Innovative Computational Intelligence-Based Model for Vulnerability Assessment of RC Frames Subject to Seismic Sequence

Abstract

Previous earthquakes have shown that for structures that experience a significant ground motion with only minor damage, there is a likelihood of severe damage or complete collapse during the aftershocks. Due to the importance of such seismic sequences in structural engineering investigations, in this study, the vulnerability of RC frames under subsequent earthquakes is evaluated using computational intelligence (CI) in three phases. In the first phase, an optimized fuzzy system (Fuzzy-TLBO) is presented to estimate the used capacity of the story by considering the damage state of beams, columns, joints, and also the drift. The second phase includes two soft computing models to determine the used capacity of the frame under mainshock using indicators of the soil, record, and the structure, simultaneously. The final phase utilizes two soft computing methods with the same goal of the second phase in which the initial vulnerability of the frame concluded from the mainshock is also introduced to the estimators. These three approaches, with high accuracy in their outputs, have several benefits and applications, especially in seismic studies. They can calculate the overall output of an RC frame under loads without needing a nonlinear time-history analysis. Furthermore, they consider the effects of beams, columns, joints, and also the drift in the final output. Therefore, the proposed models in this research are a very comprehensive approach compared to the available methodologies, which only consider drift as the engineering demand parameter.