Steel Truss Optimisation and Segmentation Strategies for Large Spans

Abstract

This thesis presents a material-efficient steel roof truss design for the redevelopment of The Pier in Scheveningen. A parametric workflow integrating Rhino-Grasshopper, Karamba3D structural analysis, and cross-section optimisation is developed to minimise total structural steel weight. Four architecturally compatible truss alternatives are explored, resulting in the selection of a Pratt truss with a total steel mass of 31 tonnes. This configuration corresponds to approximately 48 tonnes of embodied carbon (CO2-eq, lifecycle stages A1–A3) and achieves a 27% mass reduction compared to the heaviest considered alternative. Insights from four Haskoning case studies informed a practical two-segment truss segmentation strategy, with splice plates strategically located in the shear-dominant zones. This segmentation approach limits individual segment weights to 11.3 tonnes, complying with Dutch transport restrictions (maximum dimensions of 23 m × 3.5 m × 2 m x 32 t). Validation of the optimisation and segmentation methodology was performed using RFEM analysis and Eurocode-based hand calculations, demonstrating deviations below 1% for mass and deflection criteria. Interviews conducted with Dutch steel fabricators underscored steel member self-weight as the primary driver of fabrication costs, reinforcing the validity of the weight-focused optimisation objective. The study demonstrates the effectiveness of integrating parametric geometry optimisation, cross-sectional sizing, and transport-driven segmentation strategies. The resulting design approach achieves substantial reductions in material use and embodied carbon emissions while ensuring practical constructability and compliance with structural codes.