Les systèmes distribués sont de plus en plus grands et complexes, alors que leur utilisation s'étend à de nombreux domaines (par exemple, les communications, la domotique, la surveillance, le ``cloud''). Par conséquent, les contextes d'exécution des systèmes distribués sont très divers. Dans cette thèse, nous nous focalisons sur des contextes incertains, autrement dit, le contexte n'est pas complètement connu au départ ou il est changeant. Plus précisément, nous nous focalisons sur deux principaux types d'incertitudes : une identification incomplète des processus et la présence de fautes. L'absence d'identification est fréquente dans de grands réseaux composés d'appareils produits et déployés en masse. De plus, l'anonymat est souvent une demande pour la sécurité et la confidentialité. De la même façon, les grands réseaux sont exposés aux pannes comme la panne définitive d'un processus ou une perte de connexion sans fil. Néanmoins, le service fourni doit rester disponible.Cette thèse est composée de quatre contributions principales. Premièrement, nous étudions le problème de l'élection de leader dans les anneaux unidirectionnels de processus homonymes (les processus sont identifiés mais leur ID n'est pas forcément unique). Par la suite, nous proposons un algorithme d'élection de leader silencieux et autostabilisant pour tout réseau connecté. Il s'agit du premier algorithme fonctionnant sous de telles conditions qui stabilise en un nombre polynomial de pas de calcul. La troisième contribution est une nouvelle propriété de stabilisation conçue pour les réseaux dynamiques qui garantit des convergences rapides et progressives après des changements topologiques. Nous illustrons cette propriété avec un algorithme de synchronisation d'horloges. Finalement, nous considérons la question de la concurrence dans les problèmes d'allocation de ressources. En particulier, nous étudions le niveau de concurrence qui peut être atteint dans une grande classe de problèmes d'allocation de ressources, l'allocation de ressources locales. / Distributed systems become increasingly wide and complex, while their usage extends to various domains (e.g., communication, home automation, monitoring, cloud computing). Thus, distributed systems are executed in diverse contexts. In this thesis, we focus on uncertain contexts, i.e., the context is not completely known a priori or is unsettled. More precisely, we consider two main kinds of uncertainty: processes that are not completely identified and the presence of faults. The absence of identification is frequent in large networks composed of massively produced and deployed devices. In addition, anonymity is often required for security and privacy. Similarly, large networks are exposed to faults (e.g, process crashes, wireless connection drop), but the service must remain available.This thesis is composed of four main contributions. First, we study the leader election problem in unidirectional rings of homonym processes, i.e., processes are identified but their ID is not necessarily unique. Then, we propose a silent self-stabilizing leader election algorithm for arbitrary connected network. This is the first algorithm under such conditions that stabilizes in a polynomial number of steps. The third contribution is a new stabilizing property designed for dynamic networks that ensures fast and gradual convergences after topological changes. We illustrate this property with a clock synchronizing algorithm. Finally, we consider the issue of concurrency in resource allocation problems. In particular, we study the level of concurrency that can be achieved in a wide class of resource allocation problem, i.e., the local resource allocation.

