The assessment of existing reinforced concrete (RC) buildings, as well as the design of new RC structures is strongly affected by the correct evaluation of the shear capacity of RC members. For example, the RC elements of buildings designed in accordance with outdated seismic codes frequently exhibit insufficient transverse reinforcement, rendering them vulnerable to shear failure. Consequently, the evaluation of their seismic vulnerability is significantly influenced by the choice of shear capacity model employed in numerical analyses. Another pertinent example involves the application of capacity design principles to newly constructed RC buildings: in these cases, the adoption of a reliable shear capacity model is crucial to guarantee that, under the design seismic event, ductile flexural failure mechanisms develop prior to the onset of brittle shear failures. In this context, particular attention should be given to RC columns subjected to biaxial shear loading. Such loading conditions are very common in RC members, for instance in columns experiencing lateral forces from orthogonal directions—due to seismic or wind actions—or in spandrel beams supporting inclined roof structures. Nonetheless, experimental research on RC elements subjected to biaxial shear loading remains scarce, and the development of analytical models for the reliable prediction of their biaxial shear capacity has been undertaken only to a limited extent. In most practical cases, biaxial shear capacity is estimated by means of an empirical quadratic interaction domain connecting the uniaxial shear capacities associated with two principal directions of the cross section, without any clear analytical basis and without being supported by any resisting mechanism.

This contribution aims to fill this research gap by means of an experimental and analytical investigation into the biaxial shear capacity of RC columns. A set of full-scale RC columns having either square or rectangular cross section and featured by two different amounts of transverse reinforcement have been tested under constant axial compression and oblique cyclic lateral loading with respect to the principal axes of the cross-section. This experimental campaign has enabled us to investigate the influence of biaxial loading conditions not only on the shear capacity, but also on the overall cyclic behavior up to ultimate conditions, including damage patterns, strain distribution, hysteretic cycles, displacement ductility, stiffness and energy dissipation. Following these findings, a purely analytical model for the biaxial shear capacity prediction of RC members has been developed. The proposed model is founded on an appropriate extension of the variable-angle truss mechanism, used in the Eurocodes (and in Italy) for determining the uniaxial shear capacity, and herein adapted to account for biaxial shear effects. This adaptation incorporates suitable modifications to both the concrete contribution—associated with strut crushing—and the steel contribution—related to stirrup yielding. The very good predictive accuracy of the proposed model has been demonstrated by comparison against experimental results of this experimental campaign as well as other independent biaxial test results collected from the literature.