DTE AICCOMAS 2025

Integrating form-finding and evolutionary optimization for gridshell construction

  • Melchiorre, Jonathan (Department of Structural, Geotechnical and Building Engineering, Politecnico di Torino)
  • Manuello Bertetto, Amedeo (Department of Structural, Geotechnical and Building Engineering, Politecnico di Torino)
  • Trovalusci, Patrizia (Department of Structural and Geotechnical Engineering, Sapienza Università di Roma)

Please login to view abstract download link

Gridshell roofing structures are widely used in modern engineering and architecture for their ability to cover large spans with lightweight, slender elements. However, their adoption is often constrained by fabrication complexity and high costs [1]. Given these challenges, optimization techniques play a crucial role in enhancing material efficiency and simplifying construction. The Multi-body Rope Approach (MRA) [2] is a well-established form-finding method for generating funicular gridshell geometries, offering structurally efficient solutions. This study introduces an advanced optimization methodology for free-form gridshells by integrating evolutionary algorithms with an enhanced version of MRA, known as the improved Multi-body Rope Approach (i-MRA) [3]. The framework incorporates Multiple Order MRA (MO-MRA) and Repulsive Nodes MRA (RN-MRA) to minimize structural component variability, streamlining fabrication and assembly. Coupled with the Non-dominated Sorting Genetic Algorithm (NSGA-II), the approach optimizes material usage, reduces production waste, and ensures structural compliance. Additionally, a cutting stock problem formulation is implemented to further minimize material waste and optimize the manufacturing process. The proposed method was validated through case studies of increasing complexity, demonstrating its effectiveness in balancing structural performance, material efficiency, and construction feasibility. This research presents a novel strategy for sustainable, cost-effective, and manufacturable gridshell design.