Abstract During the past decade submicron and nanoparticles have aroused a wide interest and gained new applications due to their high surface area and strength. Grinding with a wet stirred media mill is usually the last process step before the submicron or nanoparticles are added to an application, and the step where the final particle size distribution is achieved. Since stirred media milling is an energy-intensive process, energy efficiency should be optimized. This can be done by determining the optimum operational parameters for the mill and using the highest possible solids concentration. The solids concentration can be increased by controlling particle-particle interactions with stabilization chemicals, e.g. polymers. This thesis concerns parameters and grinding aids affecting the particle size distribution and suspension stability of the aqueous submicron grinding of calcium carbonate (CaCO3) and titanium dioxide (TiO2) in stirred media mills. TiO2 particles are aggregates produced via a bottom-up method, while CaCO3 are primary mineral particles produced by a top-down method. The most energy efficient grinding of TiO2 to a 300 nm particle size with the narrowest possible particle size distribution was obtained with the lowest stress energy, implying the smallest grinding medium size. It was observed that electrosteric stabilization with sodium polyacrylates was effective for TiO2, and sodium polyacrylate with a molecular weight of 12500 g/mol was found to be the most effective for reducing the viscosity of the suspension. As with TiO2, electrosteric stabilization with sodium polyacrylates was also found to be effective for CaCO3, but in this case sodium polyacrylate with a lower polydispersity index was more effective, showing a better stabilization potential in micron and submicron grinding and reducing the viscosity and particle size to a greater extent. Nanogrinding experiments were performed for a CaCO3 suspension with low PDI sodium polyacrylate and it was found to be possible to obtain a particle size of 26 nm, smaller than any size previously reported when grinding CaCO3. / Tiivistelmä Viimeisen kymmenen vuoden aikana alle yhden mikrometrin partikkelit ovat herättäneet kiinnostusta ja niille on kehitetty uusia sovelluksia niiden suuren pinta-alan ja lujuuden ansiosta. Ultrahienojauhatus märkähelmimyllyllä on useimmiten viimeinen prosessivaihe ennen partikkelien lisäämistä sovelluskohteeseen ja siinä saavutetaan partikkelien lopullinen partikkelikokojakauma. Helmimyllyjauhatuksen energiankulutus minimoidaan etsimällä optimioperointiparametrit kullekin jauhatusprosessille ja käyttämällä korkeinta mahdollista suspension kuiva-ainepitoisuutta. Suspension kuiva-ainepitoisuutta voidaan nostaa hallitsemalla partikkelien välisiä vuorovaikutuksia stabilointiaineilla, kuten polymeereillä. Tässä väitöskirjassa tutkittiin operointiparametrien ja jauhatusapuaineiden vaikutusta titaanidioksidin (TiO2) ja kalsiumkarbonaatin (CaCO3) partikkelikokojakaumaan ja lietteen stabiilisuuteen submikronijauhatuksessa. Tutkitut TiO2-partikkelit olivat aggregaatteja, jotka oli valmistettu sulfaattiprosessilla saostamalla, ja tutkitut CaCO3-partikkelit olivat primäärisiä mineraalipartikkeleita. TiO2-partikkeleille saavutettiin energiatehokkain jauhatus ja samalla toivottu partikkelikokojakauma, eli mediaani 300 nm ja mahdollisimman kapea jakauma, pienillä helmillä, jotka aiheuttavat partikkeleihin pienimmän puristusenergian. Elektrosteerinen stabilointi käyttämällä natriumpolyakrylaatteja stabilointiaineena havaittiin tehokkaaksi menetelmäksi hallita TiO2-partikkelien välisiä vuorovaikutuksia. Natriumpolyakrylaatti, jonka molekyylimassa oli 12500 g/mol, oli tehokkain TiO2-partikkeleille alentaen suspension viskositeettiä eniten. Myös CaCO3-partikkeleille elektrosteerinen stabilointi natriumpolyakrylaatteja käyttäen oli tehokkain stabilointimenetelmä. Myös natriumpolyakrylaattien polydispersiteetti-indeksin vaikutusta tutkittiin CaCO3-suspensioille. Tulokset osoittivat matalan polydispersiteetti-indeksin olevan tehokkaampi alentaen viskositteettia ja pienentäen partikkelikokoa tehokkaammin kuin natriumpolyakrylaatti, jolla oli korkeampi polydispersitetti-indeksi. Tämän vuoksi natriumpolyakrylaatti, jolla oli matala polydispersiteetti-indeksi, valittiin nanojauhatuskokeisiin. Kokeissa CaCO3-partikkelit saatiin jauhettua 26 nm kokoon, joka on pienin koskaan aiemmin jauhamalla saavutettu koko CaCO3-partikkeleille.

Abstract Biomass, such as wood, binds CO2 as it grows, and is thus considered an environmentally friendly alternative fuel to replace coal. In Finland, biomass is typically co-combusted with peat, and also municipal waste is becoming more common as a fuel for power plants. Wood, peat and waste-based fuels are typically burned in fluidized bed combustion (FBC) boilers. Ash is the inorganic, incombustible residue resulting from combustion. The annual production of biomass and peat ash in Finland is 600 000 tonnes, and this amount is likely to increase in the future, since the use of coal for energy production will be discontinued during the 2020s. Unfortunately, FBC ash is still largely unutilized at the moment and is mainly dumped in landfills. The general aim of this thesis was to generate information which could potentially improve the utilization of FBC ash by alkali activation. The specific objective was to produce geopolymer aggregates by means of a simultaneous alkali activation-granulation process. It was shown that geopolymer aggregates with physical properties comparable to commercial lightweight expanded clay aggregates (LECAs) can be produced from FBC fly ash containing heavy metals. Although the ashes were largely unreactive and no new crystalline phases were formed by alkali activation, a new amorphous phase was observed in the XRD patterns, possibly representing micron-sized calcium aluminate silicate hydrate-type gels. The heavy metal immobilization efficiency of alkali activation varied with the type of fly ash. Good stabilization was generally obtained for cationic metals such as Ba, Pb and Zn, but in common with the results obtained with alkali activation of coal fly ash, anionic metals became leachable after alkali activation. The efficiency of immobilization depended on the physical and chemical properties of the fly ash and was not related to the total content of the element. All the geopolymer aggregates met the criteria for a lightweight aggregate (LWA) as defined by EN standard 13055-1. Their strength depended on the reactivity and particle size distribution of the fly ash. Mortars and concretes prepared with such geopolymer aggregates had higher mechanical strength, higher dynamic modulus of elasticity and higher density than concrete produced with commercial LECA, while exhibiting similar rheology and workability. / Tiivistelmä Biopolttoaineet, esimerkiksi puu, ovat ympäristöystävällinen vaihtoehto kivihiilelle, koska ne sitovat hiilidioksidia kasvaessaan. Suomessa biopolttoaineita poltetaan tyypillisesti turpeen kanssa, ja nykyään myös jätteen hyödyntäminen polttoaineena on yleistynyt. Puu, turve ja jätepolttoaineet poltetaan tyypillisesti leijupetipoltto-tekniikalla. Tuhka on polton epäorgaaninen, palamaton jäännös. Puun ja turpeen tuhkaa tuotetaan Suomessa 600 000 tonnia vuodessa ja määrän odotetaan kasvavan, sillä kivihiilen poltto lopetetaan 2020-luvulla. Leijupetipolton tuhkaa ei tällä hetkellä juurikaan hyödynnetä ja tuhka päätyykin pääasiassa kaatopaikoille. Tämän tutkielman päämääränä oli tuottaa tietoa, joka parantaisi leijupetipolton tuhkien hyödyntämistä alkali-aktivaatiolla. Erityisesti tavoitteena oli valmistaa geopolymeeriaggregaatteja yhtäaikaisella alkali-aktivaatiolla ja rakeistuksella. Tutkielmassa osoitettiin, että raskasmetalleja sisältävistä tuhkista valmistettujen geopolymeeriaggregaattien fysikaaliset ominaisuudet ovat vertailukelpoiset kaupallisten kevytsora-aggregaattien (LECA) kanssa. Vaikka tuhkien reaktiivisuus oli matala, ja uusia kidefaaseja ei muodostunut alkaliaktivaatiolla, uusi amorfinen faasi havaittiin XRD-mittauksissa. Uusi amorfinen faasi oli mahdollisesti mikrometrikokoluokan kalsium-aluminaatti-silikaatti-hydraatti-tyyppinen rakenne. Raskasmetallien stabiloinnin tehokkuus vaihteli tuhkien välillä. Kationiset metallit, kuten barium, lyijy ja sinkki, stabiloituivat pääasiassa hyvin, mutta anionisten metallin liukoisuus kasvoi alkali-aktivoinnin myötä. Stabiloinnin tehokkuus riippui tuhkien fysikaalisista ja kemiallisista ominaisuuksista, mutta raskasmetallin kokonaispitoisuudella ei ollu vaikutusta. Kaikki geopolymeeriaggregaatit olivat kevytsora-aggregaatteja standardin EN 13055-1 mukaisesti. Aggregaattien lujuus riippui tuhkan reaktiivisuudesta ja partikkelikokojakaumasta. Geopolymeeriaggregaateilla valmistettujen laastien ja betonien mekaaninen lujuus, Youngin moduuli ja tiheys olivat korkeampia kuin kaupallisella kevytsora-aggregaateilla valmistetut, vaikka niiden reologia ja työstettävyys olivat samanlaisia.

Bibliographic identification KUČEROVÁ, Klára. Long-term monitoring of impact of competitive swimming in statics and dynamics of the spine. Prague: Charles University, 2nd Faculty of Medicine, Department of Rehabilitation and Sports Medicine, 2017, 59 p. Supervisor of the bachelor's thesis Mgr. Magdaléna Lepšíková. Abstract Swimming is a physical activity that is often recommended by physicians as compensatory activity. It is symmetrical sport which takes place in aquatic environment. This sport has many benefits and can be done at any age. Our goal was to find out if competitive swimming had an effect on statics and dynamics of the spine. To measure static sagittal parameters we used DIER Formetric III 4D. Dynamic tests were measured by spinal development tests. The trunk stabilization was detected using Dynamic Neuromuscular Stabilization (DNS) concepts. We observed 11 children aged 11 - 12. These children were examined twice in the bachelor thesis, first one before the start of racing swimming and second one after 10 months. In our diploma thesis, these children were examined again after 3 years of racing swimming. We also watched a set of 11 adult swimmers. We compared their results with standards. After three years of racing swimming in children, there was no statistically significant change in size of...

