Refine
Has Fulltext
- yes (11)
Is part of the Bibliography
- yes (11)
Year of publication
- 2008 (11) (remove)
Document Type
- Doctoral Thesis (11)
Keywords
- Aktiver galaktischer Kern (2)
- Astrophysik (2)
- Computersimulation (2)
- Epitaxie (2)
- Festkörpertheorie (2)
- Gammastrahlung (2)
- Heteroepitaxie (2)
- Hubbard-Modell (2)
- Hydrodynamik (2)
- Kristallwachstum (2)
- Monte Carlo (2)
- Monte-Carlo-Simulation (2)
- Simulation (2)
- Strain (2)
- Turbulenz (2)
- Verspannung (2)
- AMR (1)
- BL-Lacertae-Objekt (1)
- Beam polarisation (1)
- Blazar (1)
- CP-Parität (1)
- CP-Veretzung (1)
- CP-Violation (1)
- Correlation Functions (1)
- Cuprate (1)
- Dislocations (1)
- Dynamical Cluster Approximation (1)
- Dynamische Cluster Approximation (1)
- Elektron-Elektron-Streuung (1)
- Epitaxy (1)
- FEARLESS (1)
- Flavourmischung (1)
- Galaxienentstehung (1)
- Galaxienentwicklung (1)
- Galaxienhaufen (1)
- Gamma-Astronomie (1)
- Gravitation (1)
- Hochtemperatursupraleiter (1)
- Hochtemperatursupraleitung (1)
- ILC (1)
- Interstellare Materie (1)
- Korrelationsfunktionen (1)
- Kubisch flächenzentriertes Gitter (1)
- LES (1)
- LFV (1)
- Leptonflavourverletzung (1)
- Linearbeschleuniger (1)
- Linearcollider (1)
- MAGIC-Teleskop (1)
- MSSM (1)
- Magnetismus (1)
- Maximum Entropie Methode (1)
- Maximum Entropy Method (1)
- Metall-Isolator-Phasenumwandlung (1)
- Mott-Übergang (1)
- Numerische Strömungssimulation (1)
- PG 1553+113 (1)
- Quantengravitation (1)
- Quantenkosmologie (1)
- RSB (1)
- Selbstenergiefunktional-Theorie (1)
- Self-energy-functional theory (1)
- Slepton (1)
- Spin-Spin-Wechselwirkung (1)
- Spin-Struktur (1)
- Sternentstehung (1)
- Strahlpolarisation (1)
- Subgrid-Skalen Modell (1)
- Supersymmetrie (1)
- Teilchenspektrum (1)
- Tesla <Teilchenbeschleuniger> (1)
- Variational cluster approximation (1)
- Versetzungen (1)
- active galactic nuclei (1)
- active galactic nucleus (1)
- astrophysics (1)
- cherenkov telescope (1)
- gamma astronomy (1)
- gamma radiation (1)
- glassy systems (1)
- hydrodynamics (1)
- qravitation (1)
- quantum cosmology (1)
- quantum gravity (1)
- spin glasses (1)
- star formation (1)
- subgrid-scale model (1)
- turbulence (1)
Institute
- Institut für Theoretische Physik und Astrophysik (11) (remove)
Die Untersuchung stark korrelierter Elektronensysteme anhand des zweidimensionalen Hubbard-Modells bildet das zentrale Thema dieser Arbeit. Wir analysieren das Schicksal des Mott-Isolators bei Dotierung als auch bei Reduzierung der Wechselwirkungsstärke. Die numerische Auswertung erfolgt mit Hilfe von Quanten-Cluster-Approximationen, die eine thermodynamisch konsistente Beschreibung der Grundzustandseigenschaften garantieren. Der hier verwendete Rahmen der Selbstenergiefunktional-Theorie bietet eine große Flexibilität bei der Konstruktion von Cluster-Näherungen. Eine detaillierte Analyse gibt Aufschluss über die Qualität und das Konvergenzverhalten unterschiedlicher Cluster-Näherungen innerhalb der Selbstenergiefunktional-Theorie. Wir verwenden für diese Untersuchungen das eindimensionale Hubbard-Modell und vergleichen unsere Resultate mit der exakten Lösung. In zwei Dimensionen finden wir als Grundzustand des Teilchen-Loch-symmetrischen Modells bei Halbfüllung einen antiferromagnetischen Isolator unabhängig von der Wechselwirkungsstärke. Die Berücksichtigung kurzreichweitiger räumlicher Korrelationen durch unsere Cluster-Näherung führt, im Vergleich mit der dynamischen Mean-Field-Theorie, zu einer deutlichen Verbesserung des antiferromagnetischen Ordnungsparameters. Darüberhinaus beobachten wir in der paramagnetischen Phase einen Metall-Isolator-Übergang als Funktion der Wechselwirkungsstärke, der sich qualitativ vom reinen Mean-Field-Szenario unterscheidet. Ausgehend vom antiferromagnetischen Mott-Isolator zeigt sich ein füllungsgetriebener Metall-Isolator-Übergang in eine paramagnetische metallische Phase. Abhängig von der verwendeten Cluster-Approximation tritt dabei zunächst eine antiferromagnetische metallische Phase auf. Neben langreichweitiger antiferromagnetischer Ordnung haben wir in unseren Rechnungen auch Supraleitung berücksichtigt. Das Verhalten des supraleitenden Ordnungsparameters als Funktion der Dotierung ist dabei in guter Übereinstimmung sowohl mit anderen numerischen Verfahren als auch mit experimentellen Ergebnissen.
