This dissertation investigates the seismic response of Column Base Connections in Steel Moment Frames, a critical structural component used to transfer forces from the steel superstructure to the supporting concrete foundation. The research presented is intended to highlight and address various unresolved issues on the seismic performance of column base connections, and develop methods and criteria for their design resulting in economical and reliable connections.
Column bases are arguably the most important connections in steel structures, transferring forces from the entire building to the foundation. A variety of details are commonly used for these connections ranging from exposed type with anchor rods, to embedded type. Current design/construction practices for base connections result in major conservatism in material requirements (deeper embedments or large anchor rods) and other inefficiencies stemming from the following: (1) Column base connections are generally capacity-designed to be stronger than the adjoining column presuming that the connections will be less ductile than the column, (2) Embedded Column Base connections have been designed for years without direct experimental support, such that current methods assume them to be similar to coupling beams in shear walls; disregarding many physical mechanisms that contribute to the strength of the actual connections. These include, the effect of reinforcement welded/attached to the column, and the beneficial effect of a slab-on-grade that overtops exposed-type connections, resulting in a shallowly embedded (blockout) connection. Consideration of these effects has the potential to greatly reduce costs by decreasing required embedment depth, minimizing other detailing (such as heavy anchor rods), and reducing logistical challenges.
Four studies are presented, investigating (1) the seismic response of dissipative exposed-type base connections, (2) the strength characterization of biaxially-loaded column bases, (3) the seismic performance of blockout column base connection, as well as (4) embedded-type base connection with reinforcement attachments. The first study presents full-scale tests on Exposed Column Base Plate Connections with ductile anchors, with the aim to examine the seismic performance of these connections for their prospective use as dissipative/weak bases. These connections feature upset thread anchor rods, providing a stretch length over which inelastic deformations may be distributed. The tested specimens (with varying parameters) survived, with no anchor rod failure, the application of two back-to-back lateral deformation protocols (each to drift amplitudes of 5%), followed by additional cycles to 6.5% drift amplitude. Complementary line element-based and continuum finite-element simulations are conducted to examine to what extent the experimentally observed response may be generalized to untested configurations.
The second study demonstrates a new method to characterize the internal stress distribution and anchor rod forces in Exposed Column Base Plate Connections subjected to biaxial bending and axial compression. The method is based on and validated against finite element simulations and available experimental data. The method is demonstrated to predict anchor force with good accuracy across a range of configurations (including column and base plate size) and loadings (level of axial force and bending angle).
The third study presents full-scale experiments for shallowly embedded “Blockout” base connections to investigate the effect of additional strength and stiffness provided by an overtopping slab cast over the conventional exposed base connection. The connections are subjected to combinations of axial compression and cyclic lateral deformations. Significant increases in both stiffness and strength, with stable and ductile hysteretic response are noted. Results from this study are synthesized with results from previous studies on similar connections to propose a strength model and evaluate previously proposed stiffness models.
The final study involves large-scale tests on embedded column base connections with attached reinforcement, and analyzes the effect of common reinforcement detailing for connections under axial compression and cyclic lateral deformation representative of seismic loading. It is observed that the introduction of horizontal reinforcement attached to the embedded column flanges reduces the strength and stiffness of such connections, as compared to cases where no reinforcement is attached. This is because, the horizontal reinforcement introduces a tension field in the concrete area above the uplifting region of the embedded plate, resulting in a reduced vertical resistance and an overall net reduction in strength. Based on observations, a method is developed to predict the strength of Embedded Column Base connections with various detailing features. The method shows good agreement with test data when compared with available strength models. Implications on design, detailed analysis and discussion of limitations of each study are provided.