Wind loads acting on wooden building structures need to be dealt with adequately in order to ensure that neither the serviceability limit state nor the ultimate limit state is exceeded. For the structural designer of tall buildings, avoiding the possibly serious consequences of heavy wind loading while taking account at the same time of the effects of gravitation can be a real challenge. Wind loads are usually no major problem for low buildings, such as one- to two-storey timber structures involving ordinary walls made by nailing or screwing sheets of various types to the frame, but when taller structures are designed and built, serious problems may arise. Since wind speed and thus wind pressure increases with height above the ground and the shear forces transmitted by the walls increase accordingly, storey by storey, considerable efforts can be needed to handle the strong horizontal shear forces that are exerted on the bottom floor in particular. The strong uplift forces that can develop on the wind side of a structure are yet another matter that can be critical. Accordingly, a structure needs to be anchored to the substrate or to the ground by connections that are properly designed. Since the calculated uplift forces depend very much upon the models employed, the choice of models and simplifications in the analysis that are undertaken also need to be considered carefully. The present licentiate thesis addresses questions of how wind loads acting on multi-storey timber buildings can be best dealt with and calculated for in the structural design of such buildings. The conventional use of sheathing either nailed or screwed to a timber framework is considered, together with other methods of stabilizing timber structures. Alternative ways of using solid timber elements for stabilization are also of special interest. The finite element method was employed in simulating the structural behaviour of stabilizing units. A study was carried out of walls in which sheathing was nailed onto a timber frame. Different structural levels were involved, extending from modelling the performance of a single fastener and of the connection of the sheathing to frame, to the use of models of this sort for studying the overall structural behaviour of wall elements that possess a stabilizing function. The results of models used for simulating different load cases for walls agreed reasonably well with experimental test results. The structural properties of the fasteners binding the sheathing to the frame, as well as of the connections between the members of the frame were shown to have a strong effect on the simulated behaviour of shear wall units. Regarding solid wall panels, it was concluded that walls with a high level of both stiffness and strength can be produced by use of such panels, and also that the connections between the solid wall panels can be designed in such a way that the shear forces involved are effectively transmitted from one panel to the next.

Magister Scientiae - MSc / This project "An application of multiple regression in exchange rate arrangement" focused on the processes followed by different countries when choosing an exchange rate regime for currency stabilization. It analyses the consequences faced by emerging markets as a result of changes in volatility of developed countries’ currencies (American Dollar, Japanese Yen, EURO, British Pound and the Canadian Dollar). / South Africa

Made available in DSpace on 2017-06-01T17:57:27Z (GMT). No. of bitstreams: 1 wallace de sa.pdf: 2263305 bytes, checksum: a0c0a930c5ab441157930460e793e9b6 (MD5) Previous issue date: 2006-03-28 / One of the main problems, in the Region Metropolitan of Recife (RMR), in the hillsides it is the disordered occupation, increasing the number of housings in risk areas. One of the solutions for the stabilization of the hillsides is the use of the construction of retained structure as support wall. The friction angle soil-wall is a basic parameter for the design of a support wall. Its value is used in the evaluation of the pushes active and passive in the analysis of stability of the wall. Practical the current one of projects in Brazil considers the value of the friction angle soil-wall as being equal to the angle of friction of the soil, a parcel of it or same zero to depend on the case. The values found in literature also indicate as estimative for this angle values that vary of zero to the angle of friction of the soil. Experimental values for Brazilian soils are not available in literature. Of this form the projects of support walls can be being designed against the security or of uneconomical form. In this context in the present dissertation tests of direct shear in specimens of composites of soil and another representative material of support walls had been carried through (conventional concrete, concrete with recycled aggregate of RCD and rock), with the objective to get the friction angle soil-wall. Two soils of the hillsides of Recife had been used (sandy and a other clayey one). The roughness of the surfaces in contact with soil had been evaluated, to analyze its influence in the interaction soil-wall. The gotten results are compared with values suggested in literature, having considered the