Une grande quantité d’énergie est rejetée par l’industrie à bas niveau de température, en dessous de 200 °C. Afin d’améliorer le rendement énergétique global des procédés utilisés, il est envisageable de valoriser cette chaleur perdue appelée chaleur fatale. Cependant cette valorisation est souvent rendue difficile par la présence d’un décalage temporel entre le moment où l’énergie est rejetée et le moment auquel cette énergie pourrait être de nouveau utilisée. Associant de fortes capacités de stockage ainsi qu’une possible restitution d’énergie à température constante, la solution du stockage de l’énergie thermique par des Matériaux à Changement de Phase, appelés MCP, apparaît comme particulièrement attractive. Cependant, la mise en œuvre de ces systèmes de stockage se heurte à des verrous scientifiques et technologiques tant au niveau du matériau de stockage que du système mais également de son contrôle commande et de son insertion dans les procédés industriels.L’objectif de la thèse est de mettre au point un système de stockage par MCP solide-liquide dans deux gammes de température : 70-85 °C et 120-155 °C. La première correspond aux températures des réseaux de chaleurs ou des chauffages domestiques alors que la deuxième s’applique au préchauffage des procédés industriels déjà existants. La thèse vise à démontrer la faisabilité technique du système de stockage. Le travail s’articule autour de différentes tâches allant de la sélection et la caractérisation des MCP jusqu’à leur mise en œuvre dans un organe de stockage et la simulation numérique de la solution de stockage.Les MCP recensés dans la bibliographie à ces niveaux de températures ont été caractérisés finement par calorimétrie (DSC) afin de déterminer leurs propriétés thermo-physiques sur des échantillons de grade laboratoire. L’acide stéarique pour la gamme 70-85 °C et l’acide sébacique pour la gamme 120-155 °C ont été sélectionnés. Des analyses calorimétriques plus poussées sur le grade industriel de ces matériaux ont été réalisées avec notamment des analyses de vieillissement et de compatibilité avec leur encapsulation respective au sein d’un banc expérimental. Le prototype expérimental de stockage thermique a été dimensionné et conçu pour répondre aux sollicitations simulant les rejets et les demandes d’un procédé industriel. Ce banc d’essais est composé principalement de deux organes de stockage que sont une cuve cylindrique et un échangeur multitubulaire et d’un thermorégulateur servant à simuler le fonctionnement du procédé industriel. Dans l’échangeur multitubulaire, le MCP occupe toute le volume de la calandre tandis que le fluide caloporteur circule dans les tubes. La cuve, quant à elle, contient des capsules sphériques en polyoléfines dans lesquelles le MCP est confiné. Elle est traversée par le fluide caloporteur procédant aux échanges thermiques. Ces capsules sphériques appelées nodules ne peuvent supporter plus de 100 °C et sont exclusivement réservées pour la gamme basse température. Ainsi, l’acide stéarique a été confiné dans les nodules afin de remplir la cuve de stockage. L’acide sébacique a lui été intégré dans la calandre de l’échangeur multitubulaire. Les campagnes expérimentales réalisées ont montré la faisabilité de ces types de stockage. Enfin, un modèle numérique simulant les performances du module de stockage utilisant les MCP encapsulés a été réalisé. Il constitue la première étape d’un outil de simulation complet intégrant les briques technologiques du stockage latent. / A large amount of energy is rejected by the industry at low temperature level, below a temperature of 200 °C. In order to improve the overall energy efficiency of industrial processes, it is possible to re-use this waste heat. However, this energy recovery is often made difficult because of the time difference between the process step at which the energy is lost and the process step at which this energy could be reused. Combining high energy storage capabilities and a possible energy recovery at constant temperature, thermal storage solution by phase change materials (PCM) is particularly attractive. However, this storage systems implementation faces scientific and technologic obstacles concerning both the storage material and system but also its command system and its integration into industrial processes.This thesis aims to develop a thermal energy storage system using a solid-liquid PCM technology in two temperature ranges: 70-85 °C and 120-155 °C. The first one corresponds to temperatures of heating networks or domestic heating systems, while the second one could directly preheat existing industrial processes. The thesis aims to demonstrate the technical feasibility of the storage system. The purpose is divided into different tasks such as PCMs selection and characterization, PCM implementation in a storage system but also numerical simulation of the storage solution.PCM documented in the literature at those temperature ranges were characterized by Differential Scanning Calorimetry (DSC) in order to determine thermo physical properties on laboratory grade samples. Stearic acid for the 70-85 °C temperature range and sebacic acid for the 120-155 °C temperature range were selected. Deeper differential scanning calorimetry analyses were carried out on those industrial grade materials including material ageing process analyses and their compliance with their respective encapsulation within an experimental test bench.Thermal storage experimental prototype was designed in order to meet the demands simulating the rejects and needs of industrial processes. The test bench is mainly composed of two storage systems : a cylindrical tank, a multitubular exchanger and a thermoregulator used to simulate industrial process functioning. The PCM, while in the multitubular exchanger, fills up the whole volume of the shell whereas the heat transfer fluid flows in tubes. The tank, for its part, contains polyolefin spherical capsules in which the PCM is contained. The tank is crossed by the heat transfer fluid conducting heat exchanges. Those spherical capsules called nodules cannot be exposed to temperatures exceeding 100 °C and are exclusively reserved for the low temperatures range. Thus, stearic acid was confined in nodules so as to fill the storage tank. The sebacic acid was incorporated in the multitubular exchanger shell. Experimental campaigns carried out have demonstrated the feasibility of those storage types.

