Strengthening of shear critical reinforced concrete columns by steel-rod collars

Abstract

The inadequate shear strength of the existing old reinforced concrete columns located in the seismically active areas can lead the catastrophic shear failure. In this study, strengthening of shear-critical reinforced concrete columns by the steel-rod collar method is investigated. This proposed method is less intrusive to the existing building components. The steel-rod collar comprises the four sets of welded steel angles connected by steel-rods threaded at the ends. The three specimens were divided into one unstrengthened column, and the two strengthened columns. All the specimens were tested under the constant axial load and the reversed cyclic loadings. In addition, the specimens were the vertical cantilever type and used the same section properties. The number of steel-rod collar installed on the strengthened columns were the main parameter in this study. The unstrengthened column failed in shear failure mode and the diagonal shear cracks occurred at a drift ratio of 2%, whereas, the two strengthened columns failed in flexural mode and had an increase in the lateral load capacity, the ductility and the energy dissipation capacity than the unstrengthened column. The difference between the two strengthened specimens were the number of steel-rod collars mounted on the specimens. The spacing of steel-rod collars was 200 cm in the SC-200 specimen while that of the other strengthened specimen, SC-100 was 100 cm. After testing the strengthened columns failed in flexure and they had an increase in the lateral load capacity and the ductility. The drift ratio of both specimens reached 5%. Furthermore, the finite element models for all specimens were developed using the OpenSees program. The forced based beam column element with a shear spring was established to detect the behavior of the unstrengthened column. The lumped plasticity model was used to detect the behaviour of strengthened columns. The comparison results showed that the predicted numerical responses were in close agreement with test results.