Incremental dynamic analysis of the long-span continuous beam bridge considering the fluctuating frictional force of rubber bearing

Abstract

In the seismic design of long-span bridges, the classic bi-linear model was used to simulate the frictional restoring force of the rubber bearings. However, in actual earthquake, the rubber bearing suffered fluctuating axial pressure in earthquake, even separated from the beam when vertical component of the earthquake was too strong. Employing the bi-linear model for the bearing may incorrectly estimate the seismic response of the bearings, as well as the whole bridge. This paper developed a nonlinear frictional bearing model, which can consider the variation of the frictional restoring force in the bearings, even the separation with the beam in vertical directions. A typical continuous beam bridge was modeled in ABAQUS, and incremental dynamic analysis was conducted for the quantitative comparison of the seismic responses using different bearing models. The intensity measure was selected as the ratio of the peak ground acceleration (PGA) in the vertical direction to the PGA in the horizontal direction. The analysis results indicated that the different bearing model led to the significant different seismic response for the bearings and piers, even the vertical component was small. The bi-linear bearing model would underestimate the seismic demand of the bearing and piers.

Nonlinear model of bridge bearings considering friction effect under horizontal seismic action

“…During an actual earthquake, the pot bearing may experience axial pressure fluctuations and can separate from the beam if the vertical component of the earthquake is too intense. Using the bilinear model to simulate the friction restoring force of the pot bearing will result in an incorrect estimation of the seismic response of both the bearing and the entire bridge (Liao et al, 2021). In the simulation of the shape memory alloy supplemented with lead rubber bearings (SMA-LRB) hysteretic response, it is found that the bilinear model cannot reflect the actual behavior of the SMA-LRB, which leads to a certain degree of uncertainty in the results (Dezfuli et al, 2017).…”