Predicting Reality within the Virtual: Experimental Validation of Daylight Simulation Tools For Architectural Spaces in The Tropics

Abstract

This study aims to develop a validation method to identify the predictive capacity of two daylight calculation engines applied to real-life scenarios within the Tropics. The main objective is to determine the accuracy of the calculations generated by simulation tools in correspondence to measured illuminance values, defining which conditions would render the best results. To achieve this, a field survey with photometric measurements of an indoor working space under tropical sky conditions was compared with simulation results of two software tools: VELUX Daylight Visualizer and DIVA for Rhino. The results were validated using dynamic daylight metrics and statistical analysis. Overall, the results showed a consistent overestimation of the simulated Daylight Autonomy (DA) values in the range of 9% to 36% in the afternoon hours. Useful Daylight Illuminance (UDI) values presented less deviation ranging from 7% to 9%, under-and overestimating the measured illuminance values. However, in the afternoon hours, the deviation for the UDI was as high as for the DA values. Therefore, this study concludes that direct light beam incidence can be a determining factor for program selection and sky model configuration when conducting daylight simulations in the tropical belt. Furthermore, the results of this study contribute to identifying lighting and visual discomfort problems and allow informed decisions for software tool selection, as well as adequate simulation parameters to be applied under tropical sky conditions.