Optimum Shape of Steel Single-Layer Barrel Vaults Using Genetic Algorithms

Abstract

This paper studies the optimization problem of single-layer steel barrel vaults as a type of space frame under the effects of self-weight and earthquake loads. Genetic Algorithms were used to find the optimum sections of the structure. The minimization considered the weight of the structure as an objective function and the design variables were the cross-section area of the members. Suitable sections from standard section lists of the American Institute of Steel Construction (AISC) were selected throughout the optimization process. Strength and displacement constraints were formulated according to the American Institute of Steel Construction-Load and Resistance Factor Design (AISC-LRFD). The weight of the barrel vaults increases with the increase of the depth to span ratio after 0.2. The barrel vaults with a depth to span ratio of 0.2 has the greater reduction in weight for the three studied cases (pipe, box, and angle sections) so the ratio of 0.2 presents the optimum case. The barrel vaults with box sections have an optimum than the other section shapes (about 20.4% of that of the pipe section and 13.6% of that of the angle section). The Pipe section is not economic to use in this type of barrel vault.