A malha de pontos de simulação e o cálculo das métricas de avaliação do desempenho da luz natural: estudo e proposta de distribuições não uniformes

Abstract

The study of the behavior of daylight in the built environment can be done using several tools, such as daylight performance metrics, which allow to verify minimum performance requirements and the probability of visual discomfort. Its calculation presupposes the definition of parameters, such as the points grid in the work plane, which are commonly evenly distributed in the work plane. This thesis aims to identify the relationship between the arrangement of the mesh of natural lighting calculation points and performance metrics, as well as to propose a methodology for the definition of meshes of non-uniform densities. The study was divided into three stages aiming at identifying the influence of the quantity of points, the sectorization of the work plan and parameters such as model orientation, latitude and obstruction of the surroundings in order to define the non-uniform density grid. For all phases, a standard environment model with lateral opening was used. In the first phase the 2x2 to 15x15 point grid was varied in the work plane, in the second the plane was divided into nine sectors of the same area, where the grid was varied from 1x1 to 8x8 within the sector and in the third phase two orientations, four latitudes and four situations of surrounding obstruction were evaluated. The computational simulation was performed by TropLux in version 7 for the first stage and 8 for the others. The exploratory and statistical analyses used the tools of ANOVA, Tukey's test. The results allowed the definition of the proposed methodology, which is the division into two grids, one for the calculation of the direct illuminance and the other for the reflected one. For the sectorization of the work plan, a reference value of 1.0 sect/m2 was identified for the direct one and 0.4 sect/m2 for the reflected one. For the definition of the density of points in the sector for the direct component it was used the coefficient of variation of the Factor of Sky Vision, whose divisor was 0.6, where sectors below the density of points is 1.0 pt/m2 and sectors above the density is 9.0 pt/m2. For the mesh of the reflected component the parameter was the angular distance to the window, where a dividing value of 50° was identified, where below it the density of points is 0.4 pt/m2 and sectors above it the density is 7.0 pt/m2. The application of the methodology pointed out that there is a gain in processing time without significant loss in the metric calculations with the application of the non-uniform grids. The better definition of the grid has as a consequence greater precision in the results of the metrics and provides architecture and engineering professionals with tools that allow a better representation of the performance of their projects, allowing a better definition of strategies for the use of daylight.