New Simplified Method for Designing Seismically Isolated Highway Bridges with Massive Piers

Abstract

This paper proposes two new models for the simplified seismic analysis of seismically isolated highway bridges with massive piers. Both models include two components, one of which includes the superstructure mass, the isolators, and massless pier, and the other one the pier with its distributed mass and stiffness properties. In one model, the isolator stiffness is added at the top of the pier in the second component. The total seismic response is obtained from the square root of the sum of squares (SRSS) of the two individual component results. Applicability and accuracy of the models were assessed by considering bridges with wide ranges of stiffness and mass properties of piers (Mp) and superstructures (Mss). Multimode spectral analysis (MMSA) and nonlinear time-history analysis (NLTHA) were used to define reference solutions for the superstructure displacements, vibration periods, and shears and moments at pier bases. Results from four currently available models, including the model prescribed in American and Canadian bridge codes, were also examined. The study shows that current models for the simplified analysis method yield good estimates of superstructure displacements and vibration periods; however, they generally underestimate the shear and moment demands at pier bases. The errors are more significant for more massive piers or when the stiffness of the piers is high compared to that of the isolators. In contrast, for all bridges studied, the seismic responses from the proposed models show very good agreement with those from the MMSA. Better predictions are obtained when considering the isolator stiffness in the second component of the model. The NLTHA results also show that the proposed models give overall satisfactory predictions for pier base shears and moments. The proposed models extend the range of application of the American and Canadian code simplified method to isolated bridges with massive piers while preserving its simplicity with limited extra computational effort.