CALIBRAÇÃO DE UM MODELO COMPUTACIONAL PARA ANÁLISE DE DESEMPENHO TÉRMICO

Abstract

The construction of computational models is a process that requires a large number of input variables. Therefore, the search for accurate results in building thermal simulation is a process involving many uncertainties. In this sense, the calibration is indicated as a way to provide greater safety to the models, but it was verified that few studies carried out calibrations for better accuracy of data related to thermal performance analysis. This paper presents the results of a calibration process based in infiltration and natural ventilation periods monitored in a standard kindergarten classrooms located in Maceió-AL. Comparing monitored and simulated temperatures after calibration, it was obtained a mean error equal to 0,71 and the sum of the squared errors of the various rooms it was 3,64, while the average difference between hourly temperatures was equal to 0,7°C. This results indicates a good accuracy of the model, because deviations are lower than precision of data loggers used during monitoring. Among the adjusted variables, thermal exchanges through the ground corresponds to the parameter that most influenced the calibration process.