influence of the soil and the roughness of the contact surface. The relation enters the friction angle soil-wall and the angle of friction of the ground (ä/ö) varies of 3/4 the 1, with the average roughness for the soil sandy of the Ibura and of 1/3 the 3/4 for the soil clayey of Nova Descoberta. In the soil sandy, the friction soil-wall has minor influence of the roughness of the faying surface and has value very next to the angle of internal friction of the soil. In the soil clayey the friction soil-wall strong is influenced by the roughness of the faying surface, and its value varies of 1/3 the 3/4 of the angle of friction of the soil / Um dos principais problemas, na Região Metropolitana do Recife (RMR), nas encostas, é a ocupação antrópica desordenada, aumentando o número de moradias em áreas de risco. Uma das soluções para a estabilização das encostas é o uso da construção de estrutura de contenção como muro de arrimo. O ângulo de atrito solo-muro é um parâmetro fundamental para o dimensionamento. O seu valor é utilizado na avaliação dos empuxos ativo e passivo na análise de estabilidade do muro. A prática atual de projetos considera o valor do ângulo de atrito solo-muro como sendo igual ao ângulo de atrito do solo, uma parcela dele ou mesmo nulo a depender do caso. Valores experimentais para solos brasileiros não estão disponíveis na literatura. Desta forma, os projetos de muros de arrimo podem estar sendo dimensionados contra a segurança ou de forma antieconômica. Neste contexto na presente dissertação, foram realizados ensaios de cisalhamento direto em corpos de prova de solo e de outro material representativo de muros de arrimo (concreto convencional, concreto com agregado de resíduos de construção e demolição - RCD e rocha), com o objetivo de obter os ângulos de atrito interno do solo e do contato solomuro. Foram utilizados dois solos das encostas do Recife um arenoso (Ibura) e outro argiloso (Nova Descoberta). Foram avaliadas as rugosidades das superfícies em contato com os solos, para analisar sua influência na interação solo-muro. Os resultados obtidos são comparados com valores sugeridos na literatura, considerando a influência do solo e da rugosidade da superfície de contato. A relação entre o ângulo de atrito solo-muro e o ângulo de atrito do solo (ä/ö) varia de 3/4 a 1, com a Rugosidade Média, para o solo arenoso do Ibura e de 1/3 a 3/4 para o solo argiloso de Nova Descoberta. No solo arenoso, o atrito solo-muro tem menor influência da rugosidade da superfície de contato e tem valor muito próximo do ângulo de atrito interno do solo. No solo argiloso o atrito solo-muro é fortemente influenciado pela rugosidade da superfície de contato, e o seu valor varia de 1/3 a 3/4 do ângulo de atrito do solo

Mudanças na política agrícola brasileira têm preconizado a adoção de mecanismos de mercado para o fortalecimento da comercialização, financiamento à produção e mitigação dos riscos agropecuários, tanto o climático, quanto o de mercado. O seguro rural é um dos instrumentos mais promissores nesse novo estágio da política agrícola, pois permite a administração do risco agrícola, ao mesmo tempo em que lastreia as operações de comercialização e financiamento agrícola. O seguro de receita emerge como um instrumento ainda mais robusto de estabilização a receita agrícola, na medida em que garante a variação de produtividade e preço, simultaneamente. O instrumento já é consolidado nos Estados Unidos e começa a ser estudado no Brasil. Por essa razão, a primeira parte do estudo, capítulo 2, visa analisar os planos de seguro existentes e indicar os modelos que devem ser fomentados no Brasil. Constatou-se que os modelos estadunidenses baseados no plano de Proteção de Renda (IP - Income Protection) e Receita Garantida (RA - Revenue Assurance), substituídos pelo plano Proteção de Receita (RP - Revenue Protection), são os mais adequados para iniciar o desenvolvimento dessa modalidade de seguro no Brasil. Na segunda parte do trabalho, capítulo 3, é apresentado modelo teórico de plano de seguro de receita, bem como procedimento metodológico de cálculo da taxa de prêmio, de modo univariado e bivariado. Aplicada a metodologia ao caso da soja no Paraná, concluiu-se que as taxas calculadas no estudo são inferiores às praticadas nos dois projetos experimentais existentes. O distanciamento entre as taxas praticadas no mercado e a diferença em relação às estimadas na nesta pesquisa indicam imprecisão no cálculo das taxas de prêmio e são evidências de superestimação das taxas pelas seguradoras. / Changes in Brazilian agricultural policies have advocated the adoption of market mechanisms for strengthening the marketing, the financing to production, and both climate and market farming risk mitigation. Rural insurance is one of the most promising instruments in this new stage of agricultural policy, for crop risk administration at the same time it serves as collateral to marketing operations and agricultural funding. The insurance revenue emerges as an even more robust stabilization of agricultural revenue instrument to the extent that it ensures the variation of productivity and price simultaneously. The instrument is already consolidated in the United States and begins to be studied in