Diese Arbeit beschäftigt sich mit dem Auftreten verschiedener magnetischer Phasen (Ferro-, Antiferro-, Paramagnetismus) im Kondo-Gitter-Modell (engl. Kondo-lattice model, KLM). Mit dem KLM wird die Wechselwirkung itineranter Elektronen mit lokalisierten magnetischen Momenten beschrieben. Zur Berechnung wichtiger Kenngrößen wird der Green-Funktions-Formalismus verwendet, der schließlich zu einer Näherungslösung führt. Es wird gezeigt, dass eine hinreichend gute Näherung, insbesondere die explizite Betrachtung von Doppelbesetzungen, notwendig ist, um die Resultate nicht zu verfälschen. Es ist damit möglich, Phasendiagramme in Abhängigkeit wichtiger Modellparameter bei verschwindender Temperatur zu berechnen. Um Ergebnisse bei endlichen Temperaturen zu erhalten, wird eine Möglichkeit zur Berechnung der freien Energie eingeführt, die das KLM wesentlich direkter beschreibt als die Standardabbildungsmethode der Rudermann-Kittel-Kasuya-Yoshida-Wechselwirkung. Neben der Berechnung des reinen KLMs erlauben diese Methoden auch eine Beschreibung von Systemen mit gebrochener Translationssymmetrie. Konkret werden verdünnte magnetische Halbleiter und dünne magnetische Filmsysteme betrachtet. Es entstehen hier zusätzliche Effekte wie der Ladungstransfer von magnetischen zu unmagnetischen Gitterplätzen oder umgekehrt. Insgesamt zeigt sich außerdem eine qualitative Übereinstimmung mit verschiedenen experimentellen Daten. / This thesis is concerned with the existence of magnetic phases (ferro-, antiferro-, paramagnetism) in the Kondo-lattice model (KLM). The KLM describes an interaction between itinerant electrons and localized magnetic moments. For the calculation of important quantities the Green''s function formalism is used which leads to an approximative solution of the problem. It is shown that a sufficient approximation, in particular the explicit treatment of double occupations, is necessary to avoid a falsification of the results. Thereby it is possible to calculate magnetic phase diagrams in dependence of important model parameters at vanishing temperature. To get results at finite temperature a possibility for the calculation of the free energy is introduced, which describes the KLM more directly than the standard procedure of the Rudermann-Kittel-Kasuya-Yoshida interaction. Besides the calculation of the pure KLM these methods permit a description of systems with a reduced translational symmetry. More precisely diluted magnetic semiconductors and thin magnetic films systems are treated. In these system new effects arise such as the charge transfer from magnetic to non-magnetic sites and vice versa. All in all there is also a qualitative agreement with several experimental data.

In der vorliegenden Arbeit wird die Physik von Schwer-Fermion-Systemen, die durch Lanthanid- und Aktinid-Übergangsmetallverbindungen realisiert werden, untersucht. Die Basis für eine theoretische Beschreibung bildet das Anderson-Gitter, welches das Wechselspiel freier Leitungselektronen und stark korrelierter Elektronen aus lokalisierten f-Orbitalen charakterisiert. Als Zugang zu diesem Modell wird die von Wegner vorgeschlagene Flussgleichungsmethode verwendet, ein analytisches Verfahren, welches auf der Konstruktion eines effektiven Hamilton-Operators basiert. Ein zentrales Thema dieser Arbeit ist die Beschreibung der elektronischen Struktur von Schwer-Fermion-Systemen. Insbesondere wird die Abhängigkeit statischer Größen vom Einfluss verschiedener Systemparameter betrachtet. Die Dynamik kollektiver Anregungen in Schwer-Fermion-Systemen wird an Hand der elektronischen Zustandsdichten und dynamischen magnetischen Suszeptibilitäten untersucht. / The physical properties of heavy-fermion systems are examined. These systems are mainly formed by rare earth or actinide compounds. Their essential physics can be characterized by the periodic Anderson model which describes the interplay of itinerant metal electrons and localized, but strongly correlated f-electrons. The present calculations are based on the flow equations approach proposed by Wegner. This method uses a continuous unitary transformation to derive an effective Hamiltonian of an easy to treat structure. Within this framework the electronic structure of heavy-fermion systems is calculated and the influence of external parameters is studied. Beside the derivation of static properties the density of states and dynamic magnetic susceptibilities are investigated in order to characterize the nature of collective excitations.

