Dissertations / Theses on the topic 'Fortran Language'

Orientadores: Rubens Maciel Filho, Liliane Maria Ferrareso Lona<br>Tese (doutorado) - Universidade Estadual de Campinas, Faculdade de Engenharia Quimica<br>Made available in DSpace on 2018-08-04T02:31:19Z (GMT). No. of bitstreams: 1 MendozaMarin_FlorentinoLazaro_D.pdf: 8929110 bytes, checksum: 9041b9e7f02f9a324fa10cde9a962f54 (MD5) Previous issue date: 2004<br>Resumo: O modelo determinístico e processo homopolimerização na emulsão do estireno são aplicados em reator tubular contínuo sem e com deflectores angulares internos sob condição isotérmica e não isotérmica. Os resultados de modelagem e simulação foram realizados a estado estacionário, modelo unidimensional, coordenada cilíndrica, fluxo pistão laminar completamente desenvolvido, modelo Smith-Ewart para estimar a conversão do monômero, cinética química de Arrhenius corno modelo de velocidade finita laminar para computar a geração química. O objetivo é modelar, simular e analisar o comportamento do reator de homopolimerização na emulsão do estireno com deflectores angulares inclinados internos, e comparar com reator tubular. Os métodos experimental e matemático-dedutivo foram aplicados para obter resultados, por meio de programação computacional, usando Dinâmica de Fluido Computacional através do método de volumes finitos. As seguintes variáveis como temperatura de reação constante e variável, reator tubular sem e com deflectores, temperatura de alimentação, diâmetro de reator, processo adiabático e exotérmico, calor de reação constante e velocidade axial completamente desenvolvida foram investigados. Os efeitos de conversão de monômero, área transversal interna, temperatura axial, concentração do polímero, radicais e iniciador, outros corno densidade de polímero e monômero, perda de carga e queda de pressão foram determinados e simulados. Os produtos foram caracterizados com Número de Partículas (nucleação homogênea e heterogênea), distribuição de peso molecular, tamanho de partículas de polímero e distribuição de viscosidade. Estes resultados foram validados com resultados da literatura sob condição igualou aproximada. Os resultados sob condições não isotérmicas foram melhores que os resultados isotérmicos em termos de caracterização do polímero. Isso mostra que o desenho alternativo proposto (com deflectores) permite obter o polímero com propriedades melhores em termos de número de partículas, distribuição de peso molecular, distribuição do tamanho de partículas e viscosidade<br>Abstract: Deterministic model and emulsion homopolymerization process of styrene are applied in continuous tubular reactor without and with internal angular baffles under isothermic and no isothermic conditions. The modeling and simulation results were approximate to steady state, one-dimensional model, cylindrical coordinate, fully developed laminar plug flow, Smith-Ewart model to estimate the monomer conversion, Arrhenius chemical kinetics as laminar finite-rate model to compute chemical source. The objective is to model, simulate and to analyze the emulsion homopolymerization reactor performance of styrene with internal-inc1ined angular baffles, and to compare with continuous tubular reactor. The experimental and mathematical-deductive methods were applied to obtain results, by means of computational programming, using Computational Fluid Dynamics (program code), finite volume method. The following variables such as constant and variable reaction temperature, tubular reactor without and with baffles, feed temperature, reactor diameter, adiabatic and exothermic process, constant reaction heat and fully developed axial velocity were investigated. The monomer conversion, internal transversal are a, axial temperature, concentration of polymer, radicals and initiator, others as density of polymer and monomer, head loss and pressure drop effects were determined and simulated. The products were characterized by partic1es number (homogeneous and heterogeneous nuc1eation), molecular weight distribution, polymer partic1es size and polymer viscosity distribution. These results were validated with literature results under same or approximate condition. The results under no isothermic conditions were better than isothermic results in terms of polymer characterization. It is shown that the proposed alternative design (with baffles) allow to obtain the polymer with better properties in terms of number of partic1es, molecular weight distribution, particle size distribution and viscosity<br>Doutorado<br>Desenvolvimento de Processos Químicos<br>Mestre em Engenharia Química

