Comparative Models and Geometrically Constrained Optimization of a Multi-Domain Variable Section Beam based on Timoshenko Model using Algorithm-Aided Design

Laura Sardone, Marco M. Rosso, Raffaele Cucuzza, Rita Greco and Giuseppe C. Marano

28-30 June 2021, Streamed from Athens, Greece

Abstract

This paper presents an efficient strategy to minimize the volume of a large span multi-domain variable section beam considering the geometric shape parameters as mathematical constraints. The shape optimization of the beam element has been conducted through an imposed geometry to find the best shape between the design-decision making and the structural efficiency. The study, based on the kinematic hypothesis of Timoshenko, focuses on a test case retrieved from the project designed by Paulo Mendes da Rocha and the engineer Siguer Mitsutani developed on the occasion of the Japan World Exposition, Osaka, 1970 (Osaka's Expo '70). The structural component has been remodeled and optimized through different approaches that generate comparative numerical models joining the combinations of Computational Design and Algorithm-Aided Design. Even though very abundant knowledge and literature on structural optimization already exists, this study aim not only to study the certain structural element undergone to a specific emptying function but to compute and chart the results to be used for empirical purposes. The results of the study show, in the search of the architectural optimal solutions, advantages regarding the performance of the structures and the control of the shape of the architectural component giving - at the same time - the possibility to join the needs of architectural narratives with the stability and efficiency of an optimized and correctly designed structure.