Both Part I and Part II are included. / The separation of a mixture of glass spheres in water using 2 inch hydrocyclones was studied. <p> Three operating parameters were investigated: feed concentration, volume split and feed flow rate. In addition, three design parameters were cone angle, inlet diameter, and vortex finder length. The performance criterion parameters were the efficiency with which the solids were separated from the liquid, and the energy required per unit mass flowing through the hydrocyclone. </p> <p> First the experimental data were analyzed by three different statistical methods and the results compared in an attempt to determine which statistical method was most suitable for this two criteria system. The three methods were principal component analysis, canonical correlation analysis and multiple regression analysis. The theory behind these methods is briefly outlined. Our conclusion is that using all three methods give much more insight than could be obtained from any individual method. </p> <p> Second, an analysis of the above eight hydrocyclonc parameters of hydrocyclones with cylindrical sections indicated that for the range of parameters covered in this work, feed flow rate and inlet diameter influenced the energy loss most; volume split influenced the separation efficiency the most. Energy loss and separation efficiency are quite independent; this means that it is possible to design and run a hydrocyclone with high separation efficiency and low energy loss. The dilute concentrations used in this work indicate that a hydrocyclone of conventional design can be used in waste water treatment. When the parameters were correlated, a power model gave more consistent interpretation than a linear model. </p> <p> Third, the effect of the three operating parameters on hydrocyclones with three different body shapes suggested that the most efficient cyclone was one with a straight cone and no cylindrical section. The body shape dictated which parameters would significantly affect performance. </p> / Thesis / Master of Engineering (ME)

No description available.

During recent years, new types of materials have been discovered with unique properties. One family of such materials are two-dimensional materials, which include graphene and MXene. These materials are stronger, more flexible, and have higher conductivity than other materials. As such they are highly interesting for new applications, e.g. specialized in vivo drug delivery systems, hydrogen storage, or as replacements of common materials in e.g. batteries, bulletproof clothing, and sensors. The list of potential applications is long for these new materials. As these materials are almost entirely made up of surfaces, their properties are strongly influenced by interaction between their surfaces, as well as with molecules or adatoms attached to the surfaces (surface groups). This interaction can change the materials and their properties, and it is therefore imperative to understand the underlying mechanisms. Surface groups on two-dimensional materials can be studied by Transmission Electron Microscopy (TEM), where high energy electrons are transmitted through a sample and the resulting image is recorded. However, the high energy needed to get enough resolution to observe single atoms damages the sample and limits the type of materials which can be analyzed. Lowering the electron energy decreases the damage, but the image resolution at such conditions is severely limited by inherent imperfections (aberrations) in the TEM. During the last years, new TEM models have been developed which employ a low acceleration voltage together with aberration correction, enabling imaging at the atomic scale without damaging the samples. These aberration-corrected TEMs are important tools in understanding the structure and chemistry of two-dimensional materials. In this thesis the two-dimensional materials graphene and Ti3C2Tx MXene have been investigated by low-voltage, aberration-corrected (scanning) TEM. High temperature annealing of graphene covered by residues from the synthesis is studied, as well as the structure and surface groups on single and double Ti3C2Tx MXene. These results are important contributions to the understanding of this class of materials and how their properties can be controlled.