Afin de parvenir à une meilleure compréhension des processus d’évolution des magmas réunionnais, nous avons réalisé une étude ayant pour objectifs principaux : (1) la simulation expérimentale du système d’alimentation du Piton de la Fournaise, à partir d’un matériel de départ de composition typique de Steady State Basalt et dans des conditions de pression, température, fO2 et de teneurs en volatils (H2O, CO2) réalistes ; (2) la détermination de la séquence de cristallisation d’échantillons représentatifs des roches plutoniques réunionnaises de façon à les comparer avec les résultats expérimentaux et (3) l’obtention d’une base de données volcanologiques, pétrologiques et géochimiques significative sur le groupe de laves anormales (« Abnormal Group ») permettant de confirmer son existence dans le système d’alimentation du Piton de la Fournaise. La découverte de verres silicatés ayant des compositions caractéristiques de l’Abnormal Group confirme l’implication de ces magmas dans l’activité éruptive. Toutefois, les roches plutoniques n’enregistrent pour la plupart que des séquences de cristallisation témoignant d’une évolution superficielle des magmas. Cette dernière est simulée de façon satisfaisante par les expérimentations réalisées dans la gamme 0.1 à 50 MPa qui conduisent à des modèles pétrologiques et des pressions de stockage en accord avec les données géophysiques. Les expérimentations à plus forte pression démontrent l’existence de paliers au sein du système d’alimentation qui peuvent en grande partie expliquer les diverses compositions réunionnaises, mais posent la question de la composition des magmas parentaux. / To better understand magmatic processes associated with the evolution of La Réunion magmas, we have carried out a multi-approach study aimed at (1) simulating experimentally the feeding system of the Piton de la Fournaise volcano, using a Steady State Basalt starting material and P-T-fO2-Volatiles (H2O, CO2) conditions compatible with the natural system; (2) determining crystallization sequences representative of La Réunion plutonic rocks for comparison with the experimental results and (3) constructing a volcanological, petrological and geochemical database for lavas of the Abnormal Group, to confirm the existence of Abnormal melts in the feeding system of the volcano. The discovery of glasses having chemical characteristics similar to the Abnormal Group establishes the implication of Abnormal melts in eruptive processes. However, plutonic rocks record crystallization sequences that for the most part indicate a low pressure magmatic evolution. Experiments in the pressure range 0.1 to 50 MPa satisfactorily reproduce conditions in the shallow magmatic systems and lead to petrological models and magma storage depths in agreement with geophysical data. Experiments at higher pressures demonstrate transitions in magma fractionation mechanisms in the feeding system that can explain the range of erupted compositions, but call into question the compositions of parental magmas.

Les développements récents des techniques IRM engendre des sources de potentiels induits qui « polluent » les signaux électrophysiologiques, utilisé simultanément en IRM pour surveiller le patient et synchroniser les images. Le système élaboré dans ce travail est compose de deux modules « émetteur-récepteur » IRM-compatibles. Le premier permet d’introduire dans le tunnel d’IRM des signaux EPS dont on connait les caractéristiques. Les signaux sont injectés dans un tissu conducteur placé dans le tunnel. Le second module permet de recueillir les signaux après leur contamination par les artéfacts générés par les séquences d’imagerie. Il comporte 20 canaux répartis en quatre bandes fréquentielles (40, 80, 160 et 350 Hz). Des mesures du potentiel induit ont été réalisées en environnements 1,5 T et 3 T. Nous pouvons ainsi analyser les modifications des paramètres des signaux selon les séquences, mais aussi à l’intérieur des différentes bandes de fréquences. Dans ce travail, nous présentons une caractérisation et modélisation des potentiels induits par les commutations de gradients de champ magnétique recueilli par notre dispositif expérimental. La caractérisation et la modélisation permettent d’obtenir des informations pertinentes à prendre en compte pour l’élaboration des algorithmes de filtrage efficaces et robustes. / New developments in MRI techniques create sources of induced voltages that “pollute” the simultaneously acquired electrophysiological signals (EPS), used to monitor patients and/or synchronize images. We developed a device to allow a deep study of the contamination mechanism, which would assist in elaborating new tools to obtain higher quality EPS. The system consists of three main modules: (i) a signal transmission system composed of an EPS generator and a transmission box, which transmits the EPS to a MR-compatible receiver inside the tunnel, (ii) an electro-conductive tissue-mimicking phantom in which the EPS is injected, (iii) a signal collection module composed of a MR compatible amplifier-transmitter that emits, via an optical cable, the collected signal to a receiver box placed outside the MRI room. The receiver box comprises 20 channels distributed into four frequency bands (40, 80, 160, and 350 Hz). Measurements of the induced voltages were performed in 1.5 T and 3 T MRI environments. An algorithm to extract and analyze and model the induced voltages was developed. The modeling algorithm is based on a sinusoidal decomposition of the induced voltages. This work aimed to assess the disturbance level of the EPS, when using larger bandwidth amplifiers. The characterization and modeling of the induced voltages, which represent the dominant noise, reveal relevant information which can be used to develop robust and efficient noise reduction algorithms.

[pt] Entender o processo de produção de fuligem é crucial para o projeto de novos queimadores, como os de fornos industriais. Estes queimadores, que utilizam processos de combustão turbulenta, dependem de transferência de calor via radiação das partículas de fuligem para as paredes do forno para seu bom funcionamento. A fuligem formada na região de radiação deve ser oxidada para evitar problemas de saúde e meio ambiente. Mesmo tendo havido significativo progresso no decorrer das duas últimas décadas em relação às chamas laminares, a interação entre a turbulência e a produção de fuligem ainda é um problema em aberto. Este trabalho apresenta resultados experimentais recentes da distribuição instantânea e média da distribuição de fuligem em chamas turbulentas de etileno/ar não prémisturadas estabilizadas em um queimador do tipo bluff-body. A intensidade de turbulência na região de esteira deste queimador é muito alta, levando a uma presença de fuligem intermitente e a estruturas de fuligem altamente distorcidas. A distribuição de fuligem é medida usando incandescência induzida por laser (LII), com uma excitação em 266 nm a 10 Hz e fluência de 0,8 J/cm2 e medição em 400 nm por uma câmera intensificada. Os resultados da técnica LII são comparados à técnica clássica de extinção da luz. Resultados da distribuição de hidrocarbonetos aromáticos policíclicos também são apresentados. Os resultados obtidos permitem caracterizar a distribuição da função de densidade de probabilidade de fuligem. Nas situações de escoamento onde a turbulência da esteira é controlada pelo escoamento de ar, demonstra-se que a PDF da fração volumétrica de fuligem corresponde a uma distribuição lognormal. / [en] Understanding the soot production process is crucial to the design of new burners, such as those in industrial furnaces. Indeed, these burners, which use turbulent combustion processes, rely on radiative heat transfer from the soot particles to the furnace walls to operate properly. The soot formed within the radiation region must the be oxidized in order to avoid health and environment issues. Although there has been significant progress over the past two decades in relation to laminar flames, the interaction between turbulence and soot production is still an open problem. This works presents recent experimental results of the instantaneous and mean soot distribution on non-premixed turbulent ethylene/air flames stabilized at a bluff-body burner. The turbulence intensity in the wake region of this burner is very high, leading to a soot intermittent presence and to highly distorted soot structures. The soot distribution is measured using laserinduced incandescence (LII), with 266 nm excitation at 10 Hz, 0.8 J/cm2 fluence and collected at 400 nm by an intensified camera. The results of the LII technique are compared to those of a classical of light extinction technique. Polyciclic aromatic hydrocarbon distribution results are also presented. The results obtained allow to characterize the soot probability density function distribution. In flow situations where the wake turbulence is controlled by the air flow, the soot volume fraction PDF is shown to correspond to a lognormal distribution.

La recherche permanente de solutions d’allègement des structures aéronautiques a conduit au développement de nouveaux alliages comme les Al-Cu-Li et les Al-Mg-Li qui se caractérisent par une densité inférieure aux alliages conventionnels et par des propriétés améliorées. De manière générale, la formabilité de ces alliages, que ce soit à froid ou à chaud est encore assez mal connue. Le projet « Formage avancé des alliages basses densité pour application aéronautique » se propose donc de caractériser précisément la plage de formabilité de ces nouveaux alliages et d’investiguer les voies de mise en œuvre les plus adaptées : procédé d’étirage conventionnel ou procédé de mise en forme pendant une étape de traitement thermique. Ce projet regroupe les entreprises AEROLIA SAS, ACB, AIRBUS Operation SAS et CONSTELLIUM.Ces développements seront destinés au fuselage métallique et plus particulièrement à la structure pointe-avant qui présente une géométrie en double courbure ; ils concernent à la fois les panneaux de fuselage et les profilés. / The ongoing researches to lighten aeronautical structures have led to the development of new materials like Al-Cu-Li and Al-Mg-Li alloys. These alloys exhibit lower density and higher mechanical properties than conventional alloys. However, their ability to be formed either by cold or hot forming process is not well known. The aims of this PhD research works are thus to characterize the formability of these new alloys and to suggest the most appropriate forming sequence to elaborate complex aeronautic parts. Various forming processes will be studied such as cold forming by stretching of sheet metal parts or hot forming by stamping of sheet metal parts after or during different heat treatments. An experimental as well as a numerical approach will be used to develop the forming process of these new alloys to elaborate fuselage panels with more or less complex shapes. These PhD research works are a component of a larger project led by IRT Jules Verne and entitled “Advanced forming of low density alloys for aeronautic applications”. This project includes several firms from aeronautic domain (AEROLIA SAS, ACB, AIRBUS SAS and CONSTELLIUM).

Esta dissertação resulta de uma pesquisa que investigou a interpretação atribuída a orações relativas com um núcleo [+animado], como a fada que molhou, que são ambíguas entre uma leitura de tema (a fada foi molhada) e uma leitura de agente (a fada molhou alguém). A leitura de tema é associada à estrutura absoluta, na qual o verbo transitivo sofre uma mudança de valência: não há argumento agentivo e, consequentemente, não há projeção vP. O verbo não pode checar o Caso acusativo do objeto, forçando o argumento temático a se mover para a posição pré-verbal, onde concorda com o verbo (NEGRÃO; VIOTTI, 2010). Essa estrutura é considerada mais simples do que passivas, por exemplo, uma vez que, nela, o agente não está nem mesmo implícito na sentença. Já a leitura de agente corresponde a uma estrutura em que o objeto não está expresso foneticamente. Em experimentos que eliciaram a produção de relativas de objeto no Português Brasileiro infantil (VIVANCO; PIRES, 2012; GROLLA; AUGUSTO, 2016; RANGEL, 2016), as crianças produziram relativas absolutas para evitar a complexidade estrutural da relativa de objeto. Os adultos raramente as produziram, preferindo as relativas passivas. Dada a alta taxa de produção de relativas absolutas por parte das crianças nesses testes, nosso objetivo foi investigar a preferência das crianças na interpretação de relativas ambíguas. Nós hipotetizamos que a leitura de agente seria preferida, considerando que o parser tende a atribuir lacuna na primeira posição disponível, i.e, a posição de sujeito (PHILLIPS, KAZANINA, ABADA, 2005). Uma vez que a relativa absoluta envolve uma lacuna extra (o tema se move da posição de objeto para a posição pré-verbal: a fada que _ molhou _ ), esta não deverá ser a análise preferida pelo parser. Além disso, o fato de as relativas ambíguas testadas terem um NP [+animado] como núcleo contribui para que este seja interpretado como agente (BECKER, 2014). Foi conduzido um experimento envolvendo a Tarefa de Julgamento de Valor de Verdade, no qual os participantes julgaram relativas de objeto, relativas passivas e relativas ambíguas como verdadeiras ou falsas com base em histórias contadas a eles. Nas relativas ambíguas, ambas as interpretações estavam disponíveis nas histórias, de modo que o julgamento falso se associava à leitura de tema, e o verdadeiro, à leitura de agente. Participaram do teste 46 crianças adquirindo PB (16 3;6-4;6); (30 5;0-5;11) e 26 adultos falantes nativos de PB. Ao contrário do que prevemos, as crianças preferiram a leitura de tema. Já os adultos prefeririam a leitura de agente, embora não tenham rejeitado completamente a leitura de tema. A alta frequência da leitura de tema nos dados das crianças nos leva a descartar a hipótese de que a absoluta seja apenas uma saída à qual as crianças recorrem quando têm de produzir estruturas mais complexas, como as relativas de objeto. O fato de que adultos também acessam essa interpretação evidencia que as absolutas sejam de fato estruturas lícitas no PB. / This dissertation results of a research that has investigated the interpretation of relative clauses with a [+animated] head, as \"the fairy that washed\". These relatives are ambiguous between a theme interpretation (the fairy got washed) and an agent interpretation (the fairy washed someone). The theme interpretation is associated with the absolutive sentence, in which a transitive verb suffers a valence alternation: there is no agent, and consequently, no vP is projected. The verb cannot check the object\'s accusative Case, forcing the theme to move to a pre-verbal position, agreeing with the verb (NEGRÃO; VIOTTI, 2010). The resulting structure is simpler than passives, for example, since the agent is not even implicity expressed in the sentence. The agent interpretation is associated with a structure with the object not expressed phonetically. In experiments eliciting object relative clauses in child Brazilian Portuguese (BrP), children produced absolutive relatives to avoid the structural complexity of the object relative clause (VIVANCO; PIRES, 2012; GROLLA; AUGUSTO, 2016; RANGEL, 2016). Adults rarely produced them, preferring passive relatives instead. Given children\'s high production of absolutive relatives in these experiments, our aim was to investigate children\'s preferences in interpreting ambiguous relative clauses. We hypothesize that the agent interpretation should be preferred, as it enables the parser to assign a gap in the first available position, i.e, the subject position (PHILLIPS, KAZANINA, ABADA, 2005). Since the absolutive relative involve an extra step (the NP theme moves from object to preverbal position: the fairy that _ washed _ ), it should not be the parser\'s preferred analysis. Also, the fact that the ambiguous relatives tested involved a [+animated] NP as the relative head should contribute to this NP will be interpreted as agent (BECKER, 2014). We conducted a truth-value judgment task, in which participants judged object relatives, passive relatives and ambiguous as true or false based on stories told to them. On ambiguous trials, both interpretations were entertained in the stories, with true judgments paired with the theme interpretation and false judgments associated with the agent interpretation. 46 BrP-acquiring children were tested (16 3;6-4;6); (30 5;0-5;11) and 26 adults, native BrP speakers. Contrary to our expectations, children\'s group preferred the theme interpretation. Adults preferred the agent interpretation, although they did not reject theme completely. The high frequency of the theme interpretation in children\'s data leads us to discard the hypothesis that absolutives are just an escape hatch that children resort to when forced to produce more complex structures, as object relatives. The fact that adults also accessed this interpretation brings additional evidence that the absolutive sentences are indeed licit structures in BrP.

