In this study, five 1/3-scale bridge column specimens are experimentally and numerically investigated, which includes one cast-in-place (CIP) reference and four precast segmental columns with different design details. The four precast segmental bridge columns are of the same design details in terms of dimension and mild reinforcement, while different number of segments, various full-length connection rebar grades, presence of prestressing force, and two types of shear keys are compared. Seismic performance of the five specimens is compared under the same quasi-static cyclic loading protocol. Based on the experimental results, the precast segmental specimen solely using mild reinforcement exhibits similar seismic performance to the CIP reference, while segmentation slightly reduces post-yield stiffness and dissipated energy but increases the self-centering capability. Implementation of prestressing force significantly improves the self-centering capability of the columns, and higher grades of steel also increase the overall strength of the columns. Less number of segments in the column will lead to better structural integrity with smaller joint opening widths and higher energy dissipation capacity. A new finite element model is also proposed, and maximum deviation from test results is less than 2% in terms of stiffness and peak strength. This demonstrates that the Parallel material for rebar modeling and ZeroLength element for joint interface (including bond–slip behavior of rebar) simulation are both effective.
Read full abstract