The phenomenon of energy loss that occurs when an ion interacts with matter, also called stopping power, has been investigated for more than a century, and has provided findings of interest. However, reliable procedures for obtaining accurate experimental measurements and a fully theoretical comprehension of the process are tasks still in high demand by the scientific community. Moreover, stopping power data are prerequisites in several applications in modern science, such as engineering, ion implantation and modification of materials, damage to electronics devices (e.g. space radiation), medical physics (e.g. proton therapy), among others. In this thesis we i) develop a rigorous experimental protocol to measure stopping power with high precision, and ii) investigate the collapse of the free electron gas (FEG) model in energy loss of light ions (protons) at a low energy range in transition and rare-earth metals. In the first part, we present an approach to obtain, with high accuracy, the stopping cross sections in the pure materials Al and Mo for protons in the energy range of [0.9 - 3.6] MeV by means of the transmission method. The traceability of the sources of uncertainties are fully evaluated and the final accuracy of the results is 0.63% (0.32% rand. and 0.54% syst.) for Al, and 1.5 % (0.44 % rand. and 1.4% syst.) for Mo, with both results primarily limited by the quality and homogeneity of the stopping foils. For Al, this high accuracy represents an improvement compared to the results obtained in previous studies and serves as a benchmark for our procedure. The most important sources of uncertainties were random - the uncertainty in the peak positions and in the Gaussian fits; and systematic - the non-uniformity thickness of the foils (a special procedure was developed to correct this). Even though the final uncertainty for Mo is higher than for Al, our results improve on the amount of data currently available for the energy range considered. Both data sets are compared with the most commonly employed theoretical models and Monte Carlo codes in the literature. In the second part, electronic stopping cross sections of nontrivial solids, that is, transition and rare earth metals (Ta and Gd) for slow protons are experimentally investigated, and the data were compared with the results for Pt and Au, to understand how energy losses in these metals are correlated with electronic band structures, and to understand the failure of the FEG model predictions. The higher stopping powers found for Ta and Gd cannot be explained by means of the FEG model; however, these effects are successfully correlated with the high density of states (DOS) of both the occupied and unoccupied electronic levels in these metals. For the case of Gd, the experimental data are extended in the energy range until the Bragg\'s peak is reached. The two parts of this thesis were published in Physical Review A 93 022704 (2016), and in Physical Review Letters 18 103401 (2017), respectively. / O fenômeno de perda de energia quando um íon interage com a matéria, também conhecido como poder de freamento, vem sendo investigado por mais de um século, gerando grandes descobertas. Entretanto, conseguir obter medidas experimentais com alta precisão, ou elaborar um completo entendimento teórico dos processos de perda de energia são tarefas extremamente difíceis e ainda muito requeridas pela comunidade científica. Além disso, dados de perda de energia são pré-requisitos em várias aplicações e ramos da ciência moderna, tais como: engenharia, implantação e modificação de materiais, danos em dispositivos eletrônicos (radiação espacial), física médica (próton terapia), etc. Esta tese tem dois focos: i) desenvolver um rigoroso protocolo experimental para medir stopping power com alta precisão e ii) investigar a quebra de validade do modelo de Gás de Elétrons Livres (FEG) para a perda de energia de prótons lentos em metais de transição e terra raras. Na primeira parte apresentamos uma abordagem experimental para obter com alta precisão o poder de freamento em materiais puros (Al e Mo) para prótons no intervalo de energia de [0,9 - 3,6] MeV pelo método de transmissão. A rastreabilidade das fontes de incerteza foi determinada e as incertezas finais encontradas foram: 0,63 % (0,32 % aleat. e 0,54 % sist.) para Al e 1,5 % (0,44 % aleat. e 1,4 % sist.) para Mo, ambas devido a qualidade e homogeneidade das folhas freadoras. Para Al, esta acurácia representa um avanço comparado com publicações anteriores e, assim, serviu como uma referência de nosso procedimento. As mais importantes fontes de incerteza foram: aleatória incerteza das posições dos picos e dos ajustes Gaussianos e sistemática não-uniformidade das folhas-alvo (um procedimento foi desenvolvido para corrigir isso). Embora a incerteza final do Mo é um pouco maior do que do Al, nossos resultados ajudaram a complementar a baixa quantidade de dados disponíveis para o intervalo de energia considerado. Ambos conjuntos de dados foram comparados com os mais comuns modelos teóricos e códigos de Monte Carlo na literatura. Para a segunda parte, poder de freamento em metais não tão comuns tais como transição (Ta) e terras-raras (Gd) para prótons com baixas velocidades foram experimentalmente investigados, e os dados comparados com resultados de Pt e Au, a fim de entender como o stopping power destes metais está correlacionado com as estruturas de bandas eletrônicas, e assim tentar explicar a falha do modelo de FEG. Os altos valores das perdas de energias encontradas para Ta e Gd não puderam ser explicadas pelo modelo de FEG, e portanto foram correlacionados com a densidade de estados (DOS) em ambos os níveis ocupados e não ocupados destes metais. Para o caso do Gd, os dados experimentais foram estendidos em um intervalo de energia até alcançarem o pico de Bragg. A primeira parte desta tese foi publicada na Physical Review A 93 022704 (2016), e a segunda parte na Physical Review Letters 18 103401 (2017).