O entendimento dos princípios básicos do enovelamento proteico pode conduzir a muitas aplicações importantes. Embora não se conheçam todos os aspectos significativos envolvidos neste problema, experimentos e aproximações teóricas têm produzido avanços relevantes na sua compreensão, como, por exemplo, o papel dominante das forças hidrofóbicas. Com o propósito de contribuir para a identificação de alguns parâmetros determinantes no processo de folding, o objetivo deste trabalho consiste em estudar o potencial estéreo-hidrofóbico (isto é, o potencial hidrofóbico e um conjunto de especificidades estéricas adequadas) relacionado ao processo de enovelamento de proteínas globulares, evidenciando o papel das restrições estéricas e as características topológicas do estado nativo. Para tanto, empregou-se um modelo simplificado em rede cúbica, estudando-se, através de simulações Monte Carlo, o comportamento de mais de 40 cadeias tipo-proteínas, que apresentam estados nativos caracterizados pelos seguintes parâmetros topológicos: ordem de contato, χ, ordem de longo alcance e número φ de estruturas tipo grande manivela. Claramente, os resultados das simulações para as cadeias com configurações nativas, caracterizadas por baixo valor de χ e elevado valor de φ, são bastante diferenciados daqueles obtidos para as cadeias que apresentam configurações nativas com χ elevado e φ pequeno, evidenciando que o potencial estéreo-hidrofóbico, adotado neste trabalho, permite relacionar os comportamentos termodinâmico e cinético da cadeia tipo-proteína com os atributos topológicos da configuração nativa correspondente. Adicionalmente, mostrou-se que as restrições estéricas consideradas introduzem fortes mudanças na atividade configuracional, aumentando a estabilidade do estado nativo, bem como alterando, drasticamente, a curva da capacidade térmica, em função da temperatura, em comparação com os resultados das simulações obtidos com o potencial hidrofóbico (sem restrições estéricas). / The understanding of the basic principles of the protein folding process may lead to very important applications. Although all significant aspects of this problem are not yet known, experimental and theoretical results have given important contribution on the subject, as, for instance, about the dominant role of the hydrophobic forces. In order to contribute to the identification of significant ingredients for the folding process, the main goal of this work consists in studying the stereo-hydrophobic potential (that is, the hydrophobic potential and a set of steric specificities) related to the folding process of globular proteins, revealing the importance of the steric constraints and the role of the native structure. A minimalist lattice model was employed for this purpose, and more than forty distinct protein-like chains were studied. These chains were designed based on native structures characterized by topological parameters as contact order, χ, long range order and number φ of crankshaft-like structures. The Monte Carlo simulation results show clearly that the folding process depends strongly on the topological attributes of the native structure: thermodynamical and kinetic behavior for chains designed from native structures, presenting smaller χ and higher φ, are very distinguishable from those with higher χ and lower φ. Additionally it was shown that the steric constraints significantly modify the configurational activity, increasing the general conditions for the globule stability, as well changing drastically the shape of the thermal capacity behavior, as a function of the temperature, in comparison with the corresponding results obtained using the hydrophobic potential only, i.e., without the steric specificities.

In this thesis we show how reactive processes can be affected by the presence of different types of spatial constraints, so much so that their nonlinear dynamics can be qualitatively altered or that new and unexpected behaviors can be produced. To understand how this interplay can occur in general terms, we theoretically investigate four very different examples of this situation. <p><p>The first system we study is a reversible trimolecular chemical reaction which is taking place in closed one-dimensional lattices. We show that the low dimensionality may or may not prevent the reaction from reaching its equilibrium state, depending on the microscopic properties of the molecular reactive mechanism. <p><p>The second reactive process we consider is a network of biological interactions between pigment cells on the skin of zebrafish. We show that the combination of short-range and long-range contact-mediated feedbacks can promote a Turing instability which gives rise to stationary patterns in space with intrinsic wavelength, without the need of any kind of motion.<p><p>Then we investigate the behavior of a typical chemical oscillator (the Brusselator) when it is constrained in a finite space. We show that molecular crowding can in such cases promote new nonlinear dynamical behaviors, affect the usual ones or even destroy them. <p><p>Finally we look at the situation where the constraint is given by the presence of a solid porous matrix that can react with a perfect gas in an exothermic way. We show on one hand that the interplay between reaction, heat flux and mass transport can give rise to the propagation of adsorption waves, and on the other hand that the coupling between the chemical reaction and the changes in the structural properties of the matrix can produce sustained chemomechanical oscillations. <p><p>These results show that spatial constraints can affect the kinetics of reactions, and are able to produce otherwise absent nonlinear dynamical behaviors. As a consequence of this, the usual understanding of the nonlinear dynamics of reactive systems can be put into question or even disproved. In order to have a better understanding of these systems we must acknowledge that mechanical and structural feedbacks can be important components of many reactive systems, and that they can be the very source of complex and fascinating phenomena.<p> / Doctorat en Sciences / info:eu-repo/semantics/nonPublished