Brazil. For this reason, the first part of the study, Chapter 2, aims to analyze existing insurance plans and indicate the models that should be encouraged in Brazil. It was found that models based on U.S. Income Protection (IP) and Revenue Assurance (RA), replaced by the plan Revenue Protection are best suited to start the development of this type of insurance in Brazil. In the second part of the dissertation, Chapter 3, the theoretical model of revenue insurance plan is presented, as well as a methodology for univariate and bivariate premium ratemaking. The methodology was applied to the case of soybean in Paraná, and it was concluded that the rates calculated in this study are lower than those of the two existing experimental projects. The gap between the market rates and the difference in relation to the rates estimated in the study indicate inaccuracy in the calculation of premium rates and are evidence of rate overestimation by insurers.

Au cours de dix dernières années la science aux attoseconde ou Physique au champ-fort a été l’objet d’un fort développement. La production d’impulsions laser énergétiques de courte durée à haute cadence et stabilisées en CEP constitue la première étape pour accéder aux dynamiques ultra-rapides caractérisant l’interaction de la matière avec une source de lumière cohérente, intense et ultra-rapide. Le travail de cette thèse consiste à améliorer globalement les performances d’un système laser Ti:Sa à haute cadence optimisé pour la génération des impulsions attoseconde. Nous avons développé une nouvelle configuration de cavité régénérative fonctionnant à 10 kHz qui permet une meilleure gestion des effets thermiques dans le cristal. En sortie de l’amplificateur les impulsions atteignent des valeurs de puissance de 5 W en bande étroite (35 fs), ou 2.7 W en bande spectrale large permettant une compression des impulsions proche de 17 fs. La CEP des impulsions en sortie d’amplificateur a été stabilisée ; le bruit résiduel mesuré tir-à-tir est de 210 mrad pendant trois heures.L’amplificateur peut supporter également le fonctionnement en mode accordable, en sélectionnant des spectres de 30 à 40 nm de largeur à mi-hauteur et en accordant leur longueur d’onde centrale dans une gamme de 80 nm autour de 800 nm. Nous avons conçu et mis en fonctionnement un amplificateur multi-passages non-cryogéné à imagerie par lentille thermique pour accroître la puissance des impulsions jusqu’à 10 W à une cadence de 1 kHz. Le régime de forte saturation d’amplificateur garantit une variation négligeable (±3% pic à pic) de la puissance des impulsions en sortie du module, face à une variation importante de la puissance en entrée (±25% pic à pic) sur la bande spectrale accordable. L’amplification peut encore être plus importante grâce à une ligne d’amplification à refroidissement cryogénique, qui permet d’atteindre des puissances au niveau TW, à la cadence de 1 kHz, tout en maintenant un régime de courte durée (17.5 fs) et stabilité en CEP (350 mrad de bruit résiduel tir-à-tir). Nous proposons aussi une étude des sources de bruit de CEP dans les modules hautement dispersifs: nous avons conçu une nouvelle approche numérique sur la base d’un logiciel de tracé de rayon commercial (Zemax) pour évaluer les variations de CEP dans les modules contenant réseaux de diffraction. / The last decade has seen a lot of progress in attosecond science or in strong field physics. Generating energetic, few-cycle laser pulses with stabilized Carrier-Envelope Phase at high repetition rate constitutes the first step to access the ultra-fast dynamics underlying the interaction of matter with intense, ultrashort coherent light source. The work of this thesis consists in globally improving the performances of a high repetition rate Ti:Sa laser system optimized for attosecond science. We present an original 10 kHz Ti:Sa CPA laser based on an newlydesigneddouble-crystal cavity for thermal lensing management. The amplifier delivers up to 5 W in narrow band mode (35 fs pulses), or 2.7 W in broad band mode, supporting 17 fs pulses after temporal compression. We demonstrate shot-to-shot CEP stabilization with a remaining noise of 210 mrad over three hours at the front-end output. In parallel to the short pulse duration operation mode, it is possible to use the front end in a wavelength tunability mode within a 80 nm range around 800 nm, with a resolution of 1 nm and 30 to 40 nm of bandwidth. We designed and demonstrated a complete water-cooled lens-less multipass amplifier using thermal lensing for modeadaptation boosting the pulse energy up to 10mJ at 1 kHz repetition rate (up to 10 W). The saturation regime of the amplifier ensures negligible variation (±3% peak to peak) of the output power for significant variation of the input power (±25% peak to peak) over the tunability range. The energy scalability of the front-end is demonstrated by coupling its output to cryogenically cooled amplifier, delivering 1 kHz, TW-class pulses at 17.5 fs and CEP stabilized with a residual noise of 350 mrad. A study of CEP noise sources in high dispersive module is also addressed, proposing a numerical approach based on a commercial ray-tracing software (Zemax) for predicting CEP fluctuation in grating based modules.

