Stephen P. Schneider

Experimental Behavior of Connections to Concrete-filled Steel Tubes

Six large-scale connections were tested to failure using the quasi-static test method. All details consisted of a connection-stub which was shop fabricated and field bolted for construction. Experimental results indicated that welding the connection-stub directly to the skin of the steel tube resulted in a large deformation demand on the tube wall. Large tube wall distortions made the girder flange, the flange weld, and the tube wall highly susceptible to fracture. Inelastic cyclic behavior improved when external diaphragms were used to distribute the flange forces around the tube, and the connection was able to develop the bending strength of the girder. Using embedded elements to distribute the girder flange force to the concrete core was very efficient in alleviating the stress concentration on the tube wall, however, the connection performance was sensitive to the type of embedded elements. Deformed bars welded to the girder flange and embedded into the concrete core developed a connection strength of more than 1.5 times the plastic bending strength of the connected girder, and exhibited stable hysteretic behavior up to failure. Connections with continuous flange plates showed less promise without additional detailing to anchor the flange plate into the concrete core. Extending the girder connection-stub through the entire CFT column was sufficient to develop the full plastic bending strength of the connected girder, and exhibited favorable inelastic cyclic performance.

Analytical behavior of connections to concrete-filled steel tubes

Abstract This paper presents a non-linear 3-D finite element study on a variety of details for connections to concrete-filled steel tubes. An analytical study was needed to verify connections that might be suitable for seismic conditions. This research focused on the connection to circular steel tubes, because this shape presents more detailing difficulties compared to the square tube counterpart. Analytical models were generated using eight-node shell elements for the structural steel components, and 20-node 3-D brick elements for the concrete core. Each element had the capacity for material inelastic and geometric non-linear behavior. To verify the accuracy of the analytical models, results were compared to available test data. Parameters used in this analytical study included: the diameter-to-tube wall thickness ratio, the applied axial load on the column, and the moment-to-shear ratio of the girder. Analytical results suggest that connections which transfer load from the girder to the concrete core potentially offer better seismic performance than connections to the steel tube alone. Connections to the steel tube only exhibit large distortion of the tube wall around the connection region. Components that transfer girder forces into the concrete core exhibit better strength and stiffness characteristics than a simple connection to the tube face. However, the improvement in behavior depends on the type components penetrating the concrete core. Connections in which the girder extends through the composite column offers the most effective method in developing the ideal rigid girder connection behavior.

Experimental study of identification and control of structures using neural network. Part 1: identification

Experimental verifications of a recently developed structural control method using neural network has been carried out on the earthquake simulator at the University of Illinois at Urbana—Champaign. The test specimen was a 1/4 scale model of a three-storey steel frame. The control system consisted of a tendon/pulley system controlled by a single hydraulic actuator. The model structure had a total mass of 2994 kg (6600 lb), distributed evenly among the three floors, and a total frame height of 254 cm (100 inches). The structure had three distinct lightly damped fundamental modes of vibration plus two higher modes representing the structure–control interaction and the actuator dynamics. The system identification and parameter estimation have been conducted in two experimental methods: first, the system has been identified in the time domain and the estimated parameters were used in the frequency domain methods and secondly, the system was modelled and identified using multiple emulator neural networks with different prediction capabilities. These emulators were employed in the control design. This paper describes the test set-up, the experimental validation of the identified model in the time and frequency domains, and experimentally demonstrates the performance of the multiple emulator neural networks. Copyright