Assessing seismic vulnerability of structures with damper using an ANN-based approach

The study proposes assessing the seismic vulnerability of G + 15-storey reinforced concrete (RC) buildings using finite element analysis (FEA) software (ETABS) and artificial neural networks (ANNs). The study utilizes finite element models of G + 15 buildings subjected to recent earthquake data, analysing them for seismic vulnerability and incorporating various retrofitting techniques, such as fluid viscous dampers (FVD). The damper locations are varied in the structure for the entire earthquake data considered to study the seismic vulnerability in storey displacements, storey shear, and storey drift. Key structural characteristics were systematically modified, and their impact on seismic response was evaluated through modal dynamic and non-linear time history analyses. The FEA results are used to train an ANN algorithm, creating a function that can predict the seismic behaviour of similar RC structures. This approach offers a fast and potentially generalizable method for seismic vulnerability assessment.