Die klinischen Langzeitergebnisse des vorderen Kreuzbandersatz mit dem Patellarsehnentransplantat nach durchschnittlich 13,8 Jahren waren nicht zufriedenstellend. Hauptursachen hierfür waren operativ-technische Mängel, wie Fehlpositionierung der Bohrkanäle und die Fixierung der Transplantate in einer zu hohen Kniebeugestellung. Diese Abweichungen führten zu signifikant schlechteren Bewegungsausmaßen sowohl in der Streckung als auch in der Beugung. Insbesondere bei zu weit ventral positionierter femoraler Bohrkanalanlage resultierte ein erhebliches Beugedefizit mit Stressbelastung auf das femorotibiale Kompartiment, ähnlich einem Nußknackerphänomen. Bei zu weit ventral positioniertem tibialen Bohrkanal wurde eine signifikante Verringerung der Extension beobachtet. Die Patienten mit gleichzeitiger Meniskusresektion hatten signifikant höhere arthrotische Veränderungen. Sowohl die zum damaligen Zeitpunkt allgemein anerkannte Operationstechnik als auch die postoperative Rehabilitation ist zeitlich überholt und entspricht nicht den gegenwärtigen internationalen Standards. In der experimentellen Untersuchung zum Einwachsverhalten einer autologen vorderen Kreuzbandplastik am Kaninchenmodell konnte der Einfluss und die Wirkung von manipulativ wirkenden Faktoren auf die Ausbildung eines stabilen Bindegewebsregenerates belegt werden. Sowohl der fibrinstabilisierende Faktor XIII als auch die eingesetzte Transplantat-Splitting-Technik zeigten eine positive Wirkung auf den Ein- und Umbau sowie den biomechanischen Ausrissfestigkeiten der Transplantate. Die alleinige Wirkung des Faktor XIII spielt vor allem in der Frühphase eine stimulierende Rolle auf die Initiierung der Umbauvorgänge und der ossären Intergration der Knochenblöcke. Mittel- und langfristig wird die Ausbildung von belastungsstabilen Kollagenfibrillen durch andere weitestgehend unbekannte Einflussfaktoren bestimmt. Die neuartig entwickelte Transplantat-Splitting-Technik erwies sich für den Transplantatumbau ebenfalls als vorteilhaft. Mit dieser Transplantatbearbeitung wurde die Grundlage für eine frühzeitige synoviale Ernährung der Bündel gewährleistet. Durch die artifizielle Gewebsläsion werden reparative Vorgänge beschleunigt. Auf der Grundlage der experimentell gewonnenen Daten kann die Transplantat-Splitting- Technik für die klinische Erprobung und Praxiseinführung empfohlen werden. / The clinical long term results mean 13,8 years after anterior cruciate ligament reconstructions with bone-patellar tendon-bone transplant were not satisfied. Main reasons were technical errors, malpositioning of the bone tunnels and fixation of the transplant in a too high degree of flexion. As a result the patients had significant reductions of motion in extension as well as in flexion. Especially in cases with an anterior femoral bone tunnel position the patients had a loss of flexion and high degree of shear forces in the femoro-tibial compartment similar a nutcracker phenomenon. In ACL reconstructions with an anterior tibial bone tunnel position could be observed a loss of extension. Patients with concomitant meniscal resection showed significant higher rate of osteoarthrotic changes. Both, the former operative technique and the postoperative rehabilitation programme had no clinical relevance today and are replaced by new standards. In an experimental study regarding to the remodeling process of an autologous anterior cruciate ligament plasty in a rabbit model could be examined factors which influencing the maturation process of the graft. The fibrin stabilizing factor XIII as well as the transplant splitting technique showed a positive effect in the remodelling process and the pull-out strength of the transplants. In the early phase the factor XIII activated the initial remodeling and the osseous ingrowth of the bone blocks. In middle and long term period of remodeling there plays unknown factors an important role for building of force restraint collagen fibrils. The new developed splitting technique had some additional advantages. Using these technique the basis of early synovial nutrition could be realized. The artificial tissue lesion leads to an acceleration of the reparative response. On the basis of the experimental data the transplant splitting technique can be recommended and introduce in the clinical trial.

A compact air separator demonstrator based on centrifugally enhanced distillation has been studied. The full size device is meant to be used on board of a Two Stage To Orbit vehicle launcher. The air separation system must be able to extract oxygen in highly concentrated liquid form (LEA, Liquid Enriched Air) from atmospheric air. The LEA is stored before being used in a subsequent rocket propulsion phase by the second stage of the launcher. Two reference vehicles are defined, one with a subsonic first stage and one with a supersonic first stage. In both cases, oxygen collection is performed during a cruise phase (M 0.7 and M 2.5 respectively). The aim of the project is to demonstrate the feasibility of the air separation system, investigate the separation cycle design, and assess that the separator design selected is suitable for the reference vehicles. The project is described from original base ideas to design, construction, extended testing and analysis of experimental results. Preliminary computations for a realistic layout have been performed and the motivations for the choices made during the process are explained. Test rig design, separator design and technical discussion are provided for a subscale pilot unit. Mass transport parameters and flooding limits have been estimated and experimentally measured. Performance has been assessed and shown to be sufficient for the reference Two Stage To Orbit vehicles. The technology developed is found suitable without further optimization, although some volume and mass reduction would be desirable for the supersonic first stage concept. There are many ways of optimisation that can be further investigated. The aim of this program, however, is not to fully optimize the device, but to demonstrate that a device based on a simple, robust, low-risk design is already suitable for the launch vehicles. On top of that analysis, directions for improvements are suggested and their potentials estimated. A complete assessment of those improvements requires further maturation of the technological concept through further testing and practical implementations. Directions for future work, general conclusions and a vehicle development roadmap have also been provided.