Ce travail réalisé dans le cadre d'une collaboration industrielle avec la société ALSTOM Transport porte sur l'étude de la dégradation des voies ferrées urbaines. Les composantes de voie retenus pour cette étude sont le rail et la dalle de voie en béton. Concernant le rail, différents problèmes sont abordés : contact roue – rail, usure du rail, usure ondulatoire du rail, et fatigue de contact de roulement (RCF) du rail. Un outil numérique avec des interfaces graphiques, nommé CONUS, est développé pour le problème de contact roue – rail et le problème d'usure du rail. Des théories classiques (Hertz, Kalker, Archard, etc.) sont implantées dans cet outil. La méthode stationnaire est implantée dans un code de calcul par éléments finis pour étudier l'état asymptotique de l'acier du rail sous le chargement répété des trains. Ceci nous permet de prédire les régimes de RCF du rail. La mécanique de l'endommagement est utilisée pour prédire la fatigue du matériau béton. Le formalisme de Marigo couplé avec le modèle d'endommagement de Mazars permet de modéliser la dégradation progressive de la rigidité du matériau sous chargement cyclique. Une campagne d'essais de fatigue du béton en flexion a été réalisée. Elle a pour but de valider le modèle théorique et d'identifier les paramètres du matériau. Le dimensionnement d'une dalle de voie en béton a fait l'objet d'une application de cette méthode. Le modèle de réseau de poutres (lattice model) a été utilisé pour étudier la propagation des fissures dans les structures en béton. Ce modèle a été implanté dans le logiciel de calcul par éléments finis, CESAR-LCPC. Les résultats numériques (propagation de fissures) obtenus pour les structures simples sous chargement statique sont en tout point comparables avec les résultats d'essais expérimentaux. Ce modèle a ensuite été utilisé pour étudier la fissuration sous chargement de fatigue. Pour cela un modèle d'endommagement simple modélisant la dégradation des éléments «poutres» s'est avéré suffisant pour décrire la cinématique de propagation des fissures / This work is part of the collaboration between the laboratory Navier (UMR ENPC /IFSTTAR/ CNRS) and ALSTOM Transport company (TGS/Trackway). It focuses on the study of the degradation of urban railways. The components of track considered in this study are the rail and the concrete slab. Regarding the rail, different problems are discussed : wheel – rail contact, rail wear, rail corrugation and rolling contact fatigue (RCF). A numerical tool with graphical interfaces, called CONUS, is developed to predict the behaviour of the wheel - rail contact, the rail wear, and the rail corrugation problems. Classical theories (Hertz, Kalker, Archard, etc...) are implemented in this tool. The stationary method is implemented in a finite element software to study the asymptotic state of the rail steel under repeated loading of trains.The damage mechanics is used to predict the fatigue life of concrete. Marigo's formalism coupled with Mazars' damage model is used to predict the gradual degradation of material stiffness under cyclic loading. A campaign of fatigue tests for concrete in bending was conducted. It aims at validating the theoretical model and identifying material parameters. We applied this method in order to design the concrete slabs of urban railway. The lattice model was used to study the crack propagation in concrete structures. This model was implemented in the finite element software, CESAR-LCPC. The numerical results obtained for simple structures under static loading are consistent with the results of laboratory experiments. This model was then used to study the crack propagation under fatigue loading. For that purpose, a simple damage model of degradation of the "beams" elements describes the kinematics of crack propagation with a satisfying accuracy

No description available.

In der vorliegenden Arbeit wird die Physik von Schwer-Fermion-Systemen, die durch Lanthanid- und Aktinid-Übergangsmetallverbindungen realisiert werden, untersucht. Die Basis für eine theoretische Beschreibung bildet das Anderson-Gitter, welches das Wechselspiel freier Leitungselektronen und stark korrelierter Elektronen aus lokalisierten f-Orbitalen charakterisiert. Als Zugang zu diesem Modell wird die von Wegner vorgeschlagene Flussgleichungsmethode verwendet, ein analytisches Verfahren, welches auf der Konstruktion eines effektiven Hamilton-Operators basiert. Ein zentrales Thema dieser Arbeit ist die Beschreibung der elektronischen Struktur von Schwer-Fermion-Systemen. Insbesondere wird die Abhängigkeit statischer Größen vom Einfluss verschiedener Systemparameter betrachtet. Die Dynamik kollektiver Anregungen in Schwer-Fermion-Systemen wird an Hand der elektronischen Zustandsdichten und dynamischen magnetischen Suszeptibilitäten untersucht. / The physical properties of heavy-fermion systems are examined. These systems are mainly formed by rare earth or actinide compounds. Their essential physics can be characterized by the periodic Anderson model which describes the interplay of itinerant metal electrons and localized, but strongly correlated f-electrons. The present calculations are based on the flow equations approach proposed by Wegner. This method uses a continuous unitary transformation to derive an effective Hamiltonian of an easy to treat structure. Within this framework the electronic structure of heavy-fermion systems is calculated and the influence of external parameters is studied. Beside the derivation of static properties the density of states and dynamic magnetic susceptibilities are investigated in order to characterize the nature of collective excitations.