Computational models applied to the service life prediction of External Thermal Insulation Composite Systems (ETICS)

Abstract The degradation of the built environment, associated to high maintenance and repair costs, increases the relevance of the studies about the durability of construction elements. External Thermal Insulation Composite Systems (ETICS) are a relatively recent cladding solution with a growing use in the construction industry, due to their advantages. This study proposes a methodology to predict the service life of ETICS, through two computational methods: artificial neural networks and fuzzy logic systems. The proposed models lead to an average estimated service life of 21 years, considering a sample of 378 ETICS located in Portugal, which is in accordance with the empirical knowledge and other studies related with the durability of these cladding systems. The results obtained in this study aid in the adoption of more durable and sustainable solutions at the design and maintenance stages, improving the quality of the adopted construction solutions, and reducing the costs of these elements during their life cycle.