I propose a particle-based technique for simulating incompressible uid that includes adaptive re nement of particle sampling. Each particle represents a mass of uid in its local region. Particles are split into several particles for ner sampling in regions of complex ow. In regions of smooth ow, neghboring particles can be merged. Depth below the surface and Reynolds number are exploited as our criteria for determining whether splitting or merging should take place. For the uid dynamics calculations, I use the hybrid FLIP method, which is computationally simple and e cient. Since the uid is incompressible, each particle has a volume proportional to its mass. A kernel function, whose e ective range is based on this volume, is used for transferring and updating the particle's physical properties such as mass and velocity. In addition, the particle sampling technique is extended to a fully adaptive approach, supporting adaptive splitting and merging of uid particles and adaptive spatial sampling for the reconstruction of the velocity and pressure elds. Particle splitting allows a detailed sampling of uid momentum in regions of complex ow. Particle merging, in regions of smooth ow, reduces memory and computational overhead. An octree structure is used to compute inter-particle interactions and to compute the pressure eld. The octree supporting eld-based calculations is adapted to provide a ne spatial reconstruction where particles are small and a coarse reconstruction where particles are large. This scheme places computational resources where they are most needed, to handle both ow and surface complexity. Thus, incompressibility can be enforced even in very small, but highly turbulent areas. Simultaneously, the level of detail is very high in these areas, allowing the direct support of tiny splashes and small-scale surface tension e ects. This produces a nely detailed and realistic representation of surface motion.

The Hubble Space Telescope has provided the world with incredible imagery of the surrounding universe. The aesthetic quality of this imagery is limited by production resources; by creating a method to harness the highly refined detail and quality of CG elements in live-action films, we can inspire and educate at a much greater level. In this thesis, I create a rendering approach that allows camera movement around and through elements such as nebulae and galaxies, creating a more cinematic experience. The solution will also allow for reasonable scientific accuracy, visual appeal, efficiency, and extendability to other astronomical visualizations. 3D meshes are constructed and textured using telescopic images as reference. Splats are volumetrically generated using a voxelized bounding box around the mesh. Valid splats within a user specified maximum distance receive initial color and alpha values from the texture map. Probability density functions are used to create a density falloff along the edges of the object, and modifications to the RGBA values are made to achieve the desired cloud-like appearance. The data sets are rendered using a C program developed at the Space Telescope Science Institute by Dr. Frank Summers. The methodology is applied to the test cases of a nebula, star-forming region Sharpless 2-106, and a galaxy, Messier 51, or the Whirlpool Galaxy. The results of this thesis demonstrate the visual, scientific, and technical success of this solution. The code developed during this project generates the desired imagery with reasonable efficiency. A short animation moving from outside the galaxy to a close up of the nebula exhibits the flexibility in scale and camera movement. A careful balance between scientific accuracy and visual appeal were maintained through consultation with astronomers at the Space Telescope Science Institute. The favorable efficient, flexible, visual, and scientific results presented by this work make this process extendable to most other cases of nebula and galaxy visualizations.

Conventional analysis of atmospheric data includes three-dimensional desktop-computer displays. One disadvantage is that it can reduce the ability to zoom in and see small-scale features while concurrently viewing other faraway features. This research intends to determine if using virtual environments to examine atmospheric data can improve a meteorologist's ability to analyze the given information. In addition to possibly enhancing small-scale analysis, virtual environments technology offers an array of possible improvements. Presented is the theory on developing an experiment to establish the extent to which virtual environments assist meteorologists in analysis. Following is the details of an implementation of such an experiment. Based on the quantitative results obtained, the conclusion is that immersion can significantly increase the accuracy of a meteorologist's analysis of an atmospheric data set.

No description available.

