PSI - Issue 78

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

1378

baseline enabled direct comparison of the infill contribution across the current specimens and any future infill configurations. Target drift levels were set at 0.05 %, 0.10 %, 0.20 %, 0.30 %, 0.40 %, 0.50 %, 0.65 %, 0.75 %, 1.00 %, 1.25 %, and 1.50 %, eac h applied three times in both push and pull directions, the last cycle applied for each specimen is detailed in Table 1. The out-of-plane seismic response of the masonry infills was examined through dynamic shaking-table tests in which specimens (Table 1) were subjected exclusively to out-of-plane excitation until their ultimate capacity was reached. Artificial ground motions were synthesized to match a horizontal Required Response Spectrum (RRS) based on ICC ES AC156 (2020), featuring a spectral plateau between 1.3 Hz and 10 Hz for the majority of the GM, as detailed by Morandi et al. (2025) for Batch 1. Each motion was applied incrementally by increasing the nominal Peak Floor Acceleration (PFA) in 0.10 g steps up to 0.5 g, and thereafter in 0.25 g steps until collapse. Specimens with only vertical boundary restraints were tested using a constant 0.10 g increment throughout. When a panel approached imminent failure, the preceding motion has been repeated sometimes without further acceleration increase. Prior to each ground-motion sequence, white-noise tests were performed to characterize the dynamic properties of the specimen and to fine-tune the shaking-table control. 3. Example of experimental results The in‑plane behaviour of a fully attached masonry infill is exemplified by specimen T1, which was subjected to incremental drifts until reaching ultimate conditions at a nominal drift of δ n = 1.00 %. Although identical procedures apply to all specimens, here T1 serves as a representative case. Damage first appeared at the frame – infill interfaces, with cracks forming along the mortar bed joints. These evolved into pronounced diagonal step cracks that connected to a primary horizontal fissure. With increasing drift, additional cracks developed along mortar joints, further compromising the surrounding plaster and unit shells.

Figure 4. Pictures of the damage state of infills T1 on unplastered (left) and plastered face (right) at the end of the in-plane tests. Cyclic tests on both bare and infilled frames were analysed by plotting applied force against the top‑beam displacement at mid‑height, aligned with the horizontal actuator. The hysteresis loops and peak envelope curves for the infilled frames (Figure 5) are presented alongside the isolated contribution of the infill panel, obtained by subtracting the bare‑frame response from the composite envelope. The backbone curves of infill contribution for Batch 1 (T1, T2 and T3) are reported in Figure 5 reveal consistent initial stiffness across specimens and a similar onset of nonlinearity due to degradation. In T1, maximum drifts of +0.91 % (pull) and −0.95 % (push) were recorded, and strength and stiffness degradation became evident through force reductions of 5 – 25 % between the first and second loading cycles. The bare frame remained elastic throughout the tested drift range. All parameters and observations described for T1 can likewise be determined for the other specimens under the same testing protocol; here they are reported only for T1 as an illustrative example. Figure 6 illustrates the out-of-plane damage patterns through pre- collapse sketches and video frames. Specimen T4, fully connected on all edges and not subjected to prior in-plane loading, displayed remarkable resilience: boundary cracks initiated at 0.1 g PFA, wh ile the first panel crack, a horizontal crack at mid-height, appeared only at 1.0 g. Above the 7th course, a second horizontal crack formed at 1.5 g, followed by two diagonal cracks at 1.8 g. Minor additional damage accrued up to 2.5 g, beyond whic h plaster spalling, unit crushing at the boundaries, and widespread joint cracking intensified, culminating in collapse at 2.75 g via overturning of an upper trapezoidal wedge that lost