In this thesis, parametric studies have been performed for a soil-steel compositebridge to determine and investigate the most influential parameters on the dynamicresponse.High-speed railways are currently being planned in Sweden by the Swedish TransportAdministration with train speeds up to 320 km/h. According to the European designcodes, bridges must be verified with respect to dynamic resonance behaviour for trainspeeds exceeding 200 km/h. However, there are no guidelines or design criterion forperforming dynamic verifications of soil-steel composite bridges. The aim of thisthesis has therefore been to investigate the influence of the geometry and materialproperties of soil-steel composite bridges on their dynamic response.This thesis is based upon the frequency domain approach for dynamic analysis ofa soil-steel composite bridge using finite element software. In 2018, field measurementswere performed on a soil-steel composite bridge in Hårestorp, Sweden. Areference finite element model was developed based on previous research and wasverified against these field measurements. Parametric studies where performed byextrapolating the geometry of the reference model, focusing primarily on the crownheight, culvert span width and the location of the bedrock. Sensitivity analyses ofthe density- and stiffness of the soil was also performed.The parametric studies showed that the crown height was the most influential parameterwith respect to the amplitude of the resonance peak. Increasing it from 1 mto 3 m reduced the amplitude by approximately 70 %. An increased span width ofthe culvert was found to reduce the frequency and amplitude of the resonance peak,however increasing the stiffness of the culvert increased the resonance frequency.The position of the rock layer also reduced the amplitude of the resonance peak iflowered, likely because of lessened wave reflection. The lowest rock level investigatedshowed a significant decrease of more than 70 % in amplitude. However, the modelused to calculate this response was heavily extrapolated and thus difficult to verify.The sensitivity analyses showed that the soil density- and stiffness was negativelyand positively correlated with the resonance frequency, respectively. Additionally,the soil density lowered the amplitude of the resonance peak if increased.

The Cardiac Rhythm Management Division of St. Jude Medical specializes in the development of implantable cardioverter defibrillators that improve the quality of life for patients diagnosed with a variety of cardiac arrhythmias, especially for patients prone to sudden cardiac death. With the goal to improve detection of cardiac arrhythmias, this study explored the value in time-frequency analysis of intracardiac electrogram in four steps. The first two steps characterized, in the frequency domain, the waveforms that construct the cardiac cycle. The third step developed a new algorithm that putatively provides the least computationally expensive way to identifying cardiac waveforms in the frequency domain. Lastly, this novel approach to analyzing intracardiac electrogram was compared to a threshold crossing algorithm that strictly operates in the time domain and that is currently utilized by St. Jude Medical. The new algorithm demonstrated an equally effective method in identifying the QRS complex on the ventricular channel. The next steps in pursing time-frequency analysis of intracardiac electrogram include implementing the new algorithm on a testing platform that emulates the latest implantable cardioverter defibrillator manufactured by St. Jude Medical and pursuing a similar algorithm that can be employed on the atrial channel.

Many people experience discomfort during movement when using a knee prosthesis. As a result, a study is being conducted to see if biomedical models can be used to produce an optimal prosthetic socket to reduce discomfort. This has previously been accomplished using pressure sensors embedded in the sleeve to measure leg pressure at various stages of the gait cycle. To gather more information, an EEG should be performed to find where discomfort occurs throughout a gait cycle. The purpose of the EEG is to perform measurements for a certain time during walking to then be able to analyze the signals that occur from the brain's motor cortex. The signal analysis is performed by applying the Fast Fourier transform to convert EEG data from the time domain to the frequency domain. By examining frequency differences, it is demonstrated that additional measurements with more individuals are required to ensure that the results reflect discomfort, and that the EEG is useful in the applied field of use. / Vid användning av knäproteser upplever ett stort antal personer obehag vid rörelse. På grund av detta utförs det en studie där man genom biomedicinska modeller kan producera den optimala proteshylsan för att minska obehag. Detta har tidigare gjorts med trycksensorer placerade i hylsan, för att kunna mäta tryck från benet vid olika faser av en gångcykel. EEG ska användas för att få mer information var det uppstår obehag under en gångcykel. Syftet med EEG är att genomföra mätningar under en viss tid vid gång, för att sedan kunna analysera signalerna från hjärnans motoriska cortex. Signalanalysen sker genom att transformera EEG-data från tidsdomänen till frekvensdomänen med hjälp av Snabb Fourier transform. Genom att undersöka frekvensskillnader påvisas att ytterligare mätningar med fler individer krävs för att säkerställa att resultaten återspeglar obehag och att EEG är användbart inom det tillämpade användningsområdet. / This work was supported by the research project SocketSense (https://www.socketsense.eu/), funded by the European Union’s Horizon 2020 research and innovation programme under grant agreement No 825429

No description available.

