Self-shaping building skin: Comparative environmental performance investigation of shape-memory-alloy (SMA) response and artificial-intelligence (AI) kinetic control

Abstract

This paper presents the first controlled experimental investigation on the environmental performance of self-shaping adaptive façade, which is a kinetically-responsive building skin activated by smart material. Although adaptive façade is an emerging subject in architectural design and research, its effectiveness has been rarely addressed in terms of building performance. Thus, in this study, we investigated indoor temperature and daylight illuminance of a thermo-responsive skin actuated by shape-memory alloy (SMA). For comparative analysis in a controlled situation, two identical lab-scale (1:20) building models with kinetic louver shading were designed, representing a south-oriented living room of 3.4 m✕5.2 m. The shadings were climate-adaptively operated by a conventional electromagnetic motor (baseline) and an SMA actuator (test case), respectively, for the same time period. The mechanical characterization and simulation of SMA-bias coil spring actuation were studied, and an artificial neural networks (ANNs) model was employed for the baseline's optimal dynamic control to approximate the American Society of Heating, Refrigerating, and Air-Conditioning Engineers (ASHRAE) Standard for thermal comfort (19.4–27.8 °C) and a minimum illuminance level of 500 lx. The study's findings indicate that SMA actuation is effective for controlling indoor daylight intensity, thereby improving overall indoor environmental conditions. Nevertheless, the climate-responsive self-shaping was not as instantaneous as the optimal kinetic control. Average performance was probabilistically different; however, a significant correlation was found in illuminance variation.