Form-Finding of a Negative-Gaussian Curvature Cable Dome Using a Genetic Algorithm

Abstract

AbstractCable domes have been widely used for covering large areas such as arenas and stadiums due to their lightweight, adaptable forms and their architectural impact. As a preliminary step in designing cable domes, the feasible sets of internal forces and the geometrical shape should be simultaneously defined. This work proposes a two-stage numerical method that performs form-finding of cable domes with zero self-stress states using the Singular Value Decomposition (SVD) method to obtain a feasible geometry with at least one self-stress state. The first stage is a genetic optimization algorithm that uses a NURBS surface to control the dome configuration under predefined sag and rise values, and ensure that the algorithm will not fall in a local minimum. The second stage is a local optimization aiming at enhancing the accuracy of the optimum solution. The proposed algorithm takes the advantage of the randomicity, rapidity and wholeness of the genetic algorithm and the high accuracy of the local optimization algorithm. To verify its efficiency, this method has been applied on a numerical example of a negative-Gaussian curvature cable dome, first proposed by Guo and Zhu [5]. The convergence of the proposed algorithm and the feasibility of the obtained geometry are checked and discussed in details, proving the efficiency and robustness of the proposed method.