Le besoin en isolation vibratoire des voies ferrées conduit de plus en plus souvent au recours à des systèmes d'attache de rail souples avec des designs sophistiqués. Par contre, les produits deviennent diversifiés mais leur conception se fait de manière empirique dans la pratique. Dans ce travail de thèse, deux systèmes d'attache de rail sont étudiés expérimentalement et numériquement: un système qui travaille principalement en cisaillement et l'autre en compression. Des essais quasi-statiques et dynamiques sont menés puis les résultats sont analysés à l'aide de différents modèles mécaniques. En plus, une série de tests sur les élastomères utilisés dans les deux systèmes sont réalisés. Ces tests permettent de choisir des modèles hyperélastiques et viscoélastique appropriés, et de déterminer les paramètres de propriétés mécaniques des élastomères dans ces modèles par des simulations Abaqus. En utilisant également Abaqus, des analyses par élément fini des deux systèmes d'attache sont effectués et les résultats sont comparés aux ceux des essais. Des non-linéarités d'origine matérielle et géométrique sont ainsi analysées et expliquées. Enfin, des études de sensibilité des paramètres sont effectuées, suivies d'études d'optimisation structurale. Cet étude devrait être utile pour une meilleure compréhension du comportement mécanique des systèmes d'attache de rail sous chargement statique et dynamique, et au final pour une meilleure pratique dans la conception et l'optimisation des produits / Higher demand on vibration isolation of track structure in modern railway track leads to a trend of lower stiffness of rail fastening systems, places an increasing need for better design approach as well. However the current development status of rail fastening industry is dramatically diversified but mainly empirical. In this work, a shear type and a bonded compressed type of rail fastening systems are investigated experimentally and numerically. Quasi-static and dynamic experiments are carried out and the results are analysed with different mechanical models. Besides, a series of fundamental rubber material tests is performed to accurately describe the rubber material used in the two fastening systems. The most appropriate hyperelastic and hysteresis models are chosen with the help of simulations by Abaqus. Also by using Abaqus, finite element analyses are conducted for the two fastening systems and the numerical results are compared with those of experiments. Material and geometrical nonlinear features observed in the measured displacement-force curves are analysed and explained. Finally, parameter sensitivity of the two fastening specimens are studied, followed by an optimization process to meet practical optimization objectives. The present work is believed to be helpful for understanding the mechanical behavior of rail fastening systems, while enlightening the engineering practice, and eventually improving product designing and optimizing measures

Submitted by Renata Lopes (renatasil82@gmail.com) on 2015-12-16T11:40:36Z No. of bitstreams: 1 thiagoribeirofurtadodemendonca.pdf: 2671419 bytes, checksum: 9751de9437139252fdb339ca32b4f21d (MD5) / Approved for entry into archive by Adriana Oliveira (adriana.oliveira@ufjf.edu.br) on 2015-12-16T11:57:40Z (GMT) No. of bitstreams: 1 thiagoribeirofurtadodemendonca.pdf: 2671419 bytes, checksum: 9751de9437139252fdb339ca32b4f21d (MD5) / Made available in DSpace on 2015-12-16T11:57:41Z (GMT). No. of bitstreams: 1 thiagoribeirofurtadodemendonca.pdf: 2671419 bytes, checksum: 9751de9437139252fdb339ca32b4f21d (MD5) Previous issue date: 2015-07-13 / CAPES - Coordenação de Aperfeiçoamento de Pessoal de Nível Superior / Em face ao aumento da demanda por energia elétrica e maior confiabilidade em seu fornecimento, diversas metodologias de controle e proteção da rede elétrica estão sendo desenvolvidas. Para isso é de fundamental importância a utilização de técnicas de processamento de sinais que permitam a extração de forma precisa e rápida dos parâmetros da rede elétrica. O advento de novas tecnologias permitiu a modernização dos sistemas de geração, transmissão e distribuição de energia elétrica, melhorando a eficiência energética e confiabilidade. Por outro lado, observou-se a proliferação na indústria de novas cargas, em sua maioria de natureza não linear, responsáveis pela geração de componentes harmônicos e inter-harmônicos que distorcem o sinal do