Die Synthese kolloidaler Nanopartikel ist daher eines der wichtigsten Forschungsthemen der letzten Jahre. Die gepulste Laserablation in Flüssigkeiten stellt eine Alternative zu den nasschemischen Synthesewegen dar. Merkmale der gepulsten Laserablation in Flüssigkeiten ist der geringe experimentelle Aufwand, die Vielseitigkeit und die Möglichkeit stabilisatorfreie kolloidale Nanopartikel herzustellen. Eine Weiterentwicklung ist die Verwendung von Pulversuspensionen als Ausgangsmaterial. Dies führt zu höheren Produktivitäten und neuen Materialien. Die zugrundeliegenden Prozesse sind allerdings komplex. Um diese Mechanismen aufzuklären, wurde ein chemischer Ansatz gewählt. Als Ausgangsmaterial wurden daher verschiedene Kupferverbindungen (Cu2C2, Cu5Si, Cu3N, Cu(N3)2, Cu3P, Cu2O, CuO, Cu2S, CuS und CuI) verwendet. Die hergestellten Nanopartikel wurden mit Hilfe der analytischen Transmissionselektronenmikroskopie charakterisiert. Dadurch konnten nachgewiesen werden, dass zwei Mechanismen an der Nanopartikelbildung beteiligt sind. Die Laserbestrahlung von Ausgangsmaterialien wie CuO und Cu3N führt zur Bildung von metallischen Kupfernanopartikeln. In dem dabei erzeugten Plasma nukleieren die Kupferatome zunächst zu kleinen primären Nanopartikeln. Diese Partikel koaleszieren anschließend und bilden größere sekundäre Partikel. Im Gegensatz zu dieser reduktiven Ablation, resultiert die Laserbestrahlung von CuI in der Fragmentierung des Materials. In diesem Fall wird kein Plasma erzeugt, sondern der induzierte thermische Stress führt zur Fragmentierung des Kristalls unter Beibehaltung der chemischen Zusammensetzung. Die Frage, welcher der beiden Mechanismen für ein bestimmtes Ausgangsmaterial dominiert, ist entscheidend für potentielle Anwendungen der Methode, da hiervon die chemische Zusammensetzung der erhaltenen Nanopartikel abhängt. Dies wird am Beispiel der Synthese von Bi2Te3 Nanopartikeln diskutiert, die in thermoelektrischen Elementen zur Anwendung kommen können. / The synthesis of colloidal nanoparticles has become a major topic in recent years. The pulsed laser ablation in liquids poses an alternative to the common wet-chemical approaches. Key features of the pulsed laser ablation in liquids are its simple setup, its versatility, and the possibility to generate surfactant-free colloidal nanoparticles. A further development of this technique is the use of suspended powders instead of bulk targets. This leads to higher productivities and even new materials. Although the generation of colloids by irradiating a suspension is straight forward, the underlying mechanisms of the size reduction from micrometer to nanometer sized particles appear to be quite complex. In order to reveal the mechanism a chemical approach was chosen. Hence, various copper compounds (Cu2C2, Cu5Si, Cu3N, Cu(N3)2, Cu3P, Cu2O, CuO, Cu2S, CuS and CuI) were used as a model system in order to investigate the impact of the leaving group on the ablation process. The generated nanoparticles were characterized with analytical transmission electron microscopy. These investigations clearly show that there are two distinct mechanisms involved in nanoparticle formation. The laser irradiation of precursors like CuO and Cu3N results in the formation of metallic copper nanoparticles. In the generated plasma copper atoms nucleate and form small primary particles. These particles later coalesce to larger secondary particles. In contrast to this reductive ablation, the irradiation of CuI follows a fragmentation mechanism. Here, the absorbed power of the laser beam does not produce a plasma but introduces thermal stress leading to fragmentation of the crystal while the chemical composition is preserved. The question which mechanism is predominant is of utmost importance as the chemical composition of the nanoparticles depends on the formation process. This is discussed on the example of the synthesis of Bi2Te3 nanoparticles, which can be used in thermoelectric applications.

