Seismic Behavior Of Square Concrete Filled Tubular Column-Steel Beam Composite Frames

Composite steel-concrete structures can make use of the attributes of each material, combining the speed of construction, strength, long-span capability, and light weight of steel with the inherent stiffness, damping, and economy of concrete. In the case of concrete filled tubular (CFT) columns, the tube replaces the formwork during construction, and confines the concrete infill during service, while the concrete infill restrains the local buckling of the steel tube, reducing the construction costs as well as the amount of transverse and longitudinal reinforcement required. Composite concrete filled steel tubular column-steel beam structure has proved very efficient in load capacity and construction, while the complexity of its connection and lack of large scale frame experimental research limit its further application.Previous studies on rigid CFT column-steel beam connection details focused on exterior diaphragm and interior diaphragm connections, stiffening ring connections, beam through column connections and welded or bolted split-tee connections. However, most of those connections need lots of in-situ welding labor work and/or their seismic behaviors still need further study. Prestressed high strength through bolted end-plate connection is widely used in steel structures. Several studies on its application in composite structures, such as steel-reinforced concrete column-steel beam connections and RC column-steel beam connections, have validated its superiority in seismic design. Some but still limited large scale composite CFT frame experimental investigations have been performed, especially on those with through bolted connections.Analytical and experimental studies on the seismic behavior of a composite frame with concrete filled steel tubular columns and steel beams were conducted. A prototype building was designed using current seismic provisions and recommendations from previous related research. The expected performance of the building was related to three seismic hazard levels, namely, the maximum considered earthquake (MCE) level, having a 2% probability of exceedance in 50 years, the design basis earthquake (DBE) level, having a 10% probability of exceedance in 50 years, and the frequently occurring earthquake (FOE) level, having a 50% probability of exceedance in 50 years. The building was expected to remain undamaged for FOE level earthquakes, have repairable damage and exhibit no strength degradation for DBE level earthquakes, and avoid collapse for MCE level earthquakes. A set of limit states was defined representing the degree of damage on the components of the frame, namely the H-shape beams, CFT columns, connections, and panel zones, and the occurrence of these limit states was related to each performance objective.An analytical model for composite concrete filled tubular columns-steel beam frame was developed and used to evaluate the prototype building performance. Models for the individual components of the composite CFT frame developed both by previous research (CFT plastic hinge region model, H-shape beam model, connection model) and as part of this study (a six-node nonlinear spring panel zone model) were utilized. The model accounted for material nonlinearities, including yielding and local buckling of steel and concrete cracking and crushing, as well as geometric nonlinearities. Static pushover and time-history analyses were conducted using the prototype building model to study the response of the building when subjected to a series scaled ground acceleration records representing the FOE, DBE and MCE levels, respectively. The scaling process consisted of matching the response spectrum of each record to the corresponding FOE, DBE or MCE target response spectrum, as defined in related seismic provisions. These simulations were used to evaluate the performance of the prototype building and verify it against the design requirements. The evaluation was carried out in statistical terms due to the variability associated with the seismic excitations. The prototype building was found to perform adequately for all performance objects in the three different earthquake hazard levels.A test structure model was constructed and analyzed with the same consideration of the prototype building model at a four-sevenths scale factor determined from laboratory space and loading capacity and similarity of the strength, stiffness and inelastic behavior between the prototype building members and available members in the market. It was found that the test structure model closely reproduced the response of the prototype building model in terms of element deformations, story drift, and story shear. The overall test structure was divided into two substructures with the bottom two stories to be tested physically while the upper eight stories to be simulated by trilinear model with lumped mass in a test platform NetSLab (Network Structural Laboratories), in which interactions between these two parts were considered. Based on symmetry, a one and a half bay, two-story test frame was built for the physically experimental phase of the study. The test structure was subjected to simulated seismic excitations corresponding to different earthquake hazard levels using sub-structuring pseudo-dynamic hybrid testing methodology. Quasi-static test and pushover test were also conducted to investigate the seismic behavior of the test sub-structure. The test structure performed in accordance with the performance objectives, confirming the conclusions drawn from the results of the analytical studies.Based on the results from the analytical and experimental studies, CFT frame with bolted endplate connections has adequate seismic behavior suitable for seismic resistant design.