sistema elétrico (tensão e corrente), dificultando a estimação de parâmetros importantes. Com o intuito de continuar provendo energia de forma segura e confiável, alguns pré-requisitos devem ser atendidos, tais como a baixa distorção harmônica, amplitude estável, baixa oscilação da frequência fundamental, entre outros. Hoje em dia, devido ao maior nível de cargas e equipamentos não lineares conectados a rede, rastrear estes parâmetros de qualidade de energia não é uma tarefa fácil. Para isso, novas técnicas de processamento de sinais devem ser combinadas às teorias de sistemas de potência como forma de estimar e acompanhar as mudanças destes parâmetros. Nesse contexto, diversos algoritmos vêm sendo implementados, cada um com suas especificidades e desempenho. O presente trabalho propõe o desenvolvimento de um novo filtro baseado no filtro Hanning, que é capaz de eliminar harmônicos e atenuar inter-harmônicos com maior eficiência além de ser mais estável quando se considera a implementação em ponto fixo. Este filtro foi combinado com um algoritmo capaz de ajustar seus coeficientes de acordo com a estimação da frequência fundamental de forma a garantir amostragem síncrona, ou seja, a frequência de amostragem será sempre múltipla inteira da frequência atual. Dessa forma, a estimação de amplitude e reconstrução de componentes são feitas com maior precisão mesmo quando se considera a natureza variante da frequência fundamental do sinal bem como a maior quantidade de harmônicos e inter-harmônicos presentes, retratando com maior fidelidade o sinal real do sistema de potência. Os resultados mostram-se superiores às técnicas já existentes, considerando sinais com frequências fora da nominal bem como com grande quantidade de componentes harmônicos e inter-harmônicos, atingindo uma melhoria de 74% em relação ao MAF e de quase 90% em relação ao Hanning convencional. / Due to the growing demand for energy provided in a more reliable way, several control, protection and monitoring methodologies are been developed. For this reason it is of great importance the utilization of appropriate signal processing techniques that allow a precise and fast extraction of parameters from the electrical grid. The advent of new technologies has enabled the modernization of the electric energy generation, transmission and distribution systems, improving its efficiency and service reliability. On the other hand, it was noted a widespread in the application of non-linear loads, responsible for generating harmonics and inter-harmonics components that distorts the electrical signal (voltage and current), hindering the estimation process of useful parameters. With the aim to keep providing energy in a safe and reliable way, some prerequisites must be taken into account, such as the low harmonic distortion, stable amplitude, low oscillation of the fundamental frequency among others. Nowadays, due to the widespread use of nonlinear loads and equipment coupled to the grid, tracking these power quality parameters is not an easy task. That is why novel approaches and techniques of signal processing must be combined with knowledge of power systems theory as a form to estimate and keep track of these measurements. Within this context several algorithms are been developed, each one with its specificity and performance. The present work proposes the development of a new filter, based on the Hanning filter, capable of eliminating harmonics and attenuate with more efficiency the inter-harmonic components besides being more stable for fixed point implementation. This filter is combined with an algorithm that enable the adjustment of its coefficients according to the frequency estimation, assuring a synchronous sampling process, that is, the sampling rate will always be an integer multiple of the actual system frequency. In this way, the amplitude estimation and component reconstruction are performed with better accuracy even when considering the time-varying nature of the fundamental frequency of the signal and the high amount of distortion due to harmonics and inter-harmonics, depicting with greater fidelity the true signal of power system. Results have shown better performance than the already known techniques, considering signal with off-nominal frequency as well as high level of harmonic and inter-harmonic distortion, reaching 74% of improvement in relation with MAF and almost 90% in relation of conventional Hanning.

De nos jours, le nombre d'appareils connectés nécessitant un accès aux données mobiles est en augmentation constante. L'arrivée d'encore plus d'ojects multimédias connectés et la demande croissante d'informations par appareil ont mis en évidence les limites de la quatrième génération de réseaux cellulaires (4G). Cela a poussé au développement de nouvelles méthodes, dont la 5G. L'objectif est d'être en mesure de prendre en charge la croissance des systèmes portables, des capteurs ou des sysèmes associés à l'internet des objets (IoT). La vision derrière la 5G est de permettre une société entièrement mobile et connectée avec une expérience consistente.Les petites cellules sont la base des normes de communication avancées telles que 4G et maintenant 5G. Ils résultent de l’utilisation de bandes de fréquences plus élevées pour l’accès radiofréquences (RF) afin de supporter de nouvelles normes et exigences croissantes en bande passante. La 5G utilise des ondes millimétriques et nécessite un déploiement dans un environnement urbain intérieur et urbain dense, ce qui peut s'avérer être un défi. C’est là que la 5G devra inclure des solutions de réseau hybrides et pouvoir coexister avec d’autres technologies d’accès sans fil. La communication par lumière visible (VLC) s’inscrit dans ce moule puisque la lumière visible correspond à la bande comprise entre 400 et 800 THz. Le spectre disponible est des milliers de fois plus large que le spectre RF et il n’interfère pas avec celui-ci. Le principe se base sur la combinaison de l'éclairage avec un lien de communication pouvant atteindre des dizaines de gigabits par seconde. Le potentiel est d’offrir un complément à la 5G dans un réseau hybride, offrant une vitesse élevée, aucune interférence et une sécurité accrue au prix d’une couverture limitée et d’une faible maturité technologique.L’objectif de cette thèse est donc de proposer et d’évaluer une implémentation expérimentale d’un système VLC en intérieur et multi-utilisateurs afin de répondre aux objectifs de la configuration light-fidelity (Li-Fi) dans le contexte d’une petite cellule. La première étape de cette étude est un état de l'art détaillé sur le principe de VLC dans la communication sans fil en intérieur et de l’accès multi-utilisateur. Cela permet de mieux expliquer le concept de notre désign et de comparer notre approche aux travaux existants. La deuxième étape consiste en une analyse des principes et des hypothèses pour le système VLC multi-utilisateurs en intérieur portant à la fois sur la technique de modulation et sur les schémas d’accès multi-utilisateurs. Les conclusions tirées des analyses théoriques et numériques servent de base pour la suite du travail. La troisième étape consiste en plusieurs analyses expérimentales sur l'optimisation des performances de diffusion pour un utilisateur unique, puis sur les performances multi-utilisateurs du système à l'aide de divers schémas d'accès. Le débit total avec une LED blanche commerciale atteint 163 Mb/s avec un taux d'erreur réduit d'un facteur de 3,55 grâce au processus d'optimisation des performances. Cette technique a l'avantage d'augmenter la flexibilité pour un scénario avec plusieurs utilisateurs sans augmenter la complexité car seuls les paramètres des filtres de modulation sont altérés. La taille de la cellule obtenue est de 4.56 m² à une distance de 2,15 mètres du transmetteur. Le capacité peut atteindre jusqu'à 40 utilisateurs, ou 40.62 Mb/s dans un scénario à 4 utilisateurs. Il est donc démontré que le système proposé pourrait fonctionner comme une cellule à une distance réaliste, avec un débit de données élevé et la capacité de répondre aux besoins d’un grand nombre d’utilisateurs tout en limitant les coûts de mise en œuvre. / Nowadays, the number of connected devices requiring access to mobile data is considerably increasing. The arrival of even more connected multimedia objects and the growing demand for more information per device highlighted the limits of the fourth generation of broadband cellular networks (4G). This pushed for the development of new methods, one of which is 5G. The goal is to be able to support the growth of wearable, sensors, or related internet-of-object (IoT) systems. The vision behind 5G is to enable a fully mobile and connected society with a consistent experience. In consequence, there is a fundamental need to achieve a seamless and consistent user experience across time and space.Small cells are the basis of advanced communications standards such as 4G and now, 5G. They exist as a result of using higher frequency bands for RF access in order to support new standards and the increasing demands in bandwidth. 