Seismic resilience improvement of earthquake-damaged low-reinforcement RC bridge columns retrofitted with SMA bars and sway mechanism

Abstract A considerable number of reinforced concrete (RC) bridge columns are characterized with low reinforcement ratios, rendering existing bridge is susceptible to severe damage under strong earthquakes, and challenging post-earthquake rehabilitation. In order to improve seismic resilience of these low-reinforcement bridge columns, a comprehensive retrofit technique is proposed jointly adopting high performance materials and sway mechanism to achieve a self-centering behavior. This approach integrates SMA bars replacement for existing earthquake-damaged rebars and a sway interface at the bottom of the column, forming self-centering rocking columns for bridges. First, five low-reinforcement RC columns were designed and manufactured for pre-damaged quasi-static cyclic tests. Then, the damaged columns were retrofitted in accordance with abovementioned method, and the seismic performance and resilience capacity were evaluated and analyzed. Finally, a non-iterative static analytical model is proposed, considering the compression zone height based on deformation characteristics of the retrofitted columns. The results indicated that the proposed comprehensive retrofit method was feasible and effective to achieve self-centering rocking behavior for retrofitting earthquake-damaged low-reinforcement RC bridge columns. The retrofitted columns exhibited a pronounced “flag-shaped” hysteresis behavior, with a 15% increase in lateral resistance compared to the original design state. Residual displacements were reduced by more than 75%, indicating that the self-centering capability of these retrofitted columns was significantly enhanced. Higher SMA bar ratios notably reduced residual displacement but compromised energy dissipation capacity. Similarly, shorter SMA bars resulted in premature bar fracture, reducing energy dissipation without significantly impacting peak load-bearing capacity. The proposed analytical model accurately predicts the lateral strength and multi-stage loading stiffness of the retrofitted columns, with a peak load capacity error within 4% compared with experimental values.