Direct Loss-Based Seismic Retrofit Design of Reinforced Concrete Buildings

In most earthquake-prone areas, under-designed structures that cannot sustain severe seismic demands contribute significantly to seismic risk. Therefore, risk mitigation strategies such as seismic retrofit should be employed to reduce the expected economic and human losses. This thesis introduces a procedure for the retrofit design of reinforced concrete (RC) frame buildings to achieve the desired target level of earthquake-induced loss for a given seismic hazard profile. The presented methodology is “direct” because the loss target is specified in the first step of the procedure, and, in principle, no design iterations are required. The target loss level is defined based on designer/client preferences and/or external constraints (e.g., foundation capacity). The proposed procedure relies on a simplified loss assessment enabled by a surrogate model defining the probability distribution of the seismic deformation demands of single degree of freedom (SDoF) systems given different ground-motion intensity levels. Combined with a hazard curve and a building-level damage-to-loss model, such a surrogate model is used to map candidate SDoF force-displacement curves to their earthquake-induced loss by assuming a given retrofit strategy. In this case, the considered retrofit strategy involves changing the frame’s local hierarchy of strength to ensure a global plastic structure mechanism. Under such assumptions, a designer can select a design force-displacement curve among those that comply with the chosen loss target. The detailing of the retrofitted frame is conducted according to the direct displacement-based design (DDBD) principles and the Simplified Lateral Mechanism Analysis (SLaMA). The procedure is applied to under-designed RC frames building retrofitted (globally or partially) with concrete jacketing. However, the retrofitting technique is relevant only for the design detailing and it can be chosen according to the designer preference, given that the target force-displacement curve is matched. A benchmark loss estimate is calculated using non-linear time-history analyses for loss assessment purposes. The proposed procedure shows satisfactory compliance with the benchmark loss, emphasising the procedure’s effectiveness in practice.