5G use millimeter waves and requires a deployment across indoor and dense urban environment which may prove to be a challenge. This is where 5G will need to include hybrid networking solutions and be able to coexist with other wireless access technologies. Visible light communication (VLC) fits into that mold since visible light corresponds to the band between 400 and 800 THz. The available spectrum is multiple thousand times the size of the RF spectrum and it does not interfere with it. The technique combines illumination with communication at possibly tens of gigabits per second. It has the potential to offer a synergistic pairing with 5G in a hybrid network, offering high speed, no interferences, and more security at the cost of limited coverage and low technological maturity.The goal of this thesis is thus to propose and evaluate an experimental implementation of an indoor multi-user VLC system in order to answer the objectives of Li-Fi setup in the context of a small cell. The first step of this study is a detailed state-of-the-art on VLC in indoor wireless communication and multi-user access. It allows the design of our work to be better explained and to compare our approach with existing works. The second step is an analysis of the principles and hypothesis supporting the indoor multi-user VLC system in the study both on the modulation technique and the multi-user access schemes. The conclusions drawn from theoretical and numerical analysis are used as a basis for the rest of the work. The third step is the experimental setup investigations on the single-user broadcast performances optimization and then on the multi-user performances of the system using various schemes. The total throughput using an off-the-shelf white LED reaches 163 Mb/s with a bit-error rate decreased by a factor of 3.55 thanks to the performance optimization process. This technique has the advantage of increasing the flexibility for a multi-access scenario while not augmenting the complexity as it only optimizes the modulation filter parameters. The multi-user access is obtained for a cell size of 4.56 m² at a distance of 2.15 meter away from the transmitter. The user capacity can reach up to 40 users, or 40.62 Mb/s in a 4-user scenario. It is thus demonstrated that the proposed system could function as a cell at a realistic range, with high data rate and the ability to provide for a large amount of users while limiting the cost of implementation.

Avec le développement rapide des modèles mathématiques et des outils de simulation, le besoin des processus de quantification des incertitudes a été bien augmenté. L'incertitude paramétrique et la groupe des nombreux décisions sont aujourd’hui les deux barrières principales dans la résolution des grandes problèmes systématiques.Capable de proportionner l'incertitude de la sortie sur celle des entrées, l’Analyse de Sensibilité Globale (GSA) est une solution fiable pour la quantification de l’incertitude. Parmi plusieurs algorithmes de GSA, Fourier Amplitude Sensitivity Analysis (FAST) est l’un des choix les plus populaires des chercheurs. Basé sur ANOVA-HDMR (ANalysis Of VAriance - High Dimensional Model Representation), il est solide en mathématique est efficace en calcul.Malheureusement, la décomposition unique d’ANOVA-HDMR se dépend sur l’indépendance des entrées. À cause de cela, il y a pas mal de cas industriels qui ne peut pas se traiter par FAST, particulièrement pour ceux qui donnent uniquement les échantillons mais sans lois de distribution. Sous cette demande, deux méthode extensifs de FAST avec design de corrélation sont