Seismic behaviour of steel structures equipped with traditional and innovative beam-to-column connections

Abstract

One of the most common ways to conceive seismic-resistant steel structures is by adopting the Moment Resisting Frames (MRFs). This approach ensures that the building withstands the seismic event through the development of plastic hinges at the beam ends, beam-to-column connections or column bases. The most widespread design philosophy relies on the strong-column strong-connection weak-beam approach, which ensures the development of plastic hinges only at the beam ends and first-floor column bases. Nevertheless, this approach implicitly accepts the development of structural damages during a severe seismic event to dissipate the input energy. This is a negative aspect because it affects the reparability and functionality of buildings.

For this reason, in the last decades, as an alternative to this classic design strategy relying on full- strength joints, a new design philosophy based on the use of partial-strength beam-to-column

connections was developed. This method relies on the strong-column weak-connection strong-beam approach so that the dissipation of the seismic input energy occurs only in well-defined nodal components, which can be easily substituted at the end of the earthquake. In such a way, structural resilience is also achieved. Several traditional and innovative solutions have been proposed and investigated within this framework. These beam-to-column joints have been widely studied based on experimental tests, numerical simulations, and theoretical formulations deriving from adequately defined analytical

models. In particular, the experimental tests and the corresponding simulations have regarded beam- to-column sub-assemblies under monotonic or cyclic loading histories. In such a way, the basic

information related to the analysed joints’ stiffness, resistance, ductility and energy dissipation capacity could be easily derived. Instead, very few tests on large-scale steel structures subjected to seismic inputs have been performed. In this framework, a relevant research programme has been planned at the University of Salerno. It aims at assessing the dynamic behaviour of different beam-to-column connections over the seismic response of large-scale structures. In particular, a significant part of this investigation relates to performing pseudo-dynamic tests on a mock-up building equipped with different traditional and innovative joints: the Reduced Beam Section (RBS or dog-bone) connection; the FREE from DAMage (FREEDAM) joint; the double-split dissipative T-stub (or X-shaped) connection. The configurations mentioned above represent joints connecting double-tee beam and column profiles, reflecting possible American and European applications. However, since there is widespread use in Japan of tubular columns, configurations connecting hollow sections and double-tee profiles should not remain unexplored. Under this perspective, this thesis also focuses on the static

characterisation of joints connecting circular-hollow-section (CHS) columns and through-all double- tee beams by adopting the component method approach. At the moment, the most common way of

conceiving such a kind of joint consists of simply welding the beam to the external surface of the column or using collar plates or composite solutions. However, these alternatives do not ensure relevant mechanical properties and simply structural detailing of the connections. Instead, the recent technological advancements introduced the possibility of using 3D Laser-Cutting for manufacturing the joint mentioned above, whose peculiarity is that the beam can intersect the column, enhancing the mechanical properties but with simple nodal detailing. Therefore, the need to study this connection’s behaviour through the component method approach relies on the possibility of employing this joint together with other solutions (i.e. RBS, ...). However, because of the incompatibility between the profiles of the columns, the seismic response of this connection cannot be investigated through the same mock-up building used to perform the pseudo-dynamic tests. For this reason, at the end of this thesis, a preliminary and brief introduction to the hybrid simulations with dynamic substructuring technique is reported. [...] [edited by Author]