Thesis (M.E.-Res.)--University of Wollongong, 2008. / Typescript. Includes bibliographical references: leaf 163-172.

Thesis (M.Sc.(Meteorology))--University of Pretoria, 2008. / Summary in English. Includes bibliographical references.

The observational approach to cosmology is the endeavour to reconstruct the geometry of the Universe using only data that is theoretically verifiable within the causal boundaries of a cosmological observer. Using this approach, it was shown in [36] that given ideal cosmological observations, the only essential assumption necessary to determine the geometry of the Universe is a theory of gravity. Assuming General Relativity, the full set of Einstein field equations (EFEs) can be used to reconstruct the geometry of the Universe using direct observations on the past null cone (PNC) as initial conditions. Observationally and theoretically this is a very ambitious task and therefore, current developments have been restricted to spherically symmetric dust models while only relaxing the usual assumption of homogeneity in the radial direction. These restricted models are important for the development of theoretical foundations and also useful as verification models since they avoid the circularity of verifying what has already been assumed. The work presented in this thesis is the development of such a model where numerical relativity (NR) is used to simulate the observable universe. Similar to the work of Ellis and co-workers [36], a reference frame based on the PNC is used. The reference frame used here, however, is based on that of the characteristic formalism of NR, which has developed for calculating the propagation of gravitational waves. This provides a formalism that is well established in NR, making the use of existing algorithms possible. The Bondi-Sachs coordinates of the characteristic formalism is, however, not suitable for calculations beyond the observer apparent horizon (AH) since the diameter distance used as a radial coordinate becomes multi-valued when the cosmological PNC reconverges in the history of a universe, smaller in the past. With this taken into consideration, the Bondi-Sachs characteristic formalism is implemented for cosmology and the problem approaching the AH is investigated. Further developments address the limitations approaching the AH by introducing a metric based on the Bondi-Sachs metric where the radial coordinate is replaced with an affine parameter. The model is derived with a cosmological constant Λ incorporated into the EFEs where Λ is taken as a parameter of the theory of gravity rather than as a matter source term. Similar to the conventional characteristic formalism, this model consists of a system of differential equations for numerically evolving the EFEs as a characteristic initial value problem (CIVP). A numerical code implemented for the method has been found to be second order convergent. This code enables simulations of different models given identical data on the initial null cone and provides a method to investigate their physical consistency within the causally connected region of our current PNC. These developments closely follow existing 3D schemes developed for gravitational wave simulations, which should make it natural to extend the affine CIVP beyond spherical symmetric simulations. The developments presented in this thesis is an extended version of two papers published earlier.