Cette étude, expérimentale et théorique, de physique des surfaces développe une méthode basée sur la spectroscopie de perte d'énergie des photoélectrons, pour déterminer les propriétés électroniques d'un matériau à partir de mesures XPS. Sur la base de la physique de la photoémission dans un solide homogène, la technique XPS-PEELS donne accès à la fonction de perte d'énergie ELF(E, q) liée à la partie imaginaire et la fonction diélectrique sur une large plage d’énergie (environ 50 eV) avec une sensibilité typique de ≈ 5 nm en profondeur. Dans les métaux ou les semi-conducteurs à faible gap, qui présentent un chevauchement important entre le pic quasi-élastique ZLP(E) et la région des pertes d'énergie, la technique est rendue applicable par la méthode de transformée de Fourier développée dans cette thèse. Les distributions en énergie de la source de rayons X et de la fonction d’appareil sont directement mesurées. La forme asymétrique du pic quasi-élastique ZLP(E) est obtenue à partir de la densité d’états électroniques (calcul par une méthode DFT), à l’aide du modèle de Hopfield-Wertheim-Citrin décrivant la réponse des électrons de valence à la création du trou (théorie multi-corps). L'algorithme XPS-PEELS utilise l'ensemble du spectre sans soustraction empirique d’une ligne de base. Il considère deux types d'excitations de plasmon - intrinsèques et extrinsèques - avec des taux de création différents mais impose la même distribution en énergie. Cette méthode originale permet d’accéder à des excitations électroniques de faible énergie (pertes proches du pic XPS) ; dans le cas de l'aluminium métallique, la transition inter-bandes est observée à 1.80 eV. En tant que technique de spectroscopie électronique, les effets de dispersion de la fonction de perte ELF(E, q) ont été considérés en utilisant un code TD-DFT (Exciting); quel que soit le niveau d'approximation (RPA, ALDA, LRC), les calculs ne décrivent pas correctement la largeur du spectre ELF expérimental (2.3 eV). À la fin du travail, nous comparons la méthode de la transformée de Fourier avec une méthode empirique d'élimination du pic élastique, valable pour les isolants, dans le cas de l'oxyde d'aluminium Al₂O₃. / This surface physics study, experimental and theoretical, develops a technique based on the energy loss spectroscopy of photoelectrons, to determine electronic properties of a material from XPS measurements. Based on the physics of photoemission in a homogeneous solid, the XPS-PEELS technique provides the energy loss function ELF(E, q) related to the imaginary part and the dielectric function with energy extension up to 50 eV and a typical sensitivity of ≈ 5 nm in depth. In metals or low-gap semiconductors, with important overlap between the elastic peak and the energy loss region, the technique became applicable by the Fourier transform method developed in this thesis. In addition to the distribution of the X-ray source and analyzer apparatus function, the asymmetric shape of the no-loss peak ZLP(E) is based on the calculation of the Density of Electronic States (DFT method) and the Hopfield-Wertheim-Citrin model describing the response of valence electrons to the creation of the hole (many body theory). The XPS-PEELS algorithm uses the entire spectrum without any empirical background subtraction. It considers two types of plasmon excitations - intrinsic and extrinsic - with different rates of creation but imposing the same distribution in energy. This original method was applied to the aluminum metal, allowing the analysis of the loss function and revealing interband excitations (1.80 eV) at low energy, very close to the no-loss peak. As an electron spectroscopy technique, dispersion effects of the loss function ELF(E, q) were considered using a TD-DFT code (Exciting); for any level of approximation (RPA, ALDA, LRC), the calculations do not describe properly the high width (2.3 eV) of the experimental ELF. At the end of the work, we compare the Fourier transform method with an empirical method of elastic peak elimination, valid for insulators, in the case of aluminum oxide Al₂O₃.

