Akademik Veri Yönetim Sistemi
43rd Education and Research in Computer Aided Architectural Design in Europe Conference (eCAADe), Ankara, Türkiye, 1 - 05 Eylül 2025, cilt.2, ss.435-444, (Tam Metin Bildiri)
Sustainability and optimization in material consumption are important concerns for tall building design, which are gaining importance as the number of tall buildings increases
dramatically. While numerous studies in literature address optimization in tall building design, majority of them focus on the final stage of design and conduct size optimization
for structural members. This approach neglects the effect of architectural design parameters on optimization process and cannot guide designers in preliminary stage where key decisions are made. In this study, a methodology is developed for tall buildings design, integrating parametric modelling and optimization via genetic algorithms (GAs), to address architectural design parameters and structural system requirements simultaneously. The aim of the methodology is to provide sustainable design solutions with minimum carbon footprint per leasable area, to guide designers beginning from the preliminary design phase, and to diminish drawbacks in design process. To ensure compatibility with the preliminary design and manage vast number of solutions, automated parametric modelling and structural analysis are introduced. Instead of leading to one optimum solution, it gives a group of optimum solutions to provide flexibility to designers. A solution space is defined with four design parameters: building height, plan dimensions, core ratio, and structural system. The methodology is applied in two steps, a portion of solutions is used to investigate different configurations of GAs, then the best configuration is applied to whole solution space. The results show that by applying GAs, a group of optimum solutions and optimum values for design parameters can be deducted by analysing only 40% of all solutions. Proposed method is beneficial for managing vast solution spaces, enabling designers to explore wide range of options without excessive number of analyses. Since the proposed method is based on parametric inputs, designers can adapt it to different design problems, solution spaces and design parameters.