proposées et étudiées dans la recherche. Parmi les deux méthodes, FAST-c s’est basé sur les distributions et FAST-orig s’est basé sur les échantillons.Comme applications et validations, multiples problèmes vibroacoustiques se sont traités dans la recherche. Les matériaux acoustiques avec soustructures, sont des candidats parfaits pour tester FAST-c et FAST-orig. Deux application sont présentées dans la première partie de la thèse, après l’état de l’arts. Les modèles choisis sont matérial poroélastique et structures composite sandwich, dont les propriétés mécaniques sont tous fortement influencées par les paramètres géométriques microscopique ou mesoscopique. D’avoir la méthode de FAST originale comparée avec les deux nouvelles, on trouve bien plus d’information sur la performance vibroacoustique de ces matériaux.Déjà répondu à la demande de GSA sur les modèles avecs les variables dépendantes, la deuxième partie de la thèse contient plus de recherches reliées avec FAST. D’abord FAST est pris en comparaison avec Random Forest, une algorithme bien connu de data-mining. Leurs erreurs potentiels et la possibilité de fonctioner ensemble sont discutés. Et dans les chapitres suivies, plus d’application de FAST sont présentées. Les méthodes sont appliquées sous plusieurs différente conditions. Une modèle de structure périodique qui contient des corrélation parmi les unités nous a en plus forcé à développer une nouvelle FAST-pe méthode. Dans ces applications, les designs des processus préliminaires et les stratégies d’échantillonages sont des essences à présenter. / With rapid development of mathematical models and simulation tools, the need of uncertainty quantification process has grown higher than ever before. Parametric uncertainties and overall decision stacks are nowadays the two main barriers in solving large scale systematic problem.Global Sensitivity Analysis (GSA) is one reliable solution for uncertainty quantification which is capable to assess the uncertainty of model output on its inputs’. Among several GSA algorithms, Fourier Amplitude Sensitivity Test (FAST) is one of the most popular choices of researchers. Based on ANOVA-HDMR (ANalysis Of VAriance - High Dimensional Model Representation), it is both mathematically solid and computationally efficient.One unfortunate fact is that the uniqueness of ANOVA-HDMR relies on the independency of input variables. It makes FAST unable to treat many industrial cases especially for those with only datasets but not distribution functions to be found. To answer the needs, two extended FAST methods with correlation design are proposed and further studied in this research. Among them FAST-c is distribution-based and FAST-orig is data-based.As a frame of validation and application, a number of vibroacoustic problems are dealt with in this research. Vibroacoustic materials with substructures, are perfect test candidates for FAST-c and FAST-orig. Two application cases are presented in the first part of this thesis, following the literature review. The models chosen here are poroelastic material and sandwich composite structures, both having their mechanical properties hugely influenced by their microscopic and mesoscopic geometric parameters. Getting the original FAST method compared to the two with correlation design, many different features on materials’ vibroacoustic performance are latter discovered.Having got an answer for GSA on models with dependent variables, the second part of this thesis contains more extended researches related to FAST. It is taken into comparison with Random Forest, a well-known data-mining algorithm. The potential error of both algorithms are analyzed and the possibility of joint application is discussed. In the following chapters, more applications of FAST-series methods are reported. They are applied under various conditions where another improved version named FAST-pe is developed to treat a model of periodic structures with correlation among each units. Upon these FAST application cases, the design of preliminary process and the sampling strategies is the core part to be introduced.

No description available.