O aumento do poder computacional e conseqüente desenvolvimento das técnicas de simulação numérica, aliados ao avanço tecnológico dos periféricos de medição, fizeram com que muitas áreas de pesquisa, passassem a necessitar de ferramentas gráficas e de auxílio computacional para apoiar o processo de interpretação das informações geradas. A aplicação de técnicas gráficas para ampliar a capacidade de interpretação de dados científicos tem sido denominada visualização em computação científica (ViSC - Visualization in Scientific Computing). Modelos aerodinâmicos construídos no laboratório, foram simulados em software numérico de método dos painéis e submetidos a rotinas desenvolvidas em C++ as quais serviram de superfície para uma ferramenta do tipo biblioteca, de baixo custo, muito utilizada em universidades do mundo todo, chama VTK (Visualization Tool Kit), que possui elementos gráficos para a geração de visualizações de qualquer tipo de dados. Esses códigos em C++ são responsáveis pelos tipos de visualização gerados e principalmente por permitir o uso da ferramenta. As visualizações de distribuição de pressão e isolinhas nas superfícies dos modelos são de suma importância na identificação de problemas aerodinâmicos possibilitando correções e modificações antes mesmo de o modelo ser construído. / The increase of computational power and the technical development of numerical are responsible for the creation of many new areas that use graphics tools and computational aid for the interpretation of the information generated. The application of graphics techniques to increase the capability of scientific interpretation is called ViSC or Visualization in Scientific Computation. Laboratory built freeflight and wind tunnel models were calculated using numerical software and submitted to routines developed in C++ language which produced various types of visualization while using VTK. Visualization of pressure distributions and streamlines on the model surfaces are important for identification of aerodynamic problems and making corrections and modifications possible before the construction of the physical model.

As HPC systems move towards exascale, the discrepancy between computational power and I/O transfer rate is only growing larger. Lossy in situ compression is a promising solution to address this gap, since it alleviates I/O constraints while still enabling traditional post hoc analysis. This dissertation explores the viability of such a solution with respect to a specific kind of compressor — wavelets. We especially examine three aspects of concern regarding the viability of wavelets: 1) information loss after compression, 2) its capability to fit within in situ constraints, and 3) the compressor’s capability to adapt to HPC architectural changes. Findings from this dissertation inform in situ use of wavelet compressors on HPC systems, demonstrate its viabilities, and argue that its viability will only increase as exascale computing becomes a reality.

<p>It is important for the military to know as much as possible about how easily detected their vehicles are. One way among many used to detect vehicles is the use of radar sensors. The radar reflecting characteristics of military vehicles are therefor often rigorously tested. With measurements and simulations it is possible to calculate likely detection distances to a vehicle from different angles. This process often produces very large data sets that are hard to analyze.</p><p>This thesis discusses and implements a method for visualizing the detection distance data set and also discusses a lot of related issues with a focus on computer graphics.</p><p>The main concept is called spherical displacement and the idea is to visualize the detection distances as a surface with the imagined vehicle in the center point. Detection is likely inside the surface but not on the outside. This concept is the next step from the colored sphere where the colors represent the detection distance which was previously used.</p><p>The thesis project resulted in a visualization tool that uses the new concept and can handle large data sets. The spherical displacement concept is more intuitive and shows detail better than the colored sphere visualization.</p>

It is important for the military to know as much as possible about how easily detected their vehicles are. One way among many used to detect vehicles is the use of radar sensors. The radar reflecting characteristics of military vehicles are therefor often rigorously tested. With measurements and simulations it is possible to calculate likely detection distances to a vehicle from different angles. This process often produces very large data sets that are hard to analyze. This thesis discusses and implements a method for visualizing the detection distance data set and also discusses a lot of related issues with a focus on computer graphics. The main concept is called spherical displacement and the idea is to visualize the detection distances as a surface with the imagined vehicle in the center point. Detection is likely inside the surface but not on the outside. This concept is the next step from the colored sphere where the colors represent the detection distance which was previously used. The thesis project resulted in a visualization tool that uses the new concept and can handle large data sets. The spherical displacement concept is more intuitive and shows detail better than the colored sphere visualization.