Made available in DSpace on 2016-03-15T19:37:31Z (GMT). No. of bitstreams: 1 Rodrigo Cesar Ribeiro Miraglia.pdf: 2147088 bytes, checksum: 67087392fe4b03bb24d9a164fd92ceed (MD5) Previous issue date: 2010-08-20 / Fundo Mackenzie de Pesquisa / Sensors based on optical fibers are being increasingly used in hostile environments for measuring pressure, temperature, stress, chemical and biological parameters, etc. These sensors have the advantage of having reduced size and weight, immunity to electromagnetic interference, of being chemically inert, and also allowing the distributed measurement of the respective parameter along the fiber. Microstructured optical fibers have characteristics that are relevant to sensing applications, such as freedom of design of their internal structure obtained by varying the diameter of the holes and the distances between them, among other parameters, which is not applicable in conventional optical fibers. The present work aims to use a microstructured optical fiber as a distributed sensor for hydrostatic pressure, using the POTDR (Polarization Optical Time-Domain Reflectometry) and the OFDR (Optical Frequency-Domain Reflectometry) measurement techniques. The application of hydrostatic pressure on the fiber changes its birefringence and, consequently, the evolution of the states of polarization, which is the parameter monitored by the techniques. By monitoring the changes of these states, it is possible to infer the changes in applied pressure. The study is undertaken via simulations and analysis, both performed in the software MatLab. / Sensores baseados em fibras ópticas vêm sendo cada vez mais utilizados em ambientes hostis para medição de pressão, temperatura, stress, análise química e biológica, etc. Esses sensores têm a vantagem de possuírem tamanho e peso reduzidos, imunidade à interferência eletromagnética, de serem quimicamente inertes, e também permitirem a medição distribuída do parâmetro a ser analisado ao longo da fibra. As fibras ópticas microestruturadas possuem características relevantes na aplicação de sensoriamento, tal como a liberdade de construção de sua estrutura interna variando o diâmetro dos buracos e a distância entre eles, entre outros parâmetros, que não são aplicáveis em fibras ópticas convencionais. Este trabalho tem como objetivo estudar uma fibra óptica microestruturada como sensor distribuído de pressão hidrostática, utilizando as técnicas de medição POTDR (Polarization Optical Time-Domain Reflectometry) e OFDR (Optical Frequency-Domain Reflectometry). A aplicação de pressão hidrostática sobre a fibra altera sua birrefringência e, consequentemente, a evolução do seu estado de polarização, sendo este último o parâmetro monitorado pelas técnicas. Analisando-se a mudança deste estado, pode-se inferir a mudança de pressão aplicada. O estudo é realizado através de simulações e análises efetuadas no software MatLab.

Transients encountered by transformers in power stations during normal operation can have complex oscillatory overvoltages containing a large spectrum of frequency components. These transients can coincide with the natural frequencies of the transformers windings, leading to voltages that can be greater or more severe than the current factory proof tests. This may lead to insulation breakdown and catastrophic failures. Existing lumped parameter RLCG transformer models have been proven to be less accurate for very fast transient overvoltages (VFTO) with frequencies over 1 MHz. A white box model for transient analysis of transformer windings has been developed using Multiconductor Transmission Line (MTL) Theory. This model enables the simulation of natural frequencies of the transformer windings up to frequencies of several MHz, and can be used to compute voltages between turns by representing each turn as a separate transmission line. Both continuous and interleaved disk windings have been modelled and a comparison and validation of the results is presented. / May 2017

