PSI - Issue 78

Marta Bertassi et al. / Procedia Structural Integrity 78 (2026) 1521–1528

1527

Fig. 6. Comparison between NLTHA and NLKA predictions by NTC 2018 and FprEN 1998: (a) wall W3 ( h = 4 m , t = 0.12 m); and (b) wall W4 ( h = 4 m , t = 0.24 m) .

In most of the examined cases shown in Figs. 5 and 6, the NLKA approach was found to be conservative, both in estimating displacement demand and in assessing the attainment of the collapse limit state, thus leading to a cautious assessment of OOP overturning safety. Exceptions were observed in certain configurations involving particularly slender and heavily loaded walls such as W1 ( h =3m, t = 0.12 m) and W3 ( h = 3 m, t = 0.24 m) under σ equal to 0.20 and 0.40 MPa. Under these conditions, the NTC 2018 kinematic procedure significantly underestimated displacement demands and failed to predict exceedance of the collapse limit state, which was instead captured by the NLTHAs. Moreover, from its formulation — see Eq. (1) and (2) — the FprEN 1998 approach generally results in higher displacement demands than NTC 2018. 6. Conclusion This paper presents a comparison between the results of simplified NLKAs, as defined by the Italian building code (NTC 2018) and the second-generation Eurocode 8 provisions (FprEN 1998), with those from NLTHAs using SDOF models. The comparison focuses on four case-study URM walls subjected to OOP excitation that induces local vertical bending mechanisms. Parametric analyses were carried out considering variations in wