoplados a um osciloscópio. Esse novo método de medição tem como grande ganho o fato dos equipamentos não precisarem ser retirados de operação para realização de medição. Pelo fato das transmissoras serem remuneradas pela disponibilidade das funções transmissão, o diagnóstico dos equipamentos em serviço faz com que os custos sejam reduzidos. Foi realizado um levantamento teórico onde foi definida a curva característica esperada como resposta do transformador de corrente a um sinal impulsivo. Foram levantadas curvas características, em laboratório, através de oscilogramas no domínio do tempo, das assinaturas de descargas parciais em equipamentos com isolação a óleo e a gás SF6, de modo a validar o sinal esperado na fundamentação teórica. Essas assinaturas foram utilizadas em medições em campo, que comprovaram a viabilidade da fundamentação teórica e das assinaturas obtidas em laboratório. Adicionalmente foram levantados fatores que podem influenciar nas medições e mostrados sinais que possuam características diferentes daquelas das descargas parciais, de modo a diferencia-los. / This work presents a new method of measuring partial discharges, through high frequency current transformers, installed in the equipment grounding conductor of the electrical power system and coupled to an oscilloscope. This new measurement method has as great gain, the fact that the equipment doesnt need to be removed from service to perform measurement. Because of the transmission being remunerated by the availability of the transmission functions, diagnosis of equipment in service makes the costs been reduced. A theoretical survey where the characteristic curve expected response from the current transformer to an impulsive signal was set has been done. Characteristic curves were raised in the laboratory through oscillograms in the time domain, of signatures of partial discharges in equipment with SF6 gas and oil isolation, in order to validate expected signal discharges in the theoretical foundation. These signatures were used in field measurements, which proved the feasibility of the theoretical foundation and signatures obtained in the laboratory. Further factors were raised that may influence the measurements and shown signs that have different characteristics from those of partial discharges in order to differentiate them.

Soil-steel composite bridge refers to structures where a buried flexible corrugated steel pipe works in composite action with the surrounding soil. These structures are being increasingly used in road and railway projects as an alternative to standard type bridges, e.g. short- and medium span concrete beam- and portal frame bridges. On account of their economic advantage and short and easy construction operation, soil-steel composite bridges are getting more popular as railway crossings located far from the cities at the heart of the nature. In this research, the dynamic behaviour of soil-steel composite bridges under high-speed train passages is studied. The studied case is a short span soil-steel composite railway bridge located in Märsta close to Stockholm. The behaviour of the bridge is first observed through field measurements in terms of deflections, stresses, and accelerations at several locations on the bridge. The measured responses are then analysed in order to predict the properties of the soil and steel material working in composite action. Subsequently, 2D and 3D finite element models are developed in order to simulate the behaviour of the bridge. The models are calibrated using the field measurements through several parametric studies. The 3D-model also enables estimation of the load distribution, which is found to increase at higher train speeds. An effective width to be used in 2D analyses is proposed. Finally, the response of the bridge is studied under high-speed train models according to Eurocode.

This study performed detailed measurements on and around scaled up conducting and adiabatic airfoils with and without film cooling. The conducting vane was a matched Bi airfoil, which accurately scaled the convective heat transfer and conduction through the solid, in order to produce non-dimensional surface temperatures and thermal boundary layers that were representative of an actual engine. Measurements made on all vane models included surface temperature measurements and thermal profiles above the walls. Separate measurements on non-film cooled and film cooled conducting models allowed for the individual contributions of the internal convective cooling and external film cooling to the overall cooling scheme to be quantified. Surface temperature and thermal field measurements above the wall were also performed on a film cooled adiabatic model. For the conducting model with internal cooling only, strong streamwise temperature variations were seen. The surface temperature variations were highly dependent on the local external and internal heat transfer coefficients. Spanwise temperature variations also existed, but were modest in comparison to streamwise variations. Comparing the thermal fields above the film cooled adiabatic and conducting walls allowed for the assumption that the conducting wall would not significantly affect the thermal field in the film cooling jet to be tested. Near the edge of the film cooling jet the developing thermal boundary layer had a clear effect on the overlying gas temperature, suggesting that the common assumption that the adiabatic wall temperature is the appropriate driving temperature for heat transfer to a film cooled wall was invalid. On the jet centerline thermal boundary layer effects were less influential, due to the development of a new, thin boundary layer. This suggested that the adiabatic wall temperature as driving temperature for heat transfer was a reasonable assumption on the jet centerline for most cases tested. As film cooling momentum flux ratio increase, thermal boundary layer effects became more influential on the jet centerline. Additionally, the high resolution surface temperature measurements and thermal field measurements above the wall presented in the current study represent a significant improvement in the data available for validation of computational simulations of conducting turbine airfoils. / text