La matière dense et tiède (WDM) est un régime caractérisé par une densité proche du solide pour une température avoisinant celle de Fermi. Pour étudier cet état de la matière, dans cette thèse, une expérience d'interférométrie dans le domaine des fréquences est mise en place afin de mesurer la phase et la réflectivité - dans les deux directions de polarisation S et P - d'une onde sonde en réflexion sur un échantillon chauffé de manière très brève par une impulsion laser ultra-courte (sub-100fs). Il est alors porté dans un état hors-équilibre. Une méthode basée sur les mesures de réflectivité est mise en place pour contrôler la forme de l'interface entre le vide et la matière chauffée. Pour des fluences laser de l'ordre de 1 J/cm2, l'hydrodynamique d'un échantillon chauffé est étudiée par la mesure du déplacement de la surface et comparée au code hydrodynamique à deux températures ESTHER. Ensuite, la fonction diélectrique à 800 nm et 400 nm est déduite des mesures expérimentales et certaines quantités en sont extraites comme la densité électronique, la température électronique et les fréquences de collision en régime WDM. Elles sont par la suite comparées avec des modèles couramment utilisés. / The Warm Dense Matter (WDM) regime is characterised by a density close to the solid density and an electron temperature close to the Fermi temperature. In this work, the nonequilibrium Warm Dense Matter is studied during the solid to liquid phase transition induced by an ultra short laser interacting with a solid. A 30 femtoseconds time resolution pump-probe experiment (FDI) is set up, yielding to the measurement of the heated sample complex reflectivity for both S and P polarisation.We have determined a criterion based on the measured reflectivities, which permits to control the interface shape of the probed matter. For pump laser fluences around 1 J/cm2, the hydrodynamics of the heated matter is studied and experimental results are compared to the two-temperatures code ESTHER. Furthermore, the evolution of the dielectric function at 800 nm and 400 nm is inferred from our measurements on a sub-picosecond time-scale. Within the Drude-Lorentz model for the conduction electrons, the dielectric function yields information such as ionisation state, electronic temperature and electron collision frequency.

L'objet de cette thèse est d'étudier les systèmes de cointégration fractionnaire de forme triangulaire mais également d'analyser l'apport de ces systèmes dans la modélisation des co-mouvements au sein des marchés financiers internationaux. La thèse s'articule autour de six chapitres équitablement répartis entre contributions économétriques et économiques. Concernant l'approche économétrique, un intérêt particulier est donné à l'estimation de ces systèmes en absence d'information sur les paramètres d'intérêts. Dans cette optique, plusieurs techniques d'estimation sont analysées et développées, essentiellement dans le domaine des fréquences car celui-ci permet un traitement semi-paramétrique des paramètres de nuisances. Les performances de ses estimateurs sont étudiés à travers des simulations mais également à travers l'étude des propriétés asymptotiques. Concernant l'approche économique, une première contribution exploite la cointégration fractionnaire pour révéler l'existence d'un système de taux de change entre certains pays Asiatiques. Une deuxième contribution porte sur l'analyse des interdépendances entre le marché du pétrole et divers taux de change au niveau de la volatilité. Une troisième contribution introduit un processus d'apprentissage adaptatif dans un modèle monétaire à plusieurs pays afin d'étudier sous quelles conditions un système de taux de change peut émerger. / The aim of the thesis is to study a triangular form of fractional cointegration systems and to investigate whether these systems allow to model the comovements in international financial markets. The thesis is organized around six chapters. Three of them are theory-oriented and the three others are empirics-oriented. Concerning the econometric approach, a particular interest is devoted to the estimation of these systems when all parameters of interest are unknown. To this extent, several estimation techniques are investigated and introduced, essentially in frequency domain as it allows a semi-parametric treatment of the nuisance parameters. Most of times, the performance of these estimators are studied by means of simulations but the asymptotic theory is also developed. Concerning the economic approach, a first contribution applies the fractional cointegration theory to reveal the existence of an exchange rate system between several Asian countries. A second contribution deals with the risk interdependences between the crude oil market and several exchange rates. A third contribution considers an adaptive learning mechanism in a multi-country monetary model to investigate the conditions under which an exchange rate system is likely to emerge.

