3rd Croatian Conference on Earthquake Engineering 3CroCEE

This research investigates seismic risk assessment for bridge structures using a combination of conventional and next-generation intensity measures (IMs). The study focuses on simplified bridge models, characteristic of European construction, and explores the efficacy of multiple IMs, including spectral acceleration (Sa), average spectral acceleration (Sa_avg), and filtered incremental velocity (FIV3), in predicting seismic demand. Probabilistic Seismic Hazard Analysis (PSHA) is conducted for each IM separately, allowing us to derive seismic demand hazard curves for each one, along with their corresponding dispersion versus the engineering demand parameter (EDP). This comprehensive analysis provides a deeper understanding of the reliability of each IM across various nonlinearity levels.
Our findings demonstrate that Sa exhibits the lowest dispersion in the elastic and early non-linear stages, making it most reliable for low-intensity seismic events. However, as non-linearity increases, Sa_avg demonstrates superior performance, showing the least dispersion for low to high non-linear conditions. In cases of high non-linearity and near-collapse scenarios, FIV3 outperforms the other IMs by offering the lowest dispersion.
In addition, we demonstrate hazard consistency by using a single ground motion model (GMM) for all IMs, along with correlation models derived from the same GMM and filtered ground motion database. This more unified and consistent approach was compared to the common method of employing different GMMs and correlation models from the literature for each IM.
The case study comparisons illustrate the benefits of using a unified treatment of hazard, particularly in reducing discrepancies and improving consistency in seismic demand assessments. The outcomes provide valuable insights into the selection and combination of next-generation IMs with conventional ones, offering practical guidance for more accurate seismic risk assessments and bridge design practices.