Deep excavations have been used worldwide for underground construction, but they also alter the ground conditions and induce ground movements which might cause risks to adjacent infrastructure. Field measurements are normally carried out during excavations to ensure their safety, and also provide valuable data to calibrate the results from the numerical analysis which is an effective way to investigate the performance of deep excavations. This thesis is concerned with evaluating the capability of advanced finite element analysis in reproducing various aspects of observed deep excavation behaviour in the field through back analysis of case histories. The finite element model developed considers both geotechnical and structural aspects such as (i) detailed geometry of the excavation and retaining structures, (ii) realistic material models for the soil, structures and the soil-structure interface, and (iii) correct construction sequences. Parametric studies are conducted first based on a simplified square excavation to understand the effect of several important aspects, e.g. (i) the merit of shell or solid elements to model the retaining wall, (ii) the effect of construction joints in the retaining wall, (iii) the effect of the operational stiffness of concrete structural components due to cracks, (iv) the thermal effect of concrete beams and floor slabs during curing process and due to variation of ambient temperature, (v) the effect of soil-structure interface behaviour, and (vi) the effect of stiffness and strength properties of the soil. Two more complex case histories are then investigated through fully 3D analyses to explore the influence of various factors such as (i) neglecting the small-strain stiffness nonlinearity in the soil model, (ii) the selected K_0 value to represent the initial stress state in the ground, (iii) the appropriate anisotropic wall properties to consider the joints in the diaphragm wall, (iv) the parameters governing the settlements of adjacent buildings and buried pipelines, (v) the effectiveness of ground improvement on reducing the building settlement, (vi) the variation of construction sequences, (vii) the effectiveness of earth berms, and (viii) ignoring the openings in the floor slabs. This research has strong practical implications, but cautions should also be taken in applications, e.g. element types and parameter selection.

No description available.