L'équation intégrale du champ électrique (EFIE) est très utilisée pour résoudre des problèmes de diffusion d'ondes électromagnétiques grâce à la méthode aux éléments de frontière (BEM). En domaine fréquentiel, les systèmes matriciels émergeant de la BEM souffrent, entre autres, de deux problèmes de mauvais conditionnement : l'augmentation du nombre d'inconnues et la diminution de la fréquence entrainent l'accroissement du nombre de conditionnement. En conséquence, les solveurs itératifs requièrent plus d'itérations pour converger vers la solution, voire ne convergent pas du tout. En domaine temporel, ces problèmes sont également présents, en plus de l'instabilité DC qui entraine une solution erronée en fin de simulation. La discrétisation en temps est obtenue grâce à une quadrature de convolution basée sur les méthodes de Runge-Kutta implicites.Dans cette thèse, diverses formulations d'équations intégrales utilisant notamment des conditions d'impédance aux frontières (IBC) sont étudiées et préconditionnées. Dans une première partie en domaine fréquentiel, l'IBC-EFIE est stabilisée pour les basses fréquences et les maillages denses grâce aux projecteurs quasi-Helmholtz et à un préconditionnement de type Calderón. Puis une nouvelle forme d'IBC est introduite, ce qui permet la construction d'un préconditionneur multiplicatif. Dans la seconde partie en domaine temporel, l'EFIE est d'abord régularisée pour le cas d'un conducteur électrique parfait (PEC), la rendant stable pour les pas de temps larges et immunisée à l'instabilité DC. Enfin, unerésolution efficace de l'IBC-EFIE est recherchée, avant de stabiliser l'équation pour les pas de temps larges et les maillages denses. / The Electric Field Integral Equation (EFIE) is widely used to solve wave scattering problems in electromagnetics using the Boundary Element Method (BEM). In frequency domain, the linear systems stemming from the BEM suffer, amongst others, from two ill-conditioning problems: the low frequency breakdown and the dense mesh breakdown. Consequently, the iterative solvers require more iterations to converge to the solution, or they do not converge at all in the worst cases. These breakdowns are also present in time domain, in addition to the DC instability which causes the solution to be completely wrong in the late time steps of the simulations. The time discretization is achieved using a convolution quadrature based on Implicit Runge-Kutta (IRK) methods, which yields a system that is solved by Marching-On-in-Time (MOT). In this thesis, several integral equations formulations, involving Impedance Boundary Conditions (IBC) for most of them, are derived and subsequently preconditioned. In a first part dedicated to the frequency domain, the IBC-EFIE is stabilized for the low frequency and dense meshes by leveraging the quasi-Helmholtz projectors and a Calderón-like preconditioning. Then, a new IBC is introduced to enable the development of a multiplicative preconditioner for the new IBC-EFIE. In the second part on time domain,the EFIE is regularized for the Perfect Electric Conductor (PEC) case, to make it stable in the large time step regime and immune to the DC instability. Finally, the solution of the time domain IBC-EFIE is investigated by developing an efficient solution scheme and by stabilizing the equation for large time steps and dense meshes.

<p>Diffusion is a phenomenon in which particles move from regions of higher density to regions of lower density. Many physical systems, in fields as diverse as plant biology and finance, are known to involve diffusion phenomena. Typically, diffusion systems are modeled by partial differential equations (PDEs), which include certain parameters. These parameters characterize a given diffusion system. Therefore, for both modeling and simulation of a diffusion system, one has to either know or determine these parameters. Moreover, as PDEs are infinite order dynamic systems, for computational purposes one has to approximate them by a finite order model. In this thesis, we investigate these two issues of model reduction and parameter estimation by considering certain specific cases of heat diffusion systems. </p><p>We first address model reduction by considering two specific cases of heat diffusion systems. The first case is a one-dimensional heat diffusion across a homogeneous wall, and the second case is a two-dimensional heat diffusion across a homogeneous rectangular plate. In the one-dimensional case we construct finite order approximations by using some well known PDE solvers and evaluate their effectiveness in approximating the true system. We also construct certain other alternative approximations for the one-dimensional diffusion system by exploiting the different modal structures inherently present in it. For the two-dimensional heat diffusion system, we construct finite order approximations first using the standard finite difference approximation (FD) scheme, and then refine the FD approximation by using its asymptotic limit.</p><p>As for parameter estimation, we consider the same one-dimensional heat diffusion system, as in model reduction. We estimate the parameters involved, first using the standard batch estimation technique. The convergence of the estimates are investigated both numerically and theoretically. We also estimate the parameters of the one-dimensional heat diffusion system recursively, initially by adopting the standard recursive prediction error method (RPEM), and later by using two different recursive algorithms devised in the frequency domain. The convergence of the frequency domain recursive estimates is also investigated. </p>