Dans cette thèse, nous étudions les théories étroitement liées du contrôle, de la stabilisation et de la propagation des singularités, pour des équations aux dérivées partielles dispersives linéaires et non-linéaires. Les résultats principaux proviennent des travaux de l’auteur:[1] Zhu, H., 2016. Stabilization of damped waves on spheres and Zoll surfaces of revolution. ESAIM : Control, Optimisation and Calculus of Variations (ESAIM: COCV), à paraître.[2] Zhu, H., 2017. Control of three dimensional water waves. arXiv preprint arXiv:1712.06130.[3] Zhu, H., 2018. Propagation of singularities for gravity-capillary water waves. arXiv preprint arXiv:1810.09339.Dans [1], nous avons étudié la stabilisation des ondes amorties sur les surfaces de révolution de Zoll. Nous avons donné un exemple où la région d’amortissement est à la limite de la condition du contrôle géométrique, alors que les ondes amorties présentent une décroissance exponentielle uniforme de l’énergie. Cet exemple généralise un résultat de Lebeau. Dans [2], nous avons étudié la contrôlabilité du système des ondes de surface avec tension superficielle. Nous avons démontré, en dimensions arbitraires, la contrôlabilité exacte pour des petites données spatialement périodiques à condition du contrôle géométriques. Ce résultat généralise le travail de Alazard, Baldi et Han-Kwan en dimension deux. Dans [3], nous avons étudié la propagation des singularités pour des ondes de surface avec tension superficielle. Nous avons défini le front d’onde quasi-homogène, généralisant le front d’onde de Hörmander et le front d’onde homogène de Nakamura et démontré des résultats de propagation des fronts d’onde quasi-homogènes par le système des ondes de surface avec tension superficielle. Comme corollaires, nous avons obtenu des effets régularisants locaux et micro-locaux pour les données initiales présentant une décroissance spatiale suffisante. / In this thesis, we study the closely related theories of control, stabilization and propagation of singularities for some linear and nonlinear dispersive partial differential equations. Main results come from the author’s works:[1] Zhu, H., 2016. Stabilization of damped waves on spheres and Zoll surfaces of revolution. ESAIM: Control, Optimisation and Calculus of Variations (ESAIM: COCV), to appear.[2] Zhu, H., 2017. Control of three dimensional water waves. arXiv preprint arXiv:1712.06130.[3] Zhu, H., 2018. Propagation of singularities for gravity-capillary water waves. arXiv preprint arXiv:1810.09339.In [1] we studied the stabilization of the damped wave equation on Zoll surfaces of revolution. We gave an example where the region of damping is at the borderline of the geometric control condition, yet the damped waves exhibit a uniform exponential decay of energy, generalizing an example of Lebeau.In [2] we studied the controllability of the gravity-capillary water wave equation. Under the geometric control condition, we proved in arbitrary spatial dimension the exact controllability for spatially periodic small data. This generalizes a result of Alazard, Baldi and Han-Kwan for the 2D gravity-capillary water wave equation.In [3] we studied the propagation of singularities for the gravity-capillary water wave equation. We defined the quasi-homogeneous wavefront set, generalizing the wavefront set of H¨ ormander and the homogeneous wavefront set of Nakamura, and proved propagation results for quasi-homogeneous wavefront sets by the gravity-capillary water wave equation. As corollaries, we obtained local and microlocal smoothing effects for gravity-capillary water waves with sufficient spatial decay.