Surface structure, chemical composition, bonding configuration, film polarity, and electronic properties of InN layers grown by high pressure chemical vapor deposition (HPCVD) have been investigated. Sputtering at an angle of 50-70 degrees followed by atomic hydrogen cleaning (AHC) was successful in removing the carbon contaminants. AHC is found to be the most effective cleaning process to remove oxygen contaminants from InN layers in an ultrahigh vacuum (UHV) system and produced a well ordered surface. Auger electron spectroscopy (AES) confirmed the cleanliness of the surface, and low energy electron diffraction (LEED) yielded a 1×1 hexagonal pattern demonstrating a well-ordered surface. High resolution electron energy loss spectra (HREELS) taken from the InN layers exhibited loss features at 550 cm-1, 870 cm-1 and 3260 cm-1 which were assigned to Fuchs-Kliewer phonon, N-H bending, and N-H stretching vibrations, respectively. Assignments were confirmed by observation of isotopic shifts following atomic deuterium dosing. No In-H species were observed indicating N-termination of the surface and N-polarity of the film. Broad conduction band plasmon excitations were observed centered at 3100 cm-1 to 4200 cm-1 in HREEL spectra acquired with 25 eV electrons, for a variety of samples grown with different conditions. Infrared reflectance data shows a consistent result with HREELS for the bulk plasma frequency. The plasmon excitations are shifted about 300 cm-1 higher in HREEL spectra acquired using 7 eV electrons due to the higher plasma frequency and carrier concentration at the surface than in the bulk, demonstrating a surface electron accumulation. Hydrogen completely desorbed from the InN surface upon annealing for 900 s at 425 ºC or upon annealing for 30 s at 500 ºC. Fitting the coverage versus temperature for anneals of either 30 or 900 s indicated that the desorption was best described by second order desorption kinetics with an activation energy and pre-exponential factor of 1.3±0.2 eV and 10-7.3±1.0 cm2/s, respectively. Vibrational spectra acquired from HREEL can be utilized to explain the surface composition, chemical bonding and surface termination, and film polarity of InN layers. The explanation of evidence of surface electron accumulation and extraction of hydrogen desorption kinetic parameters can be performed by utilizing HREEL spectra.

