Publication: ANÁLISIS DE CÓDIGO EN GPU DE CRECIMIENTO DE GRANOS
No Thumbnail Available
Date
2017
Authors
MUÑOZ PEÑALOZA, CESAR ALEXIS
Journal Title
Journal ISSN
Volume Title
Publisher
Abstract
El crecimiento de granos corresponde ser el proceso que experimentan losmateriales policristalinos cuando son sometidos a temperaturas elevadas, provocando un aumento o disminución de los granos que lo conforman. En esta memoria se utilizará el modelo y simulación realizado por [19] y que posteriormente ha sido modificado y paralelizado utilizando la GPU [11].En este trabajo, se comienza analizando el rendimiento del código desarrollado en [11] en busca de identificar aspectos a mejorar. Luego, se procede a realiza un análisis de convergencia de la estructura de los granos con el objetivo de determinar un valor ideal para el parámetro delta t y así tener una buena simulación númerica para la realización del análisis estadístico sobre la estructura de granos. Esto permite la obtención de estadísticas estables sobre la estructura de granos que se trabaja.
The grains growth corresponds to the process experienced by the polycrystallinematerials when subjected to high temperatures, causing an increase or decreasein the grains that make it up. In this memory it will be used the model and simulation carried out by [19] and that has subsequently been modified and parallelised using the GPU [11].In this job, we start by analyzing the performance of the code developed in [11] in search of identifying aspects to improve. Then, we proceed to an analysis of convergence of the structure of the grains in order to determine a value ideal for the parameter delta t and so have a good numerical simulation for the carring out of the statistical analysis about the structure of grains. This allows to obtain Stable statistics about the grain structure that is worked.
The grains growth corresponds to the process experienced by the polycrystallinematerials when subjected to high temperatures, causing an increase or decreasein the grains that make it up. In this memory it will be used the model and simulation carried out by [19] and that has subsequently been modified and parallelised using the GPU [11].In this job, we start by analyzing the performance of the code developed in [11] in search of identifying aspects to improve. Then, we proceed to an analysis of convergence of the structure of the grains in order to determine a value ideal for the parameter delta t and so have a good numerical simulation for the carring out of the statistical analysis about the structure of grains. This allows to obtain Stable statistics about the grain structure that is worked.
Description
Catalogado desde la version PDF de la tesis.
Keywords
CRECIMIENTO DE GRANOS , DISTANCIA DE HAUSDORFF , ESTADISTICA , GPU