In this thesis, we explore two different approaches to causal inferences. The traditional approach models the theoretical relationship between the outcome variables and their explanatory variables, i.e., the science, at the same time as the systematic differences between treated and control subjects are modeled, i.e., the assignment mechanism. The alternative approach, based on Rubin's Causal Model (RCM), makes it possible to model the science and the assignment mechanism separately in a two-step procedure. In the first step, no outcome variables are used when the assignment mechanism is modeled, the treated students are matched with similar control students using this mechanism, and the models for the science are determined. Outcome variables are only used in the second step when these pre-specified models for the science are fitted. In the first paper, we use the traditional approach to evaluate whether a husband is more prone to quit smoking when his wife quits smoking than he would have been had his wife not quit. We find evidence that this is the case, but that our analysis must rely on restrictive assumptions. In the subsequent two papers, we use the alternative RCM approach to evaluate if a Harvard freshman who does not smoke (observed potential outcome) is more prone to start smoking when he shares a suite with at least one smoker, than he would have been had he shared a suite with only smokers (missing potential outcomes). We do not find evidence that this is the case, and the small and insignificant treatment effect is robust against various assumptions that we make regarding covariate adjustments and missing potential outcomes. In contrast, we do find such evidence when we use the traditional approach previously used in the literature to evaluate peer effects relating to smoking, but the treatment effect is not robust against the assumptions that we make regarding covariate adjustments. These contrasting results in the two latter papers allow us to conclude that there are a number of advantages with the alternative RCM approach over the traditional approaches previously used to evaluate peer effects relating to smoking. Because the RCM does not use the outcome variables when the assignment mechanism is modeled, it can be re-fit repeatedly without biasing the models for the science. The assignment mechanism can then often be modeled to fit the data better and, because the models for the science can consequently better control for the assignment mechanism, they can be fit with less restrictive assumptions. Moreover, because the RCM models two distinct processes separately, the implications of the assumptions that are made on these processes become more transparent. Finally, the RCM can derive the two potential outcomes needed for drawing causal inferences explicitly, which enhances the transparency of the assumptions made with regard to the missing potential outcomes. / Diss. Stockholm : Handelshögskolan, 2006 S. 1-13: sammanfattning, s. [15]-161: 4 uppsatser

Esta tesis caracteriza experimentalmente el comportamiento del canal MIMO en diferentes escenarios y evalúa la dependencia del rendimiento del canal con la tipología del escenario y diferentes condiciones de propagación. Para facilitar el desarrollo de futuros sistemas con múltiples antenas en estos escenarios, se presentan modelos sencillos y eficientes de canales SISO y MIMO.Se ha realizado el análisis de canales MIMO 2x2 a partir de diversas campañas de medidas realizadas en distintos escenarios interiores, exteriores y mixtos. Se han estudiado dos aspectos fundamentales: la correlación espacial y la capacidad del canal. El estudio se completa con el desarrollo de un método de modelado basado en líneas de retardo. Mediante este método se obtienen modelos sencillos y eficientes de canales SISO y MIMO de los escenarios interiores y mixtos medidos. Así se contribuye al desarrollo de nuevos sistemas de comunicaciones inalámbricos en escenarios mixtos y a facilitar la simulación de sistemas MIMO gracias a los nuevos modelos propuestos. / This thesis characterizes in an experimental way the MIMO channel behavior in different scenarios, including the evaluation of the channel performance dependency with the type of scenario and different propagation conditions. In order to facilitate the development of future systems with multiple antennas in these scenarios, simple and efficient SISO and MIMO channel models are presented.The analysis of the 22 MIMO channel has been based on several measurement campaigns carried out in different outdoor, indoor and mixed environments. Two fundamental aspects are studied: the spatial correlation and the channel capacity. The analysis is completed with the development of a method to model the radio channel based on tapped delay lines. Using this method, simple and efficient SISO and MIMO channel models of indoor, outdoor and mixed scenarios have been obtained. In this way this work contributes to the development of new wireless communication systems and to facilitate the simulation of MIMO systems by means of the new proposed models.