Experiment-Anchored Probabilistic Seismic Demand Modelling and Fragility Analysis of a Non-Ductile RC Frame

Abstract

This study develops a probabilistic seismic demand model (PSDM) to establish the relationship between intensity measures (IMs) and engineering demand parameters (EDPs) for assessing the seismic vulnerability and risk of non-ductile reinforced concrete (RC) frame buildings. A three-dimensional finite element model was initially constructed and calibrated against experimental test results to ensure reliability. A comprehensive cloud analysis was then carried out using near-fault pulse-like (NF-FD), near-fault non-pulse-like (NF-NFD), and far-fault (FF) ground motion sets. In the first stage, thirty-one widely adopted IMs were systematically categorized into earthquake-based, event-based, and structure-based groups. Regression analyses were subsequently performed for five representative EDPs to quantify the predictive capacity of IMs in capturing structural response demands. The efficiency, correlation, practicality, proficiency, and sufficiency of all IMs were rigorously examined to identify the most appropriate parameter. The results indicated that the structure-based IM, Sa(T 1 ), consistently outperformed other measures and was recognized as the optimum IM across all three ground motion sets. Among earthquake-based IMs, the Acceleration Spectrum Intensity (ASI) proved optimum for both NF-FD and NF-NFD records, while the Velocity Spectrum Intensity (VSI) was identified as the most appropriate for FF records. Fragility functions were subsequently developed for the optimum and other selected IMs across the considered EDPs. Using fragility relationships, Expected Loss Ratio (ELR) values were computed and validated against shake table experiments subjected to FF records. A strong and statistically significant correlation was observed between the analytically derived ELR estimates and experimental results. The findings of this study provide critical insights into the seismic performance assessment of non-ductile RC frame structures, contributing to a more robust understanding of their vulnerability under diverse ground motion scenarios. The proposed framework also offers a valuable reference for future research efforts and the development of effective seismic risk mitigation strategies.