Determinação de métricas para reprodução da percepção visual por simulações computacionais

Abstract

Daylight is often mentioned as an important factor to be considered in architectural projects. However, it is mostly taken into consideration as regards its physical features (sun path and availability) whose metrics do not necessarily reflect subjective preferences of the occupants, such as the daylight factor. Daylight affects not only energy efficiency, but also human physiology and time setting and, above all, the way one perceives the environment. In this context, this research aims to establish the relationship between daylight, access to views and human perception throughout three seasons – winter, spring and summer – in a climate chamber in Karlsruhe, Germany. Its goal is to determine which metrics better translate visual perception of the chamber occupants and which could also be reproduced in computer simulations and later used as design parameters. For this purpose, a climate chamber, built on top of a rotating platform, was used to test the effect of façade orientation on several subjective variables of volunteers under experimental conditions. This is a longitudinal field study with a quantitative approach that uses instruments such as surveys and measurements, as well as computer simulations as technique. Preliminary computer simulations helped to identify façade orientations with the most differentiation regarding daylight access. After test orientations were validated against field data from a pilot study carried out during a concurrent field study, data collected during such study including survey results as regards light perception and measurements of thermal and lighting conditions, indoors and outdoors were used for analysis. Subjective and objective data were then compared using Pearson´s R coefficient (for grouped data) and Spearman´s rank correlation coefficient (for individual data). Correlation results point to the use of vertical and horizontal illuminance as metrics with the strongest relationship to participants’ light perception. Correlations between measured data and computer simulations confirm the predominance of vertical illuminance, as well as Daylight Availability and the Useful Daylight Illuminance as valid tools to reproduce the illuminance at the eye level.