Titolo della tesi: Optimization of Spatial Structures for Sustainable Construction: Integrating Shape and Topology Optimization to Minimize Environmental Impact and Enhance Buildability
The construction industry is one of the largest contributors to global carbon emissions, significantly impacting the environment. This dissertation addresses the challenge of optimizing spatial structures for sustainable construction, by integrating shape and topology optimization to enhance material distribution, structural efficiency, and minimize environmental impact, specifically by reducing Global Warming Potential (GWP) and improving buildability through effective segmentation. This research aims to develop innovative methods that enhance structural performance, resource efficiency, and constructability, ultimately leading to feasible and sustainable solutions for complex geometries.
The research employs a multi-disciplinary approach, involving shape optimization using Dynamic Relaxation (DR) to identify structurally efficient forms, and topology optimization using Solid Isotropic Material with Penalization (SIMP) and the Method of Moving Asymptotes (MMA) to optimize material distribution. Machine learning algorithms, particularly the Gaussian Mixture Model (GMM), are used to rationalize complex shell geometries by segmenting them into modular panels, reducing fabrication complexity. The GMM's effectiveness in identifying patterns enables the grouping of similar geometrical features, thereby streamlining the fabrication process and bridging the gap between digitally optimized designs and practical constructability.
This research is structured around three main contributions. First, an in-depth theoretical framework for shape and topology optimization is developed, extending existing models and incorporating sustainability considerations, such as minimizing embodied carbon, reducing GWP, and improving material efficiency. This advancement provides a more nuanced understanding of shell design, resulting in solutions that enhance structural strength and reduce environmental impact. Second, rationalization techniques supported by machine learning clustering simplify the fabrication and assembly processes, significantly improving buildability and the feasibility of constructing optimized, complex shell forms. Finally, the research includes a practical design application through a case study utilizing Design for Manufacturing and Assembly (DfMA) strategies, with reclaimed timber as a primary material. Advanced robotic fabrication is incorporated to enhance precision, reduce waste, and streamline construction. This case study demonstrates the feasibility of the proposed methods, highlighting their potential for real-world applications in sustainable construction.
The findings presented in this thesis advance the state of knowledge in architectural shell optimization, providing a new foundation for future research and practical implementations. By integrating form finding, material distribution, and sustainable materials like reclaimed timber, the research ensures that intricate architectural forms can be realized in a resource-efficient manner, minimizing embodied carbon and reducing GWP. Additionally, advanced robotic fabrication is used to enhance precision and streamline construction.