We model Milky Way like isolated disk galaxies in high resolution three-dimensional hydrodynamical simulations with the adaptive mesh refinement code Enzo. The model galaxies include a dark matter halo and a disk of gas and stars. We use a simple implementation of sink particles to measure and follow collapsing gas, and simulate star formation as well as stellar feedback in some cases. We investigate two largely different realizations of star formation. Firstly, we follow the classical approach to transform cold, dense gas into stars with an fixed efficiency. These kind of simulations are known to suffer from an overestimation of star formation and we observe this behavior as well. Secondly, we use our newly developed FEARLESS approach to combine hydrodynamical simulations with a semi-analytic modeling of unresolved turbulence and use this technique to dynamically determine the star formation rate. The subgrid-scale turbulence regulated star formation simulations point towards largely smaller star formation efficiencies and henceforth more realistic overall star formation rates. More work is necessary to extend this method to account for the observed highly supersonic turbulence in molecular clouds and ultimately use the turbulence regulated algorithm to simulate observed star formation relations.
It is aim of this work to develop, implement, and apply a new numerical scheme for modeling turbulent, multiphase astrophysical flows such as galaxy cluster cores and star forming regions. The method combines the capabilities of adaptive mesh refinement (AMR) and large-eddy simulations (LES) to capture localized features and to represent unresolved turbulence, respectively; it will be referred to as Fluid mEchanics with Adaptively Refined Large-Eddy SimulationS or FEARLESS.
In this thesis, the low-temperature regime of replica symmetry breaking in the SK-model has been thoroughly investigated. In order to access this regime and to perform self-consistence calculations with high accuracy at high orders of replica symmetry breaking, a formalism has been developed which reduces the numerical effort to the absolute minimum. The central idea of its derivation is the identification of asymptotic regions in which the recursion relations can be solved analytically. The new object in the numerical treatment is then the correction to this asymptotic regime, represented by a sequence of so-called kernel correction functions. This method increased the effciency of the numerics considerably so that up to 200 orders of RSB could be calculated at zero temperature and zero external field, and up to 60 (65) orders of RSB for finite temperature (external field). The remarkable high precision of these calculations allowed the extraction of several quantities with accuracy exceeding the literature values by several orders of magnitude. The results of the numerical calculations have been analyzed in great detail. Especially the convergence behavior of various observables and of the order function with respect to the RSB order has been investigated since the high but finite RSB regime has been addressed in the present work for the first time. Several unexpected features of finite order replica symmetry breaking have been observed.