Orientador: Marco Aurelio Cremasco / Dissertação (mestrado) - Universidade Estadual de Campinas, Faculdade de Engenharia Quimica / Made available in DSpace on 2018-08-04T19:17:49Z (GMT). No. of bitstreams: 1 Starquit_AxelNicolasVincentReneDesire_M.pdf: 14668743 bytes, checksum: 4db90a43a39b4d77f2e635582b6323b4 (MD5) Previous issue date: 2004 / Resumo: O Brasil é dono de inigualável diversidade biológica, da qual é possível encontrar fármacos de altos valores agregados.A obtenção desses produtos atravessa diversas etapas, desde a identificação de um determinado principio ativo, até a sua recuperação e posterior purificação. São nessas etapas finais que se podem utilizar sistemas de separação baseados na adsorção e, em particular, a técnica cromatográfica de leito móvel simulado (LMS). Na descrição do LMS, é freqüente recorrer à sua modelagem e simulação. Por outro lado, ao se tratar de modelagens rigorosas, estas levam a problemas matemáticos de equações diferenciais parciais acopladas. Tendo em vista a complexidade de tais modelos, resoluções numéricas são necessárias, sendo que o método das colocações ortogonais é o que se encontra com freqüência na literatura para a simulação do LMS. O objetivo desta Dissertação, portanto, é o de desenvolver um método numérico alternativo destinado à resolução do modelo matemático que descreve o fenômeno e a operação do LMS, utilizando-se quatro zonas de separação, cada qual formada por leitos fixos, em sistemas descritos por isotermas lineares. O método matemático proposto é hibrido, comportando parte analítica para o balanço de massa da fase estacionária, a qual é relacionada através do teorema de Duhamel ao balanço de massa da fase móvel que, por sua vez, é resolvido numericamente. O aplicativo computacional desenvolvido foi avaliado por comparação com resultados experimentais advindos da literatura para mistura binária dos aminoácidos fenilalaninae tirosina; para mistura temária desses aminoácidos mais o triptofano e para uma mistura multicomponentes contendo o Pac1itaxel, conhecido como Taxol@, um poderoso agente cancerígeno. Os resultados obtidos demonstraram desempenho satisfatório de método proposto. Além do leito móvel simulado, o aplicativo pode ser avaliado para outros sistemas cromatográficos: curvas de ruptura ("breakthrough") que descrevem a cinética de adsorção em batelada utilizando leito fixo; respostas de pulsos cromatográficos em leito fixo, o que leva às curvas de eluição do pulso. A presente Dissertação de Mestrado foi desenvolvida no Laboratório de Processos em Meios Porosos (LPMP), do Departamento de Termotluidodinâmica (DTF), da Faculdade de Engenharia Química (FEQ),na Universidade Estadual de Campinas (Unicamp) / Abstract: In Brazil, the biodiversity is huge so this country may become one of the greatest bioproducts producers worldwide. When extracted fiom their natural source, those bioproducts are generally obtained in a mixture containing many other compounds, such that, in order to obtain a commercialproduct, the mixture must go through separation and purification processes. In the field of bioprocucts separation, the simulated moving bed process (SMB) has gained much attention since the beginning of the 90th. In modem engineering, modeling and simulating processes has become very common, from the beginning of a new project to the process control in the existing production plants. For simulated moving bed, when rigorous models are used, the model takes the form of a system of coupled partials differentials equations, so that numerical resolutions of the mathematicalproblem are needed. For simulatedmoving bed (SMB) the numerical method of orthogonal collocations is the one most frequently encountered in the literature. This Dissertation presents an altemative numerical method for the simulated moving bed when isotherm can be considered linear. The method proposed, composed of both analytical and numerical schemes which are correlated through Duhamel's theorem, is hybrid. Most of the mass transfer parameters necessary for simulationsare estimated by the use of correlations from literature. Simulation's results for the separation of pharmaceutical's products, using the proposed numerical method, are obtained and compared with experimentalresults fiom literature. Those comparisons demonstrate the validity of the proposed numerical method, showing that it does not induces significant numerical errors, so that one canhave confidence in the results from those simulations.Investigations were also ~ade in order to analyze the influence some of the 5MB operating parameters on the overall equipment' s performances / Mestrado / Engenharia de Processos / Mestre em Engenharia Química

The numerical solution of elliptic boundary value problems on rectangular regions with Dirichlet boundary conditions is considered. The well-known finite difference scheme is used to discretize the continuous problem. The solution is now expressed as the unknown vector in a high order matrix equation. In general, efficient direct methods for obtaining the solution of the matrix equation are not known. There are several well-known iteration schemes commonly used to solve such problems. The main disadvantage of these methods is that the number of computations which are required to solve the matrix equation increases in a nonlinear way with the number of equations to be solved. Stone's original and symmetric strongly implicit factorization procedure are considered. The known results concerning the convergence properties of each iteration are presented. A new result concerning the symmetric factorization is presented and the results of numerical investigations are presented. / Science, Faculty of / Mathematics, Department of / Graduate

<p> The motion of energetic charged particles inside a crystalline solid is strongly dependent upon the orientation of the ion beam and target. This effect is commonly known as the "channeling" effect. In this report, the development of a computer code is presented which simulates the 3-D ion scatterings experienced by energetic particles moving in a crystalline solid. A Monte Carlo technique is incorporated in the code to calculate scattering angles, range distribution, backscattering distribution and angular distribution of incident ions. The Thomas-Fermi interatomic potential is used for binary collision process and the continuum potential is used for the potentials experienced by the channeled ions inside crystal lattices. </p> / Thesis / Master of Engineering (MEngr)