Die vorliegende Arbeit befasst sich mit Elektronenenergieverlustspektroskopie an unterdotierten Kupratsupraleitern und Übergangsmetalldichalcogeniden. Nach einem kurzen Abriss über die der experimentellen Methode zugrundeliegenden theoretischen Tatsachen folgen zwei experimentelle Kapitel. Für das prototypische Kupratsystem Ca2-xNaxCuO2Cl2 wird für verschiedene Dotierungskonzentrationen zunächst die Entwicklung der Ladungstransferanregungen untersucht. Man findet eine substanzielle Umverteilung des spektralen Gewichtes, verbunden mit einem starken Einbruch der Dispersion dieser Anregungen. Beides wird im Rahmen der Wechselwirkung mit Spinfreiheitsgraden innerhalb der Kupfer-Sauerstoff-Ebene diskutiert. Anschliessend erfolgt die Diskussion einer ausschließlich für zehnprozentige Dotierung auftretenden Symmetriebrechung der optischen Antwortfunktion, für die verschiedene mögliche Szenarien vorgeschlagen werden. Im Kapitel über die Dichalcogenide liegt der Fokus auf dem Verhalten des Ladungsträgerplasmons, das für alle Substanzen dieser Gruppe mit Ladungsordnung eine negative Dispersion aufweist. Dieses Verhalten läßt sich durch in-situ Interkalation zusätzlicher Ladungstraeger umkehren, so dass man eine dotierungsabhängige Plasmonendispersion erhält. Es werden verschiedene Szenarien für dieses Verhalten diskutiert. / The present thesis describes electron energy-loss spectroscopy on underdoped cuprate superconductors and transition-metal dichalcogenides. After a brief introduction into the experimental method there are two experimental chapters. For the prototype cuprate system Ca2-xNaxCuO2Cl2 the behavior of the charge-transfer excitations is investigated as a function of doping. The observed substantial redistribution of spectral weight and the accompanying breakdown of their dispersion is discussed in terms of a coupling to the spin degrees of freedom within the copper-oxygen plane. For x=0.1 there is a pronounced symmetry breaking in the optical response function which is discussed in terms of different possible scenarios. The chapter on the dichalcogenides focuses on the properties of the charge-carrier plasmon which shows a negative dispersion for all representatives of this family exhibiting a charge-density wave instability. This behavior can be influenced by in-situ intercalation of additional charges, the result being a doping dependent plasmon dispersion. Several approaches to reconcile these findings are considered.

Der Hochwasserschutz stellt für die gegenwärtige Gesellschaft eine Thematik von zunehmender Aktualität dar, besonders unter Berücksichtigung der Verschärfung der extremen meteorologischen Ereignisse infolge des Klimawandels einerseits und der Vergrößerung des Schadenpotenzials in den Auengebieten andererseits. Zu den möglichen Schutzmaßnahmen im Flussmittellauf zählt die in ökologischer Hinsicht sehr effektive Deichrückverlegung (DRV). Um ihre Wirksamkeit bei der Dämpfung der Hochwasserspitzen zu untersuchen, wird eine Parameterstudie mit einem eindimensionalen hydronumerischen Modell (MIKE11) durchgeführt. Dabei wird besonderes Augenmerk auf die Hüllkurve der sich einstellenden Wassertiefe gerichtet, sowohl oberhalb als auch innerhalb bzw. unterhalb des DRV-Abschnitts. Zunächst wird der Effekt einer Flussaufweitung bei stationärer Strömung theoretisch analysiert. Ausgehend von den Strömungseigenschaften und vom Breitenverhältnis wird eine Bestimmungsgleichung für die effektive Mindestlänge LAeff einer Aufweitung abgeleitet, ab der eine Wasserspiegelabsenkung stromauf hervorgerufen wird. Die darauffolgenden Untersuchungen beschreiben die Auswirkungen einer solchen Maßnahme bei instationärem Abfluss auf zwei ausgewählte synthetische Hochwasserwellen unter jeweils vier verschiedenen Ansätzen zur Erfassung einer Aufweitung. Die Beiträge zur Retention vom Hauptgerinne und von der Querschnittsvergrößerung werden getrennt berücksichtigt. Darüber hinaus werden Überlegungen über das Verhalten von Hochwasserwellen mit gleichem Abflussscheitel aber unterschiedlicher Fülle beim Durchlaufen der Flussaufweitung angestellt. Der Gültigkeitsbereich des stationären Ansatzes für die eindimensionale numerische Simulation einer DRV wird mithilfe eines Kriteriums definiert, das auf einer für die Abminderung des Wassertiefenscheitels im Hauptgerinne semi-empirisch abgeleiteten Formel und