This thesis is dedicated to a theoretical study of the 1-band Hubbard model in the strong coupling limit. The investigation is based on the Dynamical Cluster Approximation (DCA) which systematically restores non-local corrections to the Dynamical Mean Field approximation (DMFA). The DCA is formulated in momentum space and is characterised by a patching of the Brillouin zone where momentum conservation is only recovered between two patches. The approximation works well if k-space correlation functions show a weak momentum dependence. In order to study the temperature and doping dependence of the spin- and charge excitation spectra, we explicitly extend the Dynamical Cluster Approximation to two-particle response functions. The full irreducible two-particle vertex with three momenta and frequencies is approximated by an effective vertex dependent on the momentum and frequency of the spin and/or charge excitations. The effective vertex is calculated by using the Quantum Monte Carlo method on the finite cluster whereas the analytical continuation of dynamical quantities is performed by a stochastic version of the maximum entropy method. A comparison with high temperature auxiliary field quantum Monte Carlo data serves as a benchmark for our approach to two-particle correlation functions. Our method can reproduce basic characteristics of the spin- and charge excitation spectrum. Near and beyond optimal doping, our results provide a consistent overall picture of the interplay between charge, spin and single-particle excitations: a collective spin mode emerges at optimal doping and sufficiently low temperatures in the spin response spectrum and exhibits the energy scale of the magnetic exchange interaction J. Simultaneously, the low energy single-particle excitations are characterised by a coherent quasiparticle with bandwidth J. The origin of the quasiparticle can be quite well understood in a picture of a more or less antiferromagnetic ordered background in which holes are dressed by spin-excitations to allow for a coherent motion. By increasing doping, all features which are linked to the spin-polaron vanish in the single-particle as well as two-particle spin response spectrum. In the second part of the thesis an analysis of superconductivity in the Hubbard model is presented. The superconducting instability is implemented within the Dynamical Cluster Approximation by essentially allowing U(1) symmetry breaking baths in the QMC calculations for the cluster. The superconducting transition temperature T_c is derived from the d-wave order parameter which is directly estimated on the Monte Carlo cluster. The critical temperature T_c is in astonishing agreement with the temperature scale estimated by the divergence of the pair-field susceptibility in the paramagnetic phase. A detailed study of the pseudo and superconducting gap is continued by the investigation of the local and angle-resolved spectral function.
Supersymmetry is currently the best motivated extension of the Standard Model and will be subject to extensive studies in the upcoming generation of colliders. The e-e- mode would be a straight forward extension to the currently planed International Linear Collider, planned to operate in e+e- mode. The low background in this mode may prove advantageous in the study of CP- and Lepton Flavour Violtation. In this work a CP sensitive observable based on transverse beam polarisation is introduced and the impact of neutralino mixing on the total cross section in cas of non-vanishing CP-violtating phases is studied in representative scenarios including non-GUT scenarios. Additionally, the mixing of sleptons is studied in the context of LFV, an analytical approximation is developed, and possible background free measurements of these effects are investigated.
This thesis is concerned with the description of macroscopic geometries through Loop Quantum Gravity, and there particularly with the description of cosmology within full Loop Quantum Gravity. For this purpose we depart from two distinct (classically virtually equivalent) ansätze: One is phase space reduction and the other is the restriction to particular states. It turns out that the quantum analogue of these two approaches are fundamentally different: The quantum analogue of phase space reduction needs the reformulation in terms of the observable Poisson algebra, so it can be applied to the noncommutative quantum phase space: It rests on the observation that the observable Poisson algebra of classical canonical cosmology is induced by the embedding of the reduced cosmological phase space into the phase space of full General Relativity. Using techniques related to Rieffel-induction, we develop a construction for a noncommutative embedding that has a classical limit that is described by a Poisson embedding. To be able to use this class of noncommutative embeddings for Loop Quantum Gravity, one needs a complete group of diffeomorphisms for the quantum theory, which is constructed. These two results are applied to construct a quantum embedding of a cosmological sector into full Loop Quantum Gravity. The embedded cosmological sector turns out to be discrete, like standard Loop Quantum Cosmology and can be interpreted as a super-selection sector thereof; however due to pathologies of the dynamics of full Loop Quantum Gravity, one can not induce a meaningful dynamics for this cosmological sector. The quantum analogue of restricting the space of states is achieved by explicitly constructing states for Loop Quantum Gravity with smooth geometry. These states do not exist within the Hilbert space of Loop Quantum Gravity, but as states on the observable algebra of Loop Quantum Gravity. This observable algebra is built from spin network functions, area operators and a restricted set of fluxes. For this algebra to be physically complete, we needed to construct a version of Loop Quantum Geometry based on a fundamental area operator. This version of Loop Quantum Geometry is constructed. Since the smooth geometry states are not in the Hilbert space of standard Loop Quantum Gravity, we needed to calculate the Hilbert space representation that contains them using the GNS construction. This representation of the observable algebra can be illustrated as a classical condensate of geometry with quantum fluctuations thereon. Using these representations we construct a quantum-minisuperspace, which allows for an interpretation of standard Loop Quantum Cosmology in terms of these states and led us to conjecture a new approach for the implementation of dynamics for Loop Quantum Gravity.
