In numerical simulations, problems stemming from aerodynamics pose many challenges for the method used. Some of these are addressed in this thesis, such as the fluid interacting with objects, the presence of shocks, and various types of boundary conditions. Scenarios of the kind mentioned above are described mathematically by initial boundary value problems (IBVPs). We discretize the IBVPs using high order accurate finite difference schemes on summation by parts form (SBP), combined with weakly imposed boundary conditions, a technique called simultaneous approximation term (SAT). By using the energy method, stability can be shown. The weak implementation is compared to the more commonly used strong implementation, and it is shown that the weak technique enhances the rate of convergence to steady state for problems with solid wall boundary conditions. The analysis is carried out for a linear problem and supported numerically by simulations of the fully non-linear Navier–Stokes equations. Another aspect of the boundary treatment is observed for fluid structure interaction problems. When exposed to eigenfrequencies, the coupled system starts oscillating, a phenomenon called flutter. We show that the strong implementation sometimes cause instabilities that can be mistaken for flutter. Most numerical schemes dealing with flows including shocks are first order accurate to avoid spurious oscillations in the solution. By modifying the SBP-SAT technique, a conservative and energy stable scheme is derived where the order of accuracy can be lowered locally. The new scheme is coupled to a shock-capturing scheme and it retains the high accuracy in smooth regions. For problems with complicated geometry, one strategy is to couple the finite difference method to the finite volume method. We analyze the accuracy of the latter on unstructured grids. For grids of bad quality the truncation error can be of zeroth order, indicating that the method is inconsistent, but we show that some of the accuracy is recovered. We also consider artificial boundary closures on unbounded domains. Non-reflecting boundary conditions for an incompletely parabolic problem are derived, and it is shown that they yield well-posedness. The SBP-SAT methodology is employed, and we prove that the discretized problem is stable.

Conselho Nacional de Desenvolvimento CientÃfico e TecnolÃgico / Estudamos as estruturas e as propriedades dinÃmicas de um cristal clÃssico bidimensional (2D) em bicamadas de partÃculas dipolares magnÃticas carregadas em um arranjo no qual os dipolos sÃo perpendiculares Ãs camadas e com mesma densidade de partÃculas em cada camada. A energia do sistema à devido à interaÃÃo carga - carga (interaÃÃo coulom- biana) e a interaÃÃo dipolo - dipolo. Devido ao fato dessas interaÃÃes serem de longo alcance, usamos o mÃtodo da soma de Ewald para obter uma expressÃo para a energia envolvendo somas que convergem rapidamente. Comparando as energias de possÃveis geometrias do cristal, determinamos o diagrama de fase do sistema em funÃÃo do parÃmetro η (que està relacionado com a distÃncia entre as camadas de dipolos magnÃticos carregados e a densidade de partÃculas) e da intensidade relativa das interaÃÃes elÃtrica e magnÃtica. Mudando a intensidade relativa da interacÃo dipolo - dipolo com respeito à interaÃÃo elÃtrica, podemos encontrar seis diferentes estruturas cristalinas estÃveis em funÃÃo de η. Uma caracterÃstica interessante desse sistema à a possibilidade de permanecer em arranjos nos quais as camadas sÃo ou nÃo deslocadas uma em relaÃÃo a outra, apenas variando a interaÃÃo magnÃtica entre os dipolos, por exemplo, atravÃs de um campo magnÃtico externo. As transiÃÃes entre as estruturas cristalinas podem ser contÃnuas e descontÃnuas. No intuito de investigar a estabilidade das configuraÃÃes de mÃnima energia, calculamos o espectro dos fÃnons do sistema usando a aproximaÃÃo harmÃnica. Para isto, recorremos novamente a tÃcnica de Ewald para obter somas que convergem rapidamente. A anÃlise da relaÃÃo de dispersÃo (fÃnons) revela caracterÃsticas do sistema que sÃo de grande utilidade no estudo da transiÃÃo sÃlido-lÃquido (fusÃo). / We study the structural and dynamical properties of a two-dimensional (2D) classical bi-layer crystal of charged magnetic dipolar particles in a setup where the dipoles are oriented perpendicular to the layers and equal density of charged dipolar particles in each layer. The energy of the system is due to the charge - charge interaction (Coulomb interaction) and the dipole - dipole interaction. Due to the long-range nature of the interactions, we use the Ewald summation method to obtain an expression for the energy involving rapidly convergent sums. By comparing the energies of a number of possible crystal geometries, we determine the phase diagram of the system as a function of the parameter η (which is related to the separation between the layers of charged magnetic dipoles and the particle density) and the relative intensity of the magnetic and electrical interactions. By changing the relative intensity of the dipole - dipole interaction with respect to electrical one, we are able to find six diferent stable crystalline structures as a function of η. An interesting feature of the present model system is the possibility to tune between the matched and staggered arrangements by varying the magnetic interaction between the dipoles, e.g. through an external magnetic field. The phase boundaries of the crystalline structures consist of both continuous and discontinuous transitions. In order to investigate the stability of the minimum energy arrangements we also calculate the phonon spectra of the system within the harmonic approximation. In this case, we resort again on the Ewald technique to obtain the rapidly convergent sums. The analysis of the phonon spectra reveals interesting features which are useful in the study of melting.

The present thesis is dedicated to the efficient computation of electrostatic interactions in particle systems, which is of great importance in the field of molecular dynamics simulations. In order to compute the therefor required physical quantities with only O(N log N) arithmetic operations, so called particle-mesh methods make use of the well-known Ewald summation approach and the fast Fourier transform (FFT). Typically, such methods are able to handle systems of point charges subject to periodic boundary conditions in all spatial directions. However, periodicity is not always desired in all three dimensions and, moreover, also interactions to dipoles play an important role in many applications. Within the scope of the present work, we consider the particle-particle NFFT method (P²NFFT), a particle-mesh approach based on the fast Fourier transform for nonequispaced data (NFFT). An extension of this method for mixed periodic as well as open boundary conditions is presented. Furthermore, the method is appropriately modified in order to treat particle systems containing both charges and dipoles. Consequently, an efficient algorithm for mixed charge-dipole systems, that additionally allows a unified handling of various types of periodic boundary conditions, is presented for the first time. Appropriate error estimates as well as parameter tuning strategies are developed and verified by numerical examples. / Die vorliegende Arbeit widmet sich der Berechnung elektrostatischer Wechselwirkungen in Partikelsystemen, was beispielsweise im Bereich der molekulardynamischen Simulationen eine zentrale Rolle spielt. Um die dafür benötigten physikalischen Größen mit lediglich O(N log N) arithmetischen Operationen zu berechnen, nutzen sogenannte Teilchen-Gitter-Methoden die Ewald-Summation sowie die schnelle Fourier-Transformation (FFT). Typischerweise können derartige Verfahren Systeme von Punktladungen unter periodischen Randbedingungen in allen Raumrichtungen handhaben. Periodizität ist jedoch nicht immer bezüglich aller drei Dimensionen erwünscht. Des Weiteren spielen auch Wechselwirkungen zu Dipolen in vielen Anwendungen eine wichtige Rolle. Zentraler Gegenstand dieser Arbeit ist die Partikel-Partikel-NFFT Methode (P²NFFT), ein Teilchen-Gitter-Verfahren, welches auf der schnellen Fouriertransformation für nichtäquidistante Daten (NFFT) basiert. Eine Erweiterung dieses Verfahrens auf gemischt periodische sowie offene Randbedingungen wird vorgestellt. Außerdem wird die Methode für die Behandlung von Partikelsystemen, in denen sowohl Ladungen als auch Dipole vorliegen, angepasst. Somit wird erstmalig ein effizienter Algorithmus für gemischte Ladungs-Dipol-Systeme präsentiert, der zusätzlich die Behandlung sämtlicher Arten von Randbedingungen mit einem einheitlichen Zugang erlaubt. Entsprechende Fehlerabschätzungen sowie Strategien für die Parameterwahl werden entwickelt und anhand numerischer Beispiele verifiziert.

Freight transportation changes in the region of Stockholm lead municipalities to the need of finding ways to maintain and empower their current logistics role in the area. These changes are the operation of the future freight port in Norvik near Nynäshamn, the expansion of Stockholm leading to new freight transportation networks, and the extension of the double track line southern to Västerhaninge station. Haninge municipality is willing to keep the interest of existing companies in its land alive and perhaps increase its’ attractiveness by constructing an extension of the terminal in Jordbro. But before proceeding to that step, the municipality wants to investigate whether companies are aware of these changes and furthermore if they have plans for changing their current goods transportation patterns. It is also interested in finding out the factors that companies consider as the most important when they take decisions about their goods transportation plan. That way the municipality knows what is important for companies and can adjust the supply of infrastructure to their demands. The main method used for data collection is designing and conducting a questionnaire and for data analysis it is multi-criteria analysis (MCA). The questionnaire also includes questions about the nature of the company and the current ways of goods transportation. Although the number of responders is quite low, some general conclusions could be made. There are two alternatives competing and multi-criteria analysis leads to the selection of the most suitable one. There are several limitations and assumptions which can be overcome by further future research.

We consider diagonal-norm classical summation-by-parts (SBP) operators us-ing the simultaneous approximation term (SAT) method of imposing boundaryconditions. We derive a formula for the inverse of these SBP-SAT discretizationmatrices. This formula is then used to show that it is possible to construct a secondorder accurate SBP-SAT operator using only seven grid points.

No description available.

The present thesis provides a modularized view on the structure of fast numerical methods for computing Coulomb interactions between charged particles in three-dimensional space. Thereby, the common structure is given in terms of three self-contained algorithmic frameworks that are built on top of each other, namely fast Fourier transform (FFT), nonequispaced fast Fourier transform (NFFT) and NFFT based particle-mesh methods (P²NFFT). For each of these frameworks algorithmic enhancement and parallel implementations are presented with special emphasis on scalability up to hundreds of thousands of parallel processes. In the context of FFT massively parallel algorithms are composed from hardware adaptive low level modules provided by the FFTW software library. The new algorithmic NFFT concepts include pruned NFFT, interlacing, analytic differentiation, and optimized deconvolution in Fourier space with respect to a mean square aliasing error. Enabled by these generalized concepts it is shown that NFFT provides a unified access to particle-mesh methods. Especially, mixed-periodic boundary conditions are handled in a consistent way and interlacing can be incorporated more efficiently. Heuristic approaches for parameter tuning are presented on the basis of thorough error estimates. / Die vorliegende Dissertation beschreibt einen modularisierten Blick auf die Struktur schneller numerischer Methoden für die Berechnung der Coulomb-Wechselwirkungen zwischen Ladungen im dreidimensionalen Raum. Die gemeinsame Struktur ist geprägt durch drei selbstständige und auf einander aufbauenden Algorithmen, nämlich der schnellen Fourier-Transformation (FFT), der nicht äquidistanten schnellen Fourier-Transformation (NFFT) und der NFFT-basierten Teilchen-Gitter-Methode (P²NFFT). Für jeden dieser Algorithmen werden Verbesserungen und parallele Implementierungen vorgestellt mit besonderem Augenmerk auf massiv paralleler Skalierbarkeit. Im Kontext der FFT werden parallele Algorithmen aus den Hardware adaptiven Modulen der FFTW Softwarebibliothek zusammengesetzt. Die neuen NFFT-Konzepte beinhalten abgeschnittene NFFT, Versatz, analytische Differentiation und optimierte Entfaltung im Fourier-Raum bezüglich des mittleren quadratischen Aliasfehlers. Mit Hilfe dieser Verallgemeinerungen bietet die NFFT einen vereinheitlichten Zugang zu Teilchen-Gitter-Methoden. Insbesondere gemischt periodische Randbedingungen werden einheitlich behandelt und Versatz wird effizienter umgesetzt. Heuristiken für die Parameterwahl werden auf Basis sorgfältiger Fehlerabschätzungen angegeben.

In an organism's natural environment, there are always an uncountable number of stimuli, and stimulus features, available to gain control over behavior. When these component stimuli are presented simultaneously, this new stimulus compound can occasion a previously unseen effect on behavior. Stimulus compounding is a method used to better understand how variables in stimulus features may impact the final effect on an organism's responding when presented with a stimulus compound. While stimulus compounding experiments are often conducted using intermodal tone and light stimuli, it is conducted far less often using intramodal stimuli, potentially due to the competing stimulus features of same-modality stimuli. Even less conducted research has been done using two odor component stimuli, despite the large impact odor has on many species' behavior. The purpose of this study was to conduct a stimulus compounding experiment using intramodal ambient odor stimuli in rats, to see what kind of effect a mixed odor compound would have on the subject's behavior. This was done using a wind tunnel designed operant chamber, where both subjects were trained to respond to independently presented odor stimuli. Following training a compound mixture of both component odors was presented to the subjects. The results of this study demonstrated an averaged effect on behavior, producing a response rate under the compound mixture that was intermediate to the response rates under each independent component stimulus. This may be due to several factors, including the efficacy of initial training procedures, the efficacy of the apparatus, and the merging of the component stimuli's features.

No description available.

No description available.