The optimal performance of critical bridge infrastructure plays a key role in upholding the functionality and resilience of transportation networks following significant seismic events. As such reliable methods for evaluating and quantifying bridge performance are essential. This work aims to quantify the effect on structural performance of two analysis considerations not included in conventional evaluation methods. The two cases studied are (1) the effect of ground motion duration correlation to ground motion intensity and (2) the effect of accounting for isolation bearing end-plate rotations.
The first step in a performance-based evaluation is obtaining site-specific ground motions for input into nonlinear structural models. Site-specific ground motions are generally selected based on magnitude and distance for the structure of interest. While others have studied the effect of duration, the correlation with ground motion models has not been included. Researchers have found that duration and spectral acceleration are negatively correlated, which means that the target duration should decrease when the intensity of the ground motion scenario increases. As such, the current state of practice may be overestimating the damage probability for structures in high seismic zones. This work compares the effect of the duration correlation for a major bridge and a typical highway overpass.
The following step in the analysis is to develop a numerical model that can accurately capture the structural response. The case of interest in the second part of this work is the behavior of isolation bearings subject to rotational demands, as is the case for bearings placed on top of flexible bridge piers, a condition observed in the major bridge used as a case study. This is relevant because a typical design assumes no rotation of bearing end-plates and past research has shown that bearing end-plate rotation leads to a reduction in the horizontal stiffness of isolation bearings. To capture the changes in bearing behavior due to rotation, the study utilizes a hybrid simulation experiment to inform and improve numerical modeling.
Fragility curves are developed for a moderate and a major damage state based on pier drift and bearing shear strain for both studies, the effect on bridge performance is evaluated in each case by comparing the median probability of damage to that of baseline fragilities developed using traditional methods. It is shown that considering ground motion duration correlation to intensity leads to a reduction in the predicted median damage for the major bridge and the typical bridge, and the effect is dependent not only on the correlation parameters but also on the amount of degradation included in the numerical model. A low correlation coefficient (small changes in predicted duration) may lead to large changes in performance if the model includes additional degrading elements like the lead rubber bearings in the major bridge. By contrast, it is revealed that significant changes in the bearing shear and rotational stiffness, when considering bearing rotation, resulted in minimal impact on performance when measured by damage fragility functions.