Blazars are among the most luminous sources in the universe. Their extreme short-time variability indicates emission processes powered by a supermassive black hole. With the current generation of Imaging Air Cherenkov Telescopes, these sources are explored at very high energies. Lowering the threshold below 100 GeV and improving the sensitivity of the telescopes, more and more blazars are discovered in this energy regime. For the MAGIC telescope, a low energy analysis has been developed allowing to reach energies of 50 GeV for the first time. The method is presented in this thesis at the example of PG 1553+113 measuring a spectrum between 50 GeV and 900 GeV. In the energy regime observed by MAGIC, strong attenuation of the gamma-rays is expected from pair production due to interactions of gamma-rays with low-energy photons from the extragalactic background light. For PG 1553+113, this provides the possibility to constrain the redshift of the source, which is still unknown. Well studied from radio to x-ray energies, PG 1553+113 was discovered in 2005 in the very high energy regime. In total, it was observed with the MAGIC telescope for 80~hours between April 2005 and April 2007. From more than three years of data taking, the MAGIC telescope provides huge amounts of data and a large number of files from various sources. To handle this data volume and to provide monitoring of the data quality, an automatic procedure is essential. Therefore, a concept for automatic data processing and management has been developed. Thanks to its flexibility, the concept is easily applicable to future projects. The implementation of an automatic analysis is running stable since three years in the data center in Würzburg and provides consistent results of all MAGIC data, i.e. equal processing ensures comparability. In addition, this database controlled system allows for easy tests of new analysis methods and re-processing of all data with a new software version at the push of a button. At any stage, not only the availability of the data and its processing status is known, but also a large set of quality parameters and results can be queried from the database, facilitating quality checks, data selection and continuous monitoring of the telescope performance. By using the automatic analysis, the whole data sample can be analyzed in a reasonable amount of time, and the analyzers can concentrate on interpreting the results instead. For PG 1553+113, the tools and results of the automatic analysis were used. Compared to the previously published results, the software includes improvements as absolute pointing correction, absolute light calibration and improved quality and background-suppression cuts. In addition, newly developed analysis methods taking into account timing information were used. Based on the automatically produced results, the presented analysis was enhanced using a special low energy analysis. Part of the data were affected by absorption due to the Saharan Air Layer, i.e. sanddust in the atmosphere. Therefore, a new method has been developed, correcting for the effect of this meteorological phenomenon. Applying the method, the affected data could be corrected for apparent flux variations and effects of absorption on the spectrum, allowing to use the result for further studies. This is especially interesting, as these data were taken during a multi-wavelength campaign. For the whole data sample of 54 hours after quality checks, a signal from the position of PG 1553+113 was found with a significance of 15 standard deviations. Fitting a power law to the combined spectrum between 75 GeV and 900 GeV, yields a spectral slope of 4.1 +/- 0.2. Due to the low energy analysis, the spectrum could be extended to below 50 GeV. Fitting down to 48 GeV, the flux remains the same, but the slope changes to 3.7 +/- 0.1. The determined daily light curve shows that the integral flux above 150 GeV is consistent with a constant flux. Also for the spectral shape no significant variability was found in three years of observations. In July 2006, a multi-wavelength campaign was performed. Simultaneous data from the x-ray satellite Suzaku, the optical telescope KVA and the two Cherenkov experiments MAGIC and H.E.S.S. are available. Suzaku measured for the first time a spectrum up to 30 keV. The source was found to be at an intermediate flux level compared to previous x-ray measurements, and no short time variability was found in the continuous data sample of 41.1 ksec. Also in the gamma regime, no variability was found during the campaign. Assuming a maximum slope of 1.5 for the intrinsic spectrum, an upper limit of z < 0.74 was determined by deabsorbing the measured spectrum for the attenuation of photons by the extragalactic background light. For further studies, a redshift of z = 0.3 was assumed. Collecting various data from radio, infrared, optical, ultraviolet, x-ray and gama-ray energies, a spectral energy distribution was determined, including the simultaneous data of the multi-wavelength campaign. Fitting the simultaneous data with different synchrotron-self-compton models shows that the observed spectral shape can be explained with synchrotron-self-compton processes. The best result was obtained with a model assuming a log-parabolic electron distribution.
