PSI - Issue 78

Riccardo Raimondo Milanesi et al. / Procedia Structural Integrity 78 (2026) 1374–1381

1375

Keywords masonry infills; seismic behaviour; in-plane/out-of-plane interaction; shaking table; experimental test

1. Introduction Masonry infills are a common choice for enclosing reinforced concrete and steel framed buildings. In southern Europe, older “pre - code” infills often consist of thin (12– 15 cm) single or double leaves of highly perforated clay units. While some regions still use this lightweight system, others such as Italy have shifted to thicker (minimum 20 – 25 cm) perforated or lightweight blocks t o improve acoustics, thermal performance and construction efficiency. The thinner panels are generally classified as weak or slender, whereas the thicker ones are regarded as robust or strong. Earthquake surveys in L’Aquila (2009), Lorca (2011), Emilia (2012), Central Italy (2016), Albania (2019) and in Turkey and Syria (2023) consistently report severe damage and even collapse of slender infill panels under seismic loads, causing both economic loss and threats to life (Ricci 2011; De Luca 2014; Manfredi 2014; Masi 2019; Marinković 2022; Binici 2023). Collapse typically f ollows in-plane cracking that undermines out-of-plane stability, so understanding their mutual interaction is crucial to prevent failure. Although numerous static and pseudo-static studies have examined infill behaviour, only a handful have explored how in-plane damage affects subsequent out-of- plane capacity. Early work by Angel et al. (1994) and Calvi and Bolognini (2001) subjected specimens to cyclic in -plane loading before out-of-plane tests, and more recent campaigns have followed suit (da Porto 2013; Morandi 2022). Yet, static out -of-plane tests, whether by line load, point load or uniform pressure, cannot capture crucial seismic dynamics such as inertia effects, crack closure during motion reversals or variable load rates. Dynamic shake-table experiments are therefore essential. While some studies have tested infilled frames or multi storey models under bidirectional shaking (Hashemi 2006; Tu 2010; Baek 2023), almost none have focused on a single-bay, single- storey unreinforced masonry (URM) panel’s out -of-plane response after in-plane damage, exceptions include, for example, Milanesi et al. (2022) on full -scale clay units block infills with sliding joints. To address this gap, the Eucentre Foundation has launched a comprehensive campaign of full-scale tests on slender infills typical of 1960s-80s Italian RC frames. Currently 19 infill specimens have been tested. Steel-concrete composite frames were fitted with full‑scale, single‑story, single‑bay masonry infill panels and tested under a range of conditions which includes fully bonded around the perimeter, with gaps at the top or lateral edges, using 12 cm or 8 cm units, and both with and without openings. For each configuration, panels underwent in‑plane pseudo‑static cyclic loading, dynamic out‑of‑plane shake‑table tests, and combined sequences where they were first loaded in‑plane and then shaken out‑of‑p lane to collapse. This approach revealed how damage evolves under different boundary conditions, unit thicknesses and openings, and allowed the creation of chart out-of-plane resisting force vs in-plane displacement/drift behavior, map damage and acceleration patterns, and observe arching and bending failure modes. The dataset is freely available at https://experiments.builtenvdata.eu/datasets/43/. 2. Design of the specimens and summary of the experimental campaign The full-scale masonry infills were constructed within a structural frame with composite sections made of UPN400 steel elements (steel grade S355) grouted with high-strength concrete, further information about the frames and their in-plane response is reported in Morandi et al. (2025). The structural frame was designed to behave as an existing RC frame structure and, simultaneously, remain undamaged during the tests. The tested infill typology of all the specimens stands for a widely used traditional unreinforced “weak” masonry characterised by horizontally hollowed clay units of 12 cm and 8 cm of thickness. The application of a general-purpose mortar type “M2.5” (nominal compression strength of 2.50 MPa) was considered a suitable choice with respect to common construction practise at that time. All masonry panels have a length of 3.50 m, a height of 2.75 m and a thickness of 1 cm more than the unit thickness due the plaster applied on one side only. The masonry panel is aimed, indeed, to represent one layer of a two layers existing masonry infills, where the external faces are plastered and the internal ones are unplastered. The out-of-plane slenderness λ of the infills is about 21.2 (infill height over thickness with plaster) for the panels made with 12 cm thick units, whereas the slenderness increases to 30.6 for the panels built with the 8 cm thick clay units. Sixteen out of nineteen specimens were built in full adherence with the structural frame by filling the interface joint at the lateral edges with mortar. Meanwhile, the remaing three had a vertical gap of about 2.0 cm between the masonry panel and the structural columns to promote the vertical bending/arching resisting mechanism and avoid any influence of the horizontal/bidirectional bending/arching resisting mechanism. Additionally, four specimens were built with poorly