Polymer components are used in many different applications, including in industries where critical applications put high requirements regarding quality assessment. Such applications might include medical or food where the presence of discontinuities might induce bacterial growth or other unpleasantries, thus certain manufacturers must be able to maintain a zero-tolerance towards damaged components. This leads to the need for efficient testing methods of nondestructive nature capable of testing large quantities of components in a production line environment. The authors have been tasked by Acoustic Agree AB and Trelleborg AB, a world-leading producer of polymer engineered solutions, to find a nondestructive testing method capable of detecting discontinuities in thin polymer components in a production line environment. Implementation in production line environments puts requirements on test cycle time and a goal is to complete a test cycle within 3-4 seconds. Due to restrictions regarding available equipment and expertise, the focus has been put on applying nonlinear acoustic methods for nondestructive testing instead of more conventional methods. These methods utilize the nonlinear distortion of acoustic waves which causes certain characteristics to appear, such as the generation of Higher Harmonics (HH), frequency modulation (NWMS), resonance frequency shift (NRUS), and amplitude recovery (IDAR). Visual testing was used to discern visibly damaged samples from seemingly undamaged ones. The only methods which showed the possibility of discerning damaged thin polymer components were NRUS and IDAR. Only the latter has the capability to maintain the prescribed test cycle time. Nonlinear acoustic methods seem to be capable of detecting discontinuities in thin polymer components within the given time frame. More work is required to properly investigate the performance of NWMS and IDAR. The configuration used in this work was mainly focused on IDAR, resulting in specific calibration for NWMS was neglected. The sample population was also too low to collect sufficient data to ensure statistical certainty regarding the performance for either method. / Polymerkomponenter används i många olika tillämpningar, i synnerhet i industrier där kritiska tillämpningar ställer höga krav på kvalitetsbedömning. Sådana tillämpningar kan finnas inom medicinska- eller livsmedelsindustrier där förekomsten av diskontinuiteter kan inducera bakterietillväxt eller andra obehagliga egenskaper, därför måste vissa tillverkare upprätta nolltolerans mot skadade komponenter. Detta leder till behovet av effektiva provningsmetoder av oförstörande natur som kan utvärdera stora mängder komponenter i en produktionslinjemiljö. Författarna har fått uppdrag av Acoustic Agree AB och Trelleborg AB, en världsledande producent av polymertekniska lösningar, för att finna en oförstörande provningsmetod som kan detektera diskontinuiteter i tunna polymerkomponenter i en produktionslinjemiljö. Implementering i produktionslinjemiljöer ställer krav på testcykeltid och ett mål är att kunna slutföra en testcykel inom 3-4 sekunder. Begränsningen av tillgänglig utrustning och expertis har lett till att författarna har fokuserat på att tillämpa olinjära akustiska metoder för oförstörande provning istället för mer konventionella metoder. Dessa metoder utnyttjar olinjär distortion av akustiska vågor vilket medför att vissa egenskaper uppträder, såsom generering av övertoner (HH), frekvensmodulering (NWMS), resonansfrekvensskift (NRUS) och amplitudåtergång (IDAR). Visuell provning användes för att skildra synligt skadade prover från till synes oskadade. De enda metoder som visade möjligheten att urskilja skadade tunna polymerkomponenter var NRUS och IDAR. Endast den sistnämnda har förmågan att utföra provning inom den föreskrivna provcykeltiden. Olinjära akustiska metoder verkar kunna detektera diskontinuiteter i tunna polymerkomponenter inom den givna tidsramen. Mer arbete krävs för att korrekt utvärdera prestandan av NWMS och IDAR. Konfigurationen som användes i detta arbete var huvudsakligen inriktad på IDAR, vilket resulterade i att specifik kalibrering för NWMS ej utförts. Provpopulationen var även för låg för att kunna samla in tillräckligt med data för att uppnå statistisk säkerhet angående metodernas prestanda.

This dissertation thesis deals with elastohydrodynamic lubrication (EHL) of surfaces with real roughness. Systematic study of surface harmonic components behaviour, during their passage of the EHL contact, by other authors led to retrieval of a uniform principle according which individual harmonic components deform inside the contact. Although the amplitude attenuation model can provide rapid tool for more accurate surface assessment in design process it has not been validated on real surfaces till now. The thesis represents pioneering work on quantitative analysis of real surfaces EHD measurements. It contains original and new results with real and model roughness features that in many cases confirm theoretical expectations. Deeper understanding of surface roughness behaviour inside EHL contact, which is typical for machine parts like bearings, gears, cams etc., can help to improve design process of the components and ultimately enable machines and equipment to work with lower energy requirements and higher durability.

No description available.

No description available.

No description available.