5. Conclusions
This study derived the optimal number and the location of retrofitted RC columns by linking the optimization technique with a 3D non-linear time-history analysis. Various results were obtained by performing the 3D non-linear dynamic analysis. Moreover, reliable optimization results were acquired using the ACO optimization technique. The main contributions of the study are as follows:
(1) An algorithm that combines the ACO technique and the 3D nonlinear time-history analysis is proposed.
(2) The optimization using the ACO technique shows that reliable results could be derived.
(3) The structure strength varies according to the forcing direction of the earthquake. Its deviation is also large. The strength along the strong axis (Y-axis) is high because of the large number of spans. The strength significantly varies according to the forcing direction in school buildings with asymmetric loads and structures. Hence, the forcing direction must be closely examined, and the risk for the weak axis (X-axis) with fewer spans must be recognized.
(4) The seismic material proposed in this study must be installed in over 50% of the total columns for non-seismic school buildings to meet the current seismic design criteria.
(5) Further retrofitting is required for a higher PGA, which determines the strength of the seismic waves, for the weak axis (X-axis) with fewer spans and for lower floor levels. Retrofitting the outer columns is more effective than retrofitting the inner columns.
(6) A staggered positioning of the retrofitted columns over a wide area is more effective than continuous positioning. Applying different retrofitting locations to each floor contributes to economic benefits.