Spelling suggestions: "subject:"boissons sampling"" "subject:"boissonade sampling""
1 |
Development of general finite differences for complex geometries using immersed boundary methodVasyliv, Yaroslav V. 07 January 2016 (has links)
In meshfree methods, partial differential equations are solved on an unstructured cloud of points distributed throughout the computational domain. In collocated meshfree methods, the differential operators are directly approximated at each grid point based on a local cloud of neighboring points. The set of neighboring nodes used to construct the local approximation is determined using a variable search radius. The variable search radius establishes an implicit nodal connectivity and hence a mesh is not required. As a result, meshfree methods have the potential flexibility to handle problem sets where the computational grid may undergo large deformations as well as where the grid may need to undergo adaptive refinement. In this work we develop the sharp interface formulation of the immersed boundary method for collocated meshfree approximations. We use the framework to implement three meshfree methods: General Finite Differences (GFD), Smoothed Particle Hydrodynamics (SPH), and Moving Least Squares (MLS). We evaluate the numerical accuracy and convergence rate of these methods by solving the 2D Poisson equation. We demonstrate that GFD is computationally more efficient than MLS and show that its accuracy is superior to a popular corrected form of SPH and comparable to MLS. We then use GFD to solve several canonic steady state fluid flow problems on meshfree grids generated using uniform and variable radii Poisson disk algorithm.
|
2 |
Optimisation de la chaîne de numérisation 3D : de la surface au maillage semi-régulier / 3D digitization optimization : from surface to semi-regular meshPeyrot, Jean-Luc 12 December 2014 (has links)
La numérisation 3D permet de générer des représentations numériques très réalistes et fidèles géométriquement aux surfaces réelles. Cependant, cette fidélité géométrique, obtenue à l'aide d'un sur-échantillonnage de surfaces, augmente considérablement le volume de données générées. Les maillages ainsi produits sont donc très denses, et peu adaptés aux différents supports de visualisation, de transfert, de stockage, etc. La représentation semi-régulière des surfaces permet de diminuer le volume nécessaire à la représentation de ces maillages denses, et possède des qualités bien connues en matière de représentations multi-échelles et de compression. Cette thèse a pour objectif d'optimiser la chaîne de numérisation 3D classique en améliorant l'échantillonnage des surfaces tout en conservant la fidélité géométrique, et en court-circuitant les étapes fastidieuses qui conduisent à une représentation semi-régulière. Pour cela, nous avons intégré dans un système stéréoscopique, l'échantillonnage en disques de Poisson qui, grâce à ses propriétés de bruit bleu, réalise un bon compromis entre sous- et sur-échantillonnage. Ensuite, nous avons généré un mailleur semi-régulier, qui travaille directement sur les images stéréoscopiques, et non sur une version remaillée des nuages de points habituellement générés par ces systèmes. Les résultats expérimentaux montrent que ces deux contributions génèrent de façon efficace des représentations semi-régulières, qui sont géométriquement fidèles aux surfaces réelles, tout en réduisant le volume de données générées. / Nowadays, 3D digitization systems generate numeric representations that are both realistic and of high geometric accuracy with respect to real surfaces. However, this geometric accuracy, obtained by oversampling surfaces, increases significantly the generated amount of data. Consequently, the resulting meshes are very dense, and not suitable to be visualized, transmitted or even stored efficiently. Nevertheless, the semi-regular representation due to its scalable and compact representation, overcomes this problem. This thesis aims at optimizing the classic 3D digitization chain, by first improving the sampling of surfaces while preserving geometric features, and secondly shortening the number of required treatments to obtain such semi-regular meshes. To achieve this goal, we integrated in a stereoscopic system the Poisson-disk sampling that realizes a good tradeoff between undersampling and oversampling, thanks to its blue noise properties. Then, we produced a semi-regular meshing technique that directly works on the stereoscopic images, and not on a meshed version of point clouds, which are usually generated by such 3D scanners. Experimental results prove that our contributions efficiently generate semi-regular representations, which are accurate with respect to real surfaces, while reducing the generated amount of data.
|
Page generated in 0.0792 seconds