La mesure de la dose ionisante est aujourd'hui une tâche cruciale pour une large gamme d'applications fonctionnant dans des environnements de rayonnement sévères. Dans le contexte de l'amélioration de la luminosité du grand collisionneur de hadrons (LHC), la mesure des niveaux de rayonnement le long du complexe d'accélérateurs du CERN va devenir encore plus difficile. A cet effet, une connaissance plus détaillée du champ de rayonnement dans le tunnel de l'accélérateur et ses zones adjacentes devient nécessaire pour définir les exigences d'installation, de déplacement ou de blindage de l'électronique sensible au rayonnement. Dans l’objectif d’améliorer la mesure de la dose absorbée par les systèmes exposés au champ de rayonnement mixte généré par l’accélérateur, des investigations sur des nouveaux dosimètres ont été menées.Dans le cadre de cette recherche, deux dispositifs ont été étudiés et caractérisés pour être utilisés comme dosimètres et éventuellement pour compléter l'utilisation du dosimètre au silicium actuellement utilisé au CERN, à savoir le RADFET (RADiation-sensitive Field Effect Transistor) : un NMOS commercial et un ASIC (Application-specific Integrated Circuit) nommé FGDOS. Les dispositifs ont été sélectionnés selon deux approches opposées : d'une part, la réduction des coûts permettrait d'augmenter la densité des capteurs déployés. En conséquence directe, une carte des doses plus détaillée serait obtenue pour les grands systèmes distribués comme le LHC. D'autre part, la dosimétrie peut être améliorée en déployant des détecteurs plus sensibles, ce qui permettrait de mesurer la dose lorsque les niveaux sont trop faibles pour le RADFET. De plus, des capteurs à plus haute résolution permettraient de caractériser le champ de rayonnement dans un temps plus court, c'est-à-dire avec une luminosité intégrée plus faible.La première approche a été réalisée en recherchant des solutions alternatives basées sur des dispositifs COTS (Commercial Off-The-Shelf), qui réduiraient considérablement les coûts et garantiraient une disponibilité illimitée sur le marché. À cette fin, des recherches ont été menées sur un transistor NMOS discret commercial, qui s'est révélé très sensible au rayonnement.La nécessité d'améliorer la résolution de la mesure de dose a conduit à étudier le FGDOS, un dosimètre en silicium innovant à très haute sensibilité qui permet de détecter des doses extrêmement faibles.La calibration du transistor NMOS et du FGDOS a été effectuées en exposant les dosimètres à des rayons gamma. Leur réponse au rayonnement a été caractérisée en termes de linéarité, de variabilité d'un lot à l'autre et d'effet du débit de dose. L'influence de la température a été étudiée et une méthode pour compenser l'effet de la température a été développée et mise en œuvre.Le FGDOS étant un système sur puce (SoC) avec plusieurs caractéristiques qui font du dosimètre un système extrêmement flexible, la caractérisation de ses différents modes de fonctionnement (actif, passif et autonome) a été effectuée. Suite à la première caractérisation, des questions se sont posées concernant les mécanismes de dégradation de la sensibilité affectant le dosimètre. Pour étudier ce phénomène, des campagnes d’irradiations ont été effectuées avec une puce d'essai incorporant seulement le circuit sensible au rayonnement du FGDOS. L'analyse des expériences a permis de comprendre les processus responsables de la dégradation de la sensibilité, en séparant la contribution du transistor de lecture de celle du condensateur à grille flottante. Les résultats de cette étude nous ont amenés à envisager de nouvelles solutions de conception et des méthodes de compensation.L’aptitude du transistor NMOS et du FGDOS à mesurer la dose ionisante dans les champs de rayonnement mixtes produits par le complexe d’accélérateurs du CERN a été vérifiée à l’aide de test radiatifs accélérés effectués dans le centre de tests en champs mixte à haute énergie du CERN (CHARM). / The Total Ionizing Dose (TID) monitoring is nowadays a crucial task for a wide range of applications running in harsh radiation environments. In view of the High-Luminosity upgrade for the Large Hadron Collider, the monitoring of radiation levels along the CERN’s accelerator complex will become even more challenging. To this extent, a more detailed knowledge of the radiation field in the accelerator tunnel and its adjacent areas becomes necessary to design installation, relocation or shielding requirements of electronics sensitive to radiation. Aiming to improve the monitoring of the TID delivered by the mixed radiation field generated within the accelerator system, investigations on new suitable dosimeters have been carried out.With this research, two devices have been studied and characterized to be employed as dosimeter and possibly to complete the use of the silicon sensor currently employed at CERN for TID monitoring, i.e. the RADiation-sensitive Field Effect Transistor (RADFET): a commercial NMOS, and an ASIC (Application-Specific Integrated Circuit) named FGDOS. The devices have been selected following two opposite approaches: on the one hand, reducing the costs would allow the density of the deployed sensors to increase. As a direct consequence, a more detailed dose map would be obtained for large distributed systems like the LHC. On the other hand, the radiation monitoring can be further improved by deploying more sensitive detectors, which would allow to measure the dose where the levels are too low for the RADFET. Moreover, sensors with higher resolution would permit the characterization of the radiation field in a shorter time, which means within a lower integrated luminosity.The first approach has been accomplished by searching for alternative solutions based on COTS (Commercial Off-The-Shelf) devices, which would significantly reduce the costs and guarantee unlimited availability on the market. For this aim, investigations on a commercial discrete NMOS transistor, which was found to be very sensitive to the radiation, has been carried out.The need for improving the resolution of TID monitoring led to investigate the FGDOS, which is an innovative silicon dosimeter with a very high sensitivity that permits to detect extremely low doses.The calibration of the NMOS and the FGDOS have been performed by exposing the dosimeters to γ-ray. Their radiation response has been characterized in terms of linearity, batch-to-batch variability, and dose rate effect. The influence of the temperature has been studied and a method to compensate the temperature effect has been developed and implemented.Being the FGDOS is a System-On-Chip with several features that make the dosimeter an extremely flexible system, the characterization of its operational modes (Active, Passive and Autonomous) have been performed. Following the first characterization, some questions arose concerning the sensitivity degradation mechanisms affecting the dosimeter. To investigate this phenomenon, radiation experiments were performed with a test chip embedding only the radiation sensitive circuit of the FGDOS. The analysis of the experiments allowed the understating of the processes responsible for the sensitivity degradation, by separating the contribution of the reading transistor and the floating gate capacitor. The results of this investigation led us to considerer new design solution and compensation methods.The suitability of the NMOS and the FGDOS for TID measurement in the mixed radiation field produced by the CERN’s accelerator complex has been verified by performing accelerated radiation tests at the Cern High energy AcceleRator Mixed field facility (CHARM). The consistency of both sensors with the RADFET measurement has been demonstrated. The high sensitivity of the FGDOS leads to a significant improvement in terms of TID measurement in mixed radiation fields with respect to the RadFET, especially for low radiation intensities.

No description available.