Practical multi-objective optimal design of 3D reinforced concrete moment frames using NSGA-II
Jaemin So,Seungjae Lee,Koichi Kusunoki,Donwoo Lee
Abstract
This study presents a practical multi-objective optimization framework for 3D reinforced concrete (RC) moment-resisting frames using NSGA-II. Its main novelty is a decoupled genetic encoding of column rotation angles (0°/90°), combined with a section database that embeds ACI 318-19 reinforcement detailing—crossties, skin reinforcement, and seismic hooks—directly as discrete design variables. Two objectives are minimized: a normalized sum of total construction cost and embodied CO₂, and the maximum inter-story drift ratio. On 4-, 6-, and 8-story irregular benchmark frames, the decoupled approach (Scenario A) attains a 4.5% higher Hypervolume and up to 22% lower cost at equal drift than the conventional coupled approach (Scenario B), and ten independent runs confirm convergence robustness. Construction cost and embodied CO₂ are found to be almost perfectly correlated (𝑅2 ≥ 0.99), so the effective trade-off is between the combined cost–carbon objective and drift; the study also quantifies how the design freedom of this trade-off contracts as building height increases.
Jaemin So, Seungjae Lee, Donwoo Lee — Future Convergence Engineering, Department of Architectural Engineering, KOREATECH, Cheonan, 31253, Chungcheongnam-do, Republic of Korea
Koichi Kusunoki — Earthquake Research Institute, The University of Tokyo, Tokyo, 113-0032, Japan
PDF Viewer
Preview uses the same access rules as Full Text PDF (subscription, purchase, or open access).
