Surrogate-based optimum design of 3D reinforced concrete building frames to Eurocodes
Panagiotis E. Mergos
Abstract
The optimum structural design of real-world 3D concrete building frames to modern design standards is a complex and computationally expensive task. Hence, the use of surrogate-based optimization (SBO) methodologies must be investigated to reduce computational cost. The present study applies, for first time, a fully-fledged SBO algorithm to the optimum design of 3D concrete building frames. More particularly, the algorithm is applied to the minimum material cost design of a 4-storey and a 12-storey 3D RC building according to Eurocodes. It is found that the SBO algorithm can converge earlier than other well-established metaheuristic optimization algorithms reducing considerably the required computational effort. Nevertheless, it is likely to get trapped in local optima for large-scale RC frames. To overcome this drawback, a novel hybrid approach is also proposed herein that offers improved computational performance for large-scale concrete building frames.