At the beginning of regular observations with the MAGIC telescope in December 2004, all but one extragalactic sources detected at very high energy (VHE) gamma-rays belonged to the class of high frequency peaked BL Lac (HBL) objects. This motivated a systematic scan of candidate sources to increase the number of known sources and to study systematically their spectral properties. As candidate sources for VHE emission, X-ray bright HBLs were selected from a compilation of active galactic nuclei. The MAGIC observations took place from December 2004 to March 2006. The declination of the objects was restricted to values between -1.2° and +58.8° corresponding to a maximum zenith distance lower than 30° at culmination. Since gamma-rays are absorbed by photo-pair production in low energy background radiation fields, the redshift of the investigated objects was limitetd to z < 0.3. Under the assumption that HBLs generally emit the same energy flux at 1keV as at 200GeV, only the brightest X-ray sources were observed, leading to a cut in the X-ray flux of F(1keV) > 2µJy}. Of the fourteen sources observed, four have been detected: 1ES 1218+304 (for the first time at very high energies), 1ES 2344+514 (strong detection in a state of low activity), Mrk 421 and Mrk 501. A hint of a signal on a 3-sigma-level from the direction of 1ES 1011+496 has been observed. In the meantime the object has been confirmed as a source of VHE gamma-rays by a second MAGIC observation campaign triggered by an optical outburst. For ten sources, upper limits on their integral fluxes above 200GeV have been calculated on a 99% confidence level. To cross calibrate the different data samples, collected during 14 months, bright muon ring images have been used, recorded as background events by the MAGIC telescope. Based on the development by Meyer (2003), the method has been improved and implemented into the automatic data analysis as a continuous monitor of the calibration and the point spread function of the optical system. While the ring images are generated by muons with small impact parameters, it could be shown that the image parameter distributions for muons with large impact parameters and gamma showers completely overlap, revealing these muons as the dominant background for gamma-ray observations below energies of 150GeV. The sample of HBLs (including all HBLs detected at VHE so far) has been investigated for correlations between broad-band spectral indices as determined from simultaneous optical, archival X-ray and radio luminosities, finding that the VHE emitting HBLs do not differ from the non-detected ones. In general the absorption corrected HBL gamma-ray luminosities at 200GeV are not higher than their X-ray luminosities at 1keV. Based on a complete X-ray BL Lac sample, the Hamburg/ROSAT X-ray BL Lac sample, the number of expected VHE sources has been estimated for the performed scan, finding a consistent number under the assumption of a 37% completeness of the investigated sample and a 1keV-to-200GeV luminosity ratio of 1.4. An upper limit on the omnidirectional flux at 200GeV has been calculated by interpolating the sum over the observed fluxes and upper limits. Within the uncertainties, the result is in agreement with the expectations derived from the X-ray luminosity function of BL Lacs. For 1ES 1218+304 and 1ES 2344+514 the lightcurves have been derived, showing evidence for flux variability on a time scale of 17 days and 24h, respectively. In the case of 1ES 1218+304 variability has been reported for the first time at VHEs. For both sources the energy spectra have been reconstructed and discussed in the context of their broad band spectral energy distribution (SED), using a single zone synchrotron self Compton model. The SEDs are well fitted by the simulation even though the very high peak frequencies at gamma-rays push the model to its limits. The parameters derived from the simulation are in good agreement with the parameters found for similar HBLs.
In this PhD thesis, the effect of strain on heteroepitaxial growth is investigated by means of Kinetic Monte Carlo simulations. In this context the lattice misfit, arising from the different lattice constants of the adsorbate and the substrate material, is of particular interest. As a consequence, this lattice misfit leads to long-range elastic strain effects having strong influence on the entire growing crystal and its resulting surface morphology. The main focus of this work is the investigation of different strain relaxation mechanisms and their controlling parameters, revealing interesting consequences on the subsequent growth. Since epitaxial growth is carried out under conditions far away from thermodynamic equilibrium, it is strongly determined by surface kinetics. At this point the relevant kinetic microscopic processes are described, followed by theoretical considerations of heteroepitaxial growth disclosing an overview over several independent methodological streams, used to model epitaxy in different time and length scales, as well as the characterization of misfit dislocations and the classification of epitaxial growth modes based on thermodynamic considerations. The epitaxial growth is performed by means of Kinetic Monte Carlo simulations which allows for the consideration of long range effects in systems with lateral extension of few hundred atoms. By using an off-lattice simulation model the particles are able to leave their predefined lattice sites, which is an indispensable condition for simulating strain relaxation mechanisms. The main idea of our used model is calculating the activation energy of all relevant thermally activated processes by using simple pair potentials and then realizing the dynamics by performing each event according to its probability by means of a rejection-free algorithm method. In addition, the crystal relaxation procedure, the grid-based particle access method, which accelerates the simulation enormously, and the efficient implementation of the algorithm are discussed. To study the influence of long range elastic strain effects, the main part of this work was realized on the two dimensional triangular lattice, which can be treated as a cross section of the real three dimensional case. Chapter 4 deals with the formation of misfit dislocations as a strain relaxation mechanism and the resulting consequences on the subsequent heteroepitaxial growth. We can distinguish between two principally different dislocation formation mechanisms, depending strongly on the sign as well as on the magnitude of the misfit, but also the surface kinetics need to be taken into account. Additionally, the dislocations affect the lattice spacings of the crystal whose observed progression is in qualitative good agreement with experimental results. Furthermore, the dislocations influence the subsequent growth of the adsorbate film, since the potential energy of an adatom is modulated by buried dislocations. A clear correlation between the lateral positions of buried dislocations and the positions of mounds grown on the surface can be observed. In chapter 5, an alternative strain relaxation mechanism is studied: the formation of three dimensional islands enables the particles to approach their preferred lattice spacing. We demonstrate that it is possible to adjust within our simulation model each of the three epitaxial growth modes: Volmer–Weber, Frank–van der Merve or layer-by-layer, and Stranski–Krastanov growth mode. Moreover, we can show that the emerging growth mode depends in principle on two parameters: on the one hand the interaction strength of adsorbate particles with each other, compared to the interaction of adsorbate with substrate particles, and on the other hand the lattice misfit between adsorbate and substrate particles. A sensible choice of these two parameters allows the realization of each growth mode within the simulations. In conclusion, the formation of nanostructures controlled by an underlying dislocation network can be applied in the concept of self-organized pattern formation as well as by the tendency to form ordered arrays of strain-induced three dimensional grown islands. In chapter 6, we extend our model to three dimensions and investigate the effect of strain on growth on bcc(100) surfaces. We introduce an anisotropic potential yielding a stable bcc lattice structure within the off-lattice representation. We can show that the strain built up in submonolayer islands is mainly released at the island edges and the lattice misfit has strong influence on the diffusion process on the plane surface as well as on the situation at island edges with eminent consequences on the appearance of submonolayer islands.