Dans le domaine du calcul de haute performance, les architectures d’exécution sont en continuelle évolution. L’augmentation du nombre de nœuds de calcul, ou le choix d’une topologie réseau plus rapide représentent un investissement important tant en temps qu’en moyen financier. Les méthodes de prédiction de performances permettent de guider ce choix. En parallèle à ce développement, les systèmes HPC pair-à-pair (P2P) se sont également développés ces dernières années. Ce type d’architecture hétérogène permettrait la résolution des problèmes scientifiques pour un coût très faible par rapport au coût d’une architecture dédiée.Ce manuscrit présente une méthode nouvelle de prédiction de performances pour les applications réelles de calcul distribué, exécutées dans des conditions réelles. La prédiction prend en compte l’optimisation du compilateur. Les résultats sont extrapolables et ils sont obtenus pour un ralentissement réduit. Ce travail de recherche est implémenté dans un logiciel nouveau nommé dPerf. dPerf est capable de prédire les performances des applications C, C++ ou Fortran qui communiquent en utilisant les normes MPI ou P2P-SAP et qui s’exécutent sur une architecture cible pair à pair, hétérogène et décentralisée. La précision de cette contribution a été étudiée sur (i) la transformée Laplace, pour l’aspect séquentiel, (ii) le benchmark IS de NAS, pour l’aspect MPI, (iii) et le code de l’obstacle pour l’aspect calcul P2P décentralisé et l’extrapolation du nombre de nœuds<br>In the field of high performance computing, the architectures evolve continuously. In order to increase the number of computing nodes or the network speed, an important investment must be considered, from both temporal and financial point of view. Performance prediction methods aim at assisting in finding the best trade-off for such an investment. At the same time, P2P HPC systems have known an increase in development. These heterogeneous architectures would allow solving scientific problems at a low cost, with respect to dedicated systems.The manuscript presents a new method for performance prediction. This method applies to real applications for distributed computing, considered in a real execution environment. This method uses information about the different compiler optimization levels. The prediction results are obtained with reduced slowdown and are scalable. This thesis took shape in the development of the dPerf tool. dPerf predicts the performances of C, C++, and Fortran application, which use MPI or P2P-SAP to communicate. The applications modeled by dPerf are meant for execution on P2P heterogeneous architectures, with a decentralized communication topology. The accuracy of dPerf has been studied on three applications: (i) the Laplace transform, for sequential codes, (ii) the NAS Integer Sort benchmark for distributed MPI programs, (iii) and the obstacle problem, for the decentralized P2P computing and the scaling of the number of computing nodes

Esta tese, organizada em três partes, objetiva recuperar o Ensaio de Semântica de Michel Bréal evidenciando a importância de suas idéias para a Lingüística Moderna, para a Semântica e para a Pragmática, e a aplicabilidade da sua teoria na descrição da Língua Portuguesa, em sua modalidade escrita. Na primeira parte deste trabalho, apresenta-se o teor do ensaio, conforme a terceira edição publicada em 1904. Para tanto, parte-se de uma interpretação que relaciona as propostas brealinas àquelas elaboradas pelas disciplinas lingüísticas mais vulgarizadas no Brasil, de modo a comprovar a atualidade e a importância das propostas do ES. Em seguida, a reflexão volta-se para as questões do significado, a fim de demonstrar como semanticistas e pragmatistas têm entendido e interpretado as propostas do fundador da semântica. Esta análise, apresentada na segunda parte da tese, evidencia, de um lado, a importância de suas idéias para os estudos da significação e, de outro, a ausência de trabalhos nos quais haja aplicação prática de seus postulados. Ilustrando como suas propostas podem contribuir para a descrição de determinados fenômenos semânticos da Língua Portuguesa, na terceira parte da tese, sua teoria é empregada na análise de dois textos jornalísticos: um escrito em português brasileiro e outro em português europeu.<br>This thesis tries to revive Michel Bréal´s Essay of Semantics supporting the importance of it´s ideas for Modern Linguistics in general, and, specially, for Semantics and Pragmatics and the applicability of Bréal\'s postulates to language description. Due to that, contributions for semantics sciences and it´s applicability for written Portuguese language description are shown. The thesis is organized in three parts. In the first, the content of the third edition of the essay (1904) is connected to best known branches of Modern Linguistics in Brazil, connections that serves as evidence for the importance of Bréal`s ideas. In the second part - which deals especifically whit semantic´s issues - there is a critical view of semantician´s and pragmatician´s reviews of the essay. This part reveals the relevance of Breal´s postulates for Semantics (Pragmatics included), but, on the other hand, points out that his ideias have hardly been applied. In the last part, as evidence that to apply his theories is not only possible but desirable, they are used in the analysis of two journalistic papers: one in Brazilian Portuguese and other in European Portuguese.

Given the continuous usage and spread of computed tomography (CT), the potential harmful e↵ects and the radiation dose to the patient have become high interest topics among the scientific community. The main objective of this investigation was to modify existing three-dimensional (3D) voxel phantom models to resemble real patients as much as possible, trying to progress the concept of a more personalized patient dosimetry. This work focused essentially in one of the biggest and most radiosensitive organs in the thorax, the lungs. Additionally, the variations of organ doses when a standard phantom is used instead were studied. During the course of this work a FORTRAN-based program was developed, which is able to semi-automatically modify the volumetric information of organs of interest in a standard voxel phantom (Female ICRP Adult Reference). The voxel resolution was also altered so the phantom’s diameters match the patient’s ones. Monte Carlo (MC) PENELOPE simulation code was used to mimic CT scan conditions and, therefore, generate 2D projections, used for visual organ matching with clinical patient CT images, and access organ dose in both phantoms (ICRP standard and ICRP modified). The main results reported that matching the voxel phantom’s size and lungs provides organ dose values significantly di↵erent from the ones measured in the ICRP reference phantom. Voxel models matched to patients’ size and overall anatomy allow increased accuracy in organ dose estimation, which, as reported by this study, can su↵er from up to 20% underestimation and 40% overestimation. This study demonstrates that voxel phantoms developed using single patient data provide a better and more precise organ dose assessment by MC methods than a standard phantom. The presented methodology should be of interest for dose optimization studies and quick enough for routine clinical use.