Pattern Recognition of the Seismic Demands for Tall Pier Bridge Systems

Abstract

ABSTRACT An essential and emerging task in the area of performance-based seismic assessment of bridges is to improve the estimation of the seismic demands. Existing research proved that the bridge column is one of the dominant components in estimating the bridge system vulnerability. However, the seismic performance of bridges with tall piers is not deeply investigated yet. This study aims to address this problem through analysis of multi-span concrete box-girder bridges with tall piers. This study implements a clustering algorithm that identifies optimized groupings which can set up the basis to establish the best choice for generating MLPSDMs. The evaluation results implicate better performance of cluster-wise multiple models than a single model to represent the variation in the bridge demands. The insights gained from the findings of this study can pave the way toward a better realization of the seismic performance of bridges having tall piers.