Journal of Constructional Steel Research 66 (2010) 1431–1444 Contents lists available at ScienceDirect Journal of Constructional Steel Research journal homepage: www.elsevier.com/locate/jcsr Seismic behaviour of beam-to-column partial-strength joints for steel–concrete composite frames Aurelio Braconi a,* , Ahmed Elamary b , Walter Salvatore c a Corporate Research Policies—RIVA Group S.p.A., RIVA Acciaio–Lesegno Works, Strada Statale 28, I-12076 Lesegno (CN), Italy b Department of Structural Engineering, Al-Azhar University, Egypt c Department of Civil Engineering, University of Pisa, L.go L. Lazzerino 1—I-56123, Italy article info Article history: Received 27 August 2009 Accepted 7 May 2010 Keywords: Steel–concrete composite frames Experimental testing Seismic design Beam-to-column joint abstract The paper presents a large experimental campaign carried out on ten steel–concrete composite beam-to-column sub-assemblages employing monotonic and cyclic loading test protocols. Structural members (beams, columns and slabs) were defined through the design of a full-scale 3D prototype frame subjected to PSD testing campaign; main design hypothesis was to dissipate seismic energy in the joints designed as partial-strength. Testing programme on beam-to-column sub-assemblages was executed in order to assess seismic performance varying structural details at beam-to-column connection level and material qualities. Different mechanical connecting systems between concrete slab and column, two end-plate configurations, weak and strong column web panel, two steel qualities and different concrete strengths. The analysis of sub-assemblages performance was realized in two steps: a first step in which the joints behaviour was assessed, characterizing response at local level (e.g. moment–rotation curves); a second step in which the response was assessed at global level (e.g. force–displacement curves). The structural behaviour (i.e. resistance, plastic deformation and stiffness) was evaluated at three stages – identified as initial, service and maximum load – in order to monitor the evolution of sub-assemblage response increasing solicitation level. Moreover, seismic behaviour of specimens – in terms of dissipated energy, ductility, over-strength and equivalent viscous damping – was also executed. Comparison between experimental results was made in order to identify those parameter suitable for improved and reliable seismic behaviour of steel–concrete composite partial-strength joints. © 2010 Elsevier Ltd. All rights reserved. 1. Introduction Steel Moment Resisting Frames (MRFs) represent one of the most adopted solutions in seismic areas due to their intrinsic ductility [1] based on material properties and on its static scheme. Moreover, MRFs require lower size beams when compared to Braced Frames (BFs), designed employing pinned connections and, at the same time, guarantee higher dissipative capacities through hysteresis cycles more stable and large than cycles produced by braces [1,2]. On the contrary, steel MRFs have more difficulties in the satisfaction of damage limitation for moderate earthquakes respect to BFs. In spite of this, the adoption of a steel–concrete composite solution for MRFs furnishes a higher lateral stiffness, guaranteeing in such a way better performance in the damage control limit state [3,4]: the presence of the concrete slab, acting compositely with the steel beam, ensures a high level of initial * Corresponding author. Tel.: +39 0 174 718 154; mobile: +39 393 32 50 794. E-mail address: ricerca.lunghi@rivagroup.com (A. Braconi). rotational stiffness at the beam-to-column connections [5,6]; the concrete encasement also cooperates in increasing the lateral stiffness. Seismic behaviour of MRF structures strongly depends on gen- eral design approach adopted in the preliminary step: the lo- calization of plastic phenomena necessary for the dissipation of energy at the Ultimate Limit State [7]. In Europe, EN1998-1-1 proposes two approaches: one assumes to localize plastic defor- mation in the beams; the other assumes the localization of plastic deformation in the beam-to-column joint. In the first approach, the whole joint shall have a resistance higher than 1.3 times the plastic resistance of connected beams (full-strength — FS joint), imposing also a high demand on the column. The second approach, on the contrary, assumes the resistance of the joint as an additional pa- rameter (partial-strength — PS joint) that shall be determined on the basis of seismic demand; in such a way, the resistance of joint need not recover the plastic resistance of the beam and strength demand imposed on the column can be relaxed. FS joint solutions require heavy and complex steel details in order to recover beam resistance while PS joint solutions are 0143-974X/$ – see front matter © 2010 Elsevier Ltd. All rights reserved. doi:10.1016/j.jcsr.2010.05.004