Chi and Liu 2012
This paper presents experimental and analytical investigation of cyclic response of a post-tensioned (PT) column base connection. The PT column base connection is designed to eliminate structural damage at column bases in SC-MRFs under seismic loading; the softening behavior at the connection is provided by gap opening and elongation of PT bars rather than yielding in the column.
System Concept
The proposed PT column base connections are incorporated at the grade beam level. The PT bars are anchored between half-height of the first story columns to close to the bottoms of the below-grade columns. Therefore, the gap opening and closing behavior occurs at the grade beam-column interface under seismic loading. This configuration was deliberately chosen to allow for PT bars that are long enough such that they would not yield even in a MCE while also preventing yielding at the top of the first story column (i.e., location of high axial and moment demand) and avoiding congestion at the beam-column joint (i.e., interference with the PT strands for the beam-column connections of the SC-MRF).
The components of the PT column base connection include PT high strength bars for clamping and restoring forces, buckling restrained steel plates for energy dissipation and damping, column flange reinforcing plates for increased bearing and shear yielding resistance, and keeper plates for increased shear resistance. Buckling restrained steel (BRS) plates and keeper plates are welded together and bolted to the column flange and grade beam, respectively. The BRS plate, cover plate, and column flange are expected to move together as the gap opens at the column base, therefore friction between these elements is assumed to have a negligible effect.
The basic design philosophy for the PT column base is that no column axial or flexural yielding, no PT bar yielding, column slip, or total column uplift is permitted, even for the MCE. Some limit states such as localized yielding in the panel zone and at the column flange due to concentrated bearing forces are permitted. For the design basis earthquake, limit states that do not affect an immediate occupancy performance objective are permitted; these include connection decompression and yielding of the BRS plates, which serve as replaceable fuses.
Experimental Study, Results, and Discussion
The PT column base connection subassembly was based on design of the prototype 6-story 4-bay SC-MRF with PT column base connections. Two-thirds scale PT column base subassemblies were subjected to axial load and cyclic lateral displacements. Inflection points were assumed to be located at column mid-height and beam mid-span. Axial force in the column was assumed to be constant for an interior column base connection. It is noted that the column axial force was applied at the centroid of the column during the test. Also, the horizontal movement at grade level was restrained. Varying axial force in the column was considered for an exterior column base.
Specimens were tested up to a lateral drift of 4%, at which point no significant yielding was observed in the column or beam except for localized yielding in the panel zone.
The limit states that influenced the moment rotation response the most were PT bar yielding and BRS plate fracture. In general, PT bar yielding occurred in cases for which the initial PT force was relatively high. Test results confirmed the need to consider contribution to the moment from the axial force in the column as well as the BRS plate and the PT bars. The moment contribution of each component changes as the drift increases.
Although it is contrary to the design objectives for PT column bases, the effect of PT bar yielding at an early stage on the behavior of the PT column base connection was also investigated. Significant degradation did not occur during 4% drift cycles. Even as the PT bars yielded, the column did self-center after the test since the PT bar yielding was not considerable. If the specimen had been displaced further, the loss of the PT force may have been significant due to the plastic deformation of the PT bars.
Analytical Study, Results, and Discussion
Connection moments predicted by the simplified equations presented in this paper compare well to the test results, with good correlation to the backbone curves from the experimental results. The unloading behavior is also captured reasonably well, with some underestimation of the energy dissipation.
Reference
Chi, H., & Liu, J. (2012). Seismic behavior of post-tensioned column base for steel self-centering moment resisting frame. Journal of Constructional Steel Research, 78, 117-130.