Use of Genetic Optimization-Algorithms in the Design of 3D Concrete Printed Shell-Structures (2024-09)¶

A transition from disparate data to interconnected and contextually integrated data is currently causing a paradigm shift in the architecture industry. The need for fabricationaware architectural representation models, that enable designers to interface with today's data-intensive manufacturing technologies, is a direct consequence of new concepts such as smart fabrication, automation and vertical integration. Compared to conventional concrete casting methods, 3D Concrete Printing (3DCP) offers a wide range of advantages, particularly the ability to create complex geometry. A lack of computational modelling techniques that link design and production for 3DCP is currently making it difficult to predict the printability of designs. This research presents a unified design-tofabrication framework using machine learning (ML) that is customized for freeform steelreinforced 3DCP shell structures. 3DCP is used to create incrementally cast sacrificial formwork. In particular, the segmentation process is fed back into the design process using genetic optimization for a fabrication-aware design model. The framework is validated with a series of physical experiments.