Theoretical and Experimental Analysis of Dissipative Beam-to-Column Joints in Moment Resisting Steel Frames

Book excerpt: Before the seismic events of Northridge (Los Angeles, 17 January 1994) and Hyogoken-Nanbu (Kobe, 17 January 1995), MRFs were supposed to be the most reliable seismic resistant systems due to the high number of dissipative zones that are able to develop. Before these earthquakes, especially in the United States, MRFs were realized, generally, by adopting fully welded connections, which, at the time, were retained to perform better compared to other joint typologies. In addition, the economic advantages deriving from the adoption of field fully welded connections strongly influenced choices of building owners and, as a result, led to the adoption of this joint typology in almost all pre-Northridge steel MRFs. After the Northridge earthquake, even though the loss of life was limited, the unexpected amount of damages occurred in structures adopting as seismic resistant system welded Moment Resisting Frames put into question the role played by welded connections on the whole of structural behavior. Therefore, after the seismic events, two strategies were identified to improve the behavior of fully welded connections. The first one is related to the improvement of the welding technique, usually strengthening the critical area subjected to fracture. The second one is based on the possibility of concentrating the energy dissipation in the beam, reducing the bending resistant area of beams by properly cutting the flanges in a zone close to beam-to-column connection. This weakening approach is commonly called RBS. A new design approach, which has been the subject of many studies in the last few decades, has gained growing interest in recent years. In fact, Eurocode 8 has opened the door to the idea of dissipating the seismic input energy in the connecting elements of beam-to-column joints. In this work, attention is focused on this last approach. The first part of the work is descriptive and deals with the historical development and, in general, with the seismic behavior of Moment Resisting Frames. In the same chapter, general concepts concerning the component method, as introduced by last version of Eurocode 3, are given. Finally, the influence of the joint behaviour on main characteristics of partial strength and/or semi-rigid MRFs is evaluated by properly accounting for existing literature. The third chapter deals with an experimental analysis on the cyclic behaviour of classical partial strength beam-to-column joints. The main scope of the experimental campaign is to show how to control the dissipative behaviour of joints by properly designing the weakest joint component and by over-strengthening the other connecting elements. A design procedure is pointed out and the comparison among the results obtained by cyclic tests is presented in terms of energy dissipation capacity. In addition, by monitoring during the experimental tests both the whole joint and the single joint components it is shown that the energy dissipated by the joint is equal to the sum of the energy dissipated by the joint components. This result assures that the first phase of the component approach, i.e. the component identification, is properly carried out and that interaction between components under cyclic loads is negligible. Chapter 4 represents the extension of the work carried out in the previous chapter. In fact, on the base of the obtained results, the goal is to provide a mechanical cyclic model for the prediction of the overall joint behaviour, starting from existing literature models. Hence, a state-of-the-art review is first presented and then, a model employed to set up a computer program devoted to the prediction of the cyclic behaviour of steel beam-to-column joints is defined. In particular, the proposed cyclic model adopts Kim & Engelhardt's approach to model the shear panel behavior, Cofie & Krawinkler's model to predict Panels in Tension and Compression cyclic behavior, and Piluso et al.'s model for the prediction of the T-stub modelling