[pt] A visualização de horizontes sísmicos constitui uma importante área de conhecimento amplamente aplicada na prospecção de hidrocarbonetos pela indústria do petróleo. Diferentes técnicas são atualmente empregadas na apresentação destas superfícies, sendo usualmente utilizadas as soluções baseadas na geração de malhas poligonais, que se beneficiam da otimização das placas gráficas atuais no desenho de triiângulos. Este trabalho faz uma avaliação do uso da renderização baseada em pontos, no lugar de polígonos, para a visualização de horizontes sísmicos. Para isso as técnicas de cada etapa do processo são avaliadas, levando-se em conta a natureza específica dos dados de interpretação de horizontes em volumes sísmicos e o resultado final esperado para a visualização deste tipo de dados. O algoritmo utilizado baseia-se no método conhecido como Surface Splatting para a renderização dos pontos originais, sendo estudados a estruturação apropriada para os dados a serem visualizados, a técnica para obtenção de normais, a abordagem adequada para o cálculo da iluminação e mecanismos adicionais necessários ao processo. Resultados da aplicação do método em dados reais são, ao final do trabalho, analisados e comparados à renderização tradicional para os horizontes avaliados. / [en] Seismic horizon visualization stands as an important knowledge area used to support exploration on the oil industry. Different techniques currently employed to render this kind of surfaces are usually based on polygonal meshes generation, which benefits from graphics boards optimization on drawing triangles. This work is an evaluation of Point Based rendering techniques to replace polygonal approaches in seismic horizons visualization. To do so, this study revisits each stage of the seismic visualization process. The algorithm adopted here is based on the Surface Splatting with the EWA filter. This work also presents a study on normal evaluation and data structures to store points and normal. Special care is taken in shading techniques. The implementation yielded results that are used to support the evaluation of the Point Based Techniques on real 3D Seismic data. Traditional triangle based rendering is also presented to compare results.

La simulation informatique est un outil incontournable dans les sciences expérimentales. La puissance de calcul croissante des ordinateurs associée au parallélisme et aux avancées dans la modélisation mathématique des phénomènes physiques permet de réaliser virtuellement des expériences de plus en plus complexes. De plus, l'émergence de la programmation GPU a considérablement accru la qualité et la rapidité de l'affichage. Ceci a permis de démocratiser la visualisation sous forme graphique des résultats de simulation. La visualisation scientifique peut être passive : l'utilisateur peut suivre l'évolution de la simulation ou bien observer ses résultats après que le calcul soit terminé. Elle peut aussi être interactive lorsque le chercheur peut agir sur la simulation alors qu'elle se déroule. Créer de telles applications complexes n'est cependant pas à la portée de tout scientifique non informaticien. La programmation par composants est, depuis des années, mise en avant comme une solution à ce problème. Elle consiste à construire des applications en interconnectant des programmes exécutant des tâches élémentaires. Ce mémoire présente un modèle de composants et une méthode de composition d'applications de visualisation scientifique interactive. Elle s'intéresse, en particulier, à la conciliation de deux contraintes majeures dans la coordination de ces applications : la performance et la cohérence. / Computer simulation is an essential tool in experimental sciences. The increasing computing power, parallelism and the advances in the mathematical modeling of physical phenomena allow to virtually run always more complex experiments. In addition, the rise of GPU programming has greatly increased the quality and performance of display. This has allowed to spread the graphical visualization of simulation results. Scientific visualization can be passive: the user can only follow the simulation's progress or observe its results when it is done. It can also be interactive in which case the researcher can act on the simulation while it is running. Creating such complex applications can, however, be tedious for non-computer-scientists. Component-based development is, for years, highlighted as a solution to this problem. It consists in building applications by interconnecting small programs completing elementary tasks. This thesis presents a component model and a method for composing interactive scientific visualization applications. It particularly focuses on the balance between two major constraints of these applications: performance and coherence.