auf einer Abschätzung der Retention in der DRV basiert. Die Anwendung des o. g. Kriteriums auf die betrachteten Hochwasserereignisse zeigt, dass eine stationäre Betrachtungsweise lediglich für extrem flache Hochwasserwellen (ab einer Anstiegszeit von einigen Tagen) gerechtfertigt ist. Durch die Ergebnisauswertung von 960 Kombinationen der geometrischen bzw. hydrodynamischen Parameter bei einer DRV im Hochwasserfall konnte der Zusammenhang zwischen den einzelnen Kenngrößen (Länge und Breite der DRV, Rauheit im Hauptgerinne bzw. im Vorland, Hochwasserganglinie) und der Auswirkung der DRV auf Wasserstand und Durchfluss hergestellt werden. Während oberhalb einer DRV mit einer Länge größer LAeff immer mit einer Wasserspiegelabsenkung zu rechnen ist, profitieren die Unterlieger von dieser Schutzmaßnahme nur bei kleineren Hochwasserereignissen oder bei Deichrückverlegungen mit erheblicher Ausdehnung. Dank einer gezielten Auswahl der Ausgangsgeometrie lassen sich die numerischen Ergebnisse auch auf den Elbabschnitt bei Dresden als Anhaltswerte übertragen. / Flood protection is a topic of increasing urgency in today’s society considering two main factors. Firstly, the growth in damage potential on the riparian zones and secondly, the increase in extreme meteorological events due to climate change. Among the possible flood protection measures for a river’s middle reach, dike relocation (DRV) is especially effective from an ecological point of view. To investigate its effectiveness in flood peak attenuation, a parametric study is performed using a one-dimensional hydro-numerical model (MIKE11). Special attention is given to the envelope of the resulting water depths upstream, downstream and within the DRV stretch. Initially, the effects of river enlargement under steady flow conditions are analysed theoretically. Key hydraulic parameters and the width ratio are used in deriving an equation to determine the minimum effective length LAeff of an enlargement that will reduce the upstream water depth. An investigation of an enlargement in unsteady flow conditions is then made, focussing on two specific synthetic flood waves with four different models for the enlargement. The contributions to discharge retention from the main channel and the cross section expansion are successfully separately examinated. Considerations are then made for the behaviour of flood waves passing the river enlargement with the same peak discharge but different volumes. A criterion for the validity range of the steady flow approach of the one-dimensional, numerical simulation of a DRV is then defined. This is based on a semi-empirically derived formula for water level attenuation in the main channel, and an estimate of discharge retention in the DRV. The application of this criterion to the observed flood events shows that the steady flow assumption is only valid for extremely flat flood waves (with time to peak of a few days or more). An analysis of the results from 960 combinations of geometric and hydraulic parameters for the DRV during a flood event is then made. This allowed to derive the relation between individual characteristics (length and width of the DRV, roughness in the main channel and flood plain, flood hydrograph) and the effect of the DRV on water level and discharge. Whilst one can always expect a decrease in the water level upstream of a DRV with length greater than LAeff , the riverside dwellings downstream would only benefit in small scale flood events or with a dike relocation of extensive dimensions. Thanks to the purposefully selected initial geometry, the numerical results can now provide reference values for Dresden’s reach of the river Elbe. / La difesa dalle piene rappresenta per la società odierna una tematica di crescente attualità se si considera in particolare da una parte la crescita del potenziale di danno nelle zone di pertinenza fluviale e dall’altra l’acuirsi di eventi meteorologici estremi dovuto al cambiamento climatico in atto. Tra i possibili interventi di protezione dalle piene nei fiumi di pianura si colloca l’espansione golenale, misura che risulta oltretutto assai efficace dal punto di vista ecologico. Al fine di indagarne anche l’efficacia nei confronti della laminazione del colmo di piena si effettua uno studio parametrico con un programma di modellazione idrodinamica monodimensionale (MIKE11), ponendo particolare attenzione all’inviluppo dei tiranti nel tratto dell’espansione, nonché nei tratti confinanti rispettivamente a monte e a valle della stessa. Dapprima si analizza dal punto di vista teorico l’effetto di un allargamento fluviale in moto stazionario. A partire dalle condizioni idrauliche della corrente e dal rapporto di restringimento si ricava una formula per la lunghezza minima efficace LAeff dell’allargamento, a partire dalla quale si genera un abbassamento del pelo libero nel tratto di monte. Nella successiva analisi di un allargamento in moto vario se ne studia l’effetto su due onde di piena sintetiche e per quattro differenti schematizzazioni dell’allargamento, considerando separatamente i contributi alla laminazione dovuti all’alveo principale e all’allargamento della sezione. Ciò consente tra l’altro di trarre conclusioni sul comportamento di onde di piena aventi la stessa portata al colmo ma differente volume al passare per l’allargamento. Per mezzo di un criterio basato su una formula semiempirica della riduzione del tirante al colmo per l’alveo principale e su una stima della laminazione all’interno dell’espansione golenale, sviluppate nell’ambito di questo lavoro, é possibile definire il campo di validità dell’approccio stazionario alla simulazione monodimensionale di un’espansione golenale. L’applicazione di tale criterio agli eventi di piena presi in considerazione rivela che l’approccio stazionario si giustifica solo per onde di piena estremamente piatte (tempo di crescita a partire da alcuni giorni). Attraverso l’analisi dei risultati di 960 combinazioni dei parametri geometrico-idraulici di un’espansione golenale al passaggio di un’onda di piena si è potuto ricavare il rapporto tra le singole grandezze significative (lunghezza e larghezza dell’espansione golenale, scabrezza dell’alveo e della golena, idrogramma dell’onda di piena) e l’effetto dell’espansione sui tiranti e sulle portate. Se a monte di un’espansione golenale di lunghezza superiore a LAeff si presenta sempre una abbassamento del pelo libero, a valle della stessa si può trarre profitto da un simile intervento solo nel caso di onde di piena modeste o per espansioni golenali di dimensioni notevoli. Una scelta mirata della geometria iniziale consente di estendere i risultati numerici, quali valori indicativi, al tratto del fiume Elba all’altezza di Dresda.

Structural characterization is essential for the understanding of the magnetic properties of thin films and multilayers. In this thesis, both crystalline and amorphous thin films and multilayers were analyzed utilizing transmission electron microscopy (TEM). High resolution TEM and electron diffraction studies emphasize on the growth of amorphous Fe91Zr9 and Co68Fe24Zr8 on both Al2O3 and Al70Zr30 in multilayer structures by magnetron sputtering. The properties of the growth surfaces were found to strongly influence the formation of nano-crystallites of the magnetic material at interfaces. Field induced uniaxial magnetic anisotropy was found to be possible to imprint into both fully amorphous and partially crystallized Co68Fe24Zr8 layers, yielding similar magnetic characteristics regardless of the structure. These findings are important for the understanding of both growth and magnetic properties of these amorphous thin films. As magnetic systems become smaller, new analysis techniques need to be developed. One such important step was the realization of electron energy-loss magnetic circular dichroism (EMCD) in the TEM, where information about the ratio of the orbital to spin magnetic moment (mL/mS) of a sample can be obtained. EMCD makes use of angular dependent inelastic scattering, which is characterized using electron energy-loss spectroscopy. The work of this thesis contributes to the development of EMCD by performing quantitative measurements of the mL/mS ratio. Especially, methods for obtaining energy filtered diffraction patterns in the TEM together with analysis tools of the data were developed. It was found that plural inelastic scattering events modify the determination of the mL/mS ratio, wherefore a procedure to compensate for it was derived. Additionally, utilizing special settings of the electron gun it was shown that EMCD measurements becomes feasible on the nanometer level through real space maps of the EMCD signal.