Multi-objective aerodynamic optimization of the exterior shape of tall buildings with trilateral cross-section

Abstract

Wind-induced loads are largely dependent upon the exterior shape of buildings, and one highly effective procedure to mitigate them is to apply aerodynamic shape modifications in the aerodynamic optimization procedure (AOP). This study presents the framework of an AOP for shape modifications of the trilateral cross-section tall buildings. The AOP is comprised of a combination of multi-objective optimization algorithm named non-dominated sorting genetic algorithm II (NSGA-II), artificial neural networks, and computational fluid dynamics. The building shape is designed based on the geometric description of its vertical and horizontal profile using seven geometric parameters (design variables) to apply different types and sizes of modifications. In addition, the mean moment coefficients in drag and lift directions are considered as the objective functions. The proposed procedure investigates the effect of the three types of modifications including varying cross-section sizes along the height, twisting, and curved-side on the reduction of objective functions. Finally, a set of optimal building shapes is presented as the Pareto front solutions, which enables the designers to select the optimal shape of the building with additional considerations. The results indicate the high capability of the proposed framework to make appropriate use of various aerodynamic modifications in order to upgrade the aerodynamic performance of the trilateral cross-section tall buildings.

Keywords

• wind load
• tall building
• computational fluid dynamics
• multi-objective optimization
• artificial neural networks