Orientador: Kamal Abdel Radi Ismail / Tese (doutorado) - Universidade Estadual de Campinas, Faculdade de Engenharia Mecânica / Made available in DSpace on 2018-08-19T02:00:06Z (GMT). No. of bitstreams: 1 Rosolen_CeliaVandaAlvesdeGodoy_D.pdf: 6132404 bytes, checksum: 6af57a0cb75e4c1476dd73ce697ed5a5 (MD5) Previous issue date: 2011 / Resumo: O presente trabalho tem como objetivos principais a elaboração de uma metodologia de cálculo de perfis e de cascatas em escoamento potencial baseado na distribuição de vórtices na superfície, o desenvolvimento de um código computacional próprio e a convalidação dos resultados obtidos. A metodologia baseia-se nos trabalhos de Martensen (1959), de Murugesan e Railly (1969) e de Lewis (1991). O comportamento aerodinâmico e os parâmetros característicos de um perfil isolado são totalmente conhecidos e existem várias técnicas analíticas e numéricas para avaliar e determinar o comportamento desses perfis. Essas técnicas analíticas podem ser estendidas para tratar os perfis experimentais como os perfis NACA com bons resultados. Neste trabalho o cálculo de perfis isolados compreende o uso do método de vórtices na superfície numa aproximação inversa. Inicialmente uma distribuição de velocidade é especificada a partir do ponto de estagnação do bordo de ataque. A distribuição de velocidade calculada sobre a superfície de um perfil inicial arbitrário é comparada com aquela requerida. A diferença entre elas é utilizada como sendo o valor de uma distribuição de vórtices que é imposta sobre o contorno. Devido ao campo desta distribuição de vórtices adicional, aparece uma distribuição de velocidades normais à superfície o que viola o conceito de linha de corrente. O contorno volta a ser uma linha de corrente quando sua inclinação local é ajustada de acordo com a razão entre os valores das velocidades normal e requerida. O perfil obtido é submetido a uma nova análise direta e a distribuição de velocidade obtida é comparada com a distribuição requerida. Os cálculos são repetidos até alcançar a convergência. O algoritmo numérico resultante foi aplicado a dois tipos de perfis, analítico (Joukowski) e experimental (NACA), para os casos de aerofólios simétrico e assimétrico. Os perfis NACA utilizados são NACA 65-010 e NACA 65-1210. Em todos os casos os perfis foram gerados usando a metodologia de distribuição de vórtices e os resultados foram comparados com os respectivos perfis originais, mostrando boa concordância. Tendo verificado que a técnica proposta e o código elaborado são válidos e produzem bons resultados, eles são aplicados à análise de cascatas de aerofólios NACA 65-1210 e de uma cascata de aerofólios de Gostelow / Abstract: This work presents a method to calculate airfoils and cascades in potential flow based upon the distribution of vortices on the profile surface, develop a computational code and validate the predictions with available well accepted results. The technique is based upon Martensen (1959), Murugesan and Railly (1969) and Lewis (1991). Cascades are usually specified according to the application and they can be stationary or moving. The aerodynamic behavior and the aerodynamic characteristics of isolated foil are known and many techniques to calculate these foils are available. These analytic techniques can be extended to handle nicely experimental foils such as the NACA's. In the present work, the method of distributing the vortices over the surface of the foil is used in the reverse sense. Initially a velocity profile is specified starting from the leading edge stagnation point. It is calculated the velocity distribution over the surface of an initial arbitrary foil and it is compared with that required. The difference between them is used as an additional distribution of vortices imposed over the surface. A distribution results in a velocity component normal to the surface and this violates the concept of the stream line. The contour turns to be a streamline when its local inclination is adopted according to the ratio of the normal velocity and the required velocity. The resultant foil is reanalyzed and the velocity distribution obtained is compared with the required distribution. The calculations are repeated until convergence is achieved. The numerical code was applied to two types of foils, analytical (Joukowski) and experimental (NACA) for the cases of symmetrical and asymmetrical foils. NACA foils used are the NACA 65-010 and NACA 65-1210. In all the cases the foils were generated using the vortex distribution technique and the resulting profiles were compared with the original profiles indicating a good agreement. These comparisons validate the proposed technique and the developed code. They were applied to analyse cascades of NACA 65-1210 foils and a cascade of the Gostelow foils / Doutorado / Termica e Fluidos / Doutor em Engenharia Mecânica

In this paper numerical flow calculations with respect to the annular gaps with added radial grooves normaly used on the spools of directional control valves were carried out. The impact of various annular gap geometries and radial grooves during variable pressure conditions, and while using different hydraulic fluids, on the flows through annular gaps were investigated for different flow regimes. Samples with different geometries and numbers of radial groves on the spool of the directional control valve were also made for the purpose of carrying out flow measurements. The two different hydraulic fluids that were used in the numerical simulations and for the flow measurements were a hydraulic mineral oil and tap water. The results of the numerical calculations for the different models of the radial grooves with axially symmetric geometries show their impact on the internal leakage with respect to three different regimes of flow. The results of the numerical calculations based on the use of a hydraulic oil show a trend that was established by the experimental investigation.

This master thesis was provided by Scania AB. The objective of this thesis was to modify an application in the free Computational Fluid Dynamics software OpenFOAM to be able to handle spray and wall film modeling of a Urea Water Solution together with Conjugate Heat Transfer. The basic purpose is to widen the knowledge of the vaporization process of a Urea Water Solution in the exhaust gas after treatment system for a diesel engine by using OpenFOAM. First, urea has been modeled as a very viscous liquid at low temperature to mimic the solidication process of urea. Second, the development of the new application has been done. At last, test simulations of a simple test case are performed with the new application. The results are then compared with simplied hand calculations to verify a correct behavior of certain exposed source terms. The new application is working properly for the test case but to ensure the reliability, the results need to be compared with another Computational Fluid Dynamics software or more preferable, real experiments. For more advanced geometries, the continued development presented last in this thesis is highly recommended to follow.

Cette thèse présente une étude théorique de la structure électronique et de la tendance à l'ordre dans les nanoalliages de métaux de transition en se référant au cas des systèmes à base de cobalt-platine (CoPt) qui présentent un intérêt particulier dans les domaines du magnétisme et de la catalyse. Il est ainsi important de décrire au mieux l'évolution de la structure électronique en lien avec la structure atomique ou l'arrangement chimique en fonction de la taille de tels nanoalliages en vue de prédire des propriétés potentielles pouvant différer fortement de celles du matériau massif correspondant. Dans ce contexte, des calculs systématiques de DFT ont été mis en œuvre sur des systèmes modèles simples, alliages massifs, surfaces et nanoparticules, qui ont permis de montrer qu'une règle de neutralité de charge locale, par site espèce et orbitale, s'applique au système CoPt massif et s'étend aux nanoalliages. Sur cette base, des calculs auto cohérents de liaisons fortes ont été développés et ont permis de proposer un moyen précis de prédire les caractéristiques de nanoalliages réalistes, en termes de redistribution des états électroniques et de tendance à l'ordre. Les grandeurs caractéristiques déduites de ces calculs, telles que le désordre diagonal et non diagonal, peuvent être en effet être déterminées à partir de lois simples linéaires de variations des centres et des largeurs de bandes sur les sites Co et Pt. Les valeurs issues de ces lois peuvent être placées sur des cartographies de domaines de tendance à l'ordre et l'ensemble de la méthodologie devrait être étendue facilement à d'autres systèmes binaires. / The interest in alloys of late transition metals arises from their potential applications in high-density magnetic storage devices where they can be used as supported magnetic nanoparticle arrays and as stable, efficient and selective catalysts. The preparation of materials with optimal properties faces a number of technological and physical restrictions and requires an in-depth knowledge of the interplay between structural features on the atomic level and the desired macroscopic properties. In the thesis, after extensive discussion of Density Functional Theory and Tight Binding approaches the work focused on DFT calculations of bulk systems, surfaces and small clusters. The results allow to conclude on general validity of the method and especially to justify the local neutrality assumption in the case of low-coordinated sites in nanoparticles. Basic structural, magnetic and energetic properties were also studied and compared with the experimental data. Subsequently TB calculations were performed and verified with DFT results. The scope of the calculations was then extended for the case of nanoclusters of realistic sizes, unavailable in DFT. Local Densities of State on sites with different chemical environment and coordination numbers were analyzed. The observations prove that basic features of LDOS (d-band centre and d-band width) can be predicted by simple laws on the basis of two terms: a structural term represented by the linear function of the site coordination and a chemical term as a rigid shift which opens a new way to predict the ordering tendency (mapping of the ordering domains) for any transition metal nanoalloy as a function of its size.