Identifying buildings with high collapse risk based on samos earthquake damage inventory in İzmir

Abstract

Samos Earthquake caused the collapse of about fifty buildings in İzmir city center, resulting in over 120 fatalities. The response spectra of the ground motions at the soft soil sites in Izmir revealed that the spectral accelerations in the period range of 0.5–1.5 s are similar to the spectral accelerations defined by the response spectrum corresponding to the 72-year return period. Despite experiencing accelerations lower than those defined by the Turkish Code design spectrum (475-year return period), significant damage was observed due to the deficiencies of the building stock. Several factors such as soft story, lack of code-compliant transverse steel reinforcement, and low concrete strength have contributed to the devastating loss. Studies in the last two decades on deformation capacity estimation of RC columns showed that key parameters for collapse drift limits are the axial load ratio (i.e., axial force demand to capacity ratio) and transverse reinforcement amount. Most of the collapsed buildings in the past Turkish earthquakes had low concrete strength resulting in high column axial load ratios with limited drift capacity. In order to identify the collapse of vulnerable buildings, a simple procedure is proposed based on vertical and lateral pushover analysis results. The method relies on estimating the average axial load ratios of the most critical columns and their seismic drift demands. An axial load ratio-drift demand limit state model was developed to identify the buildings with high collapse risk. The model was validated with the damaged buildings in İzmir after the Samos earthquake to capture the buildings with poor performance even under service level earthquakes. The proposed model is practical as it only requires the floor plan and concrete compressive strength. It is found to be legitimately accurate to identify and prioritize the buildings prone to collapse.