Integrated Structural and Daylight Optimisation of Variable Transmittance Façades

Abstract

Traditional fabrication methods for plastic building panels, such as moulding and extrusion, have recently been advanced by large-scale robotic 3D printing (LSR3DP), enabling mass customisation and the production of complex architectural geometries. While existing research on LSR3DP has primarily focused on single-material printing, the exploration of multi-material or multi-property applications remains limited, especially at full architectural scale. This study addresses this gap by developing a performance-driven digital workflow for PETG-based façades that integrates structural efficiency with solar-responsive transmittance gradients. A multiobjective optimisation process using the Non-dominated Sorting Genetic Algorithm II (NSGA-II) generated 16 optimal façade geometries across four orientations (north, east, south, west), achieving up to 14% reduction in summer solar radiation and 26% increase in winter solar gain compared to a conventional vertical façade, while minimising structural displacement. The optimal south-facing solution was selected for detailed daylight performance assessment. A procedural gradient generation workflow was developed to discretise solar-based transmittance values across varying mesh densities and gradient resolutions. The best-performing variable transmittance configuration achieved 46.24% Useful Daylight Illuminance (UDI-a) and 69.21% spatial Daylight Autonomy (sDA), representing a 25.94% improvement in UDI-a over a conventional uniform-transmittance curtain wall. This integrated approach demonstrates LSR3DP’s potential to produce unified, materially expressive façades that embed environmental performance directly into form and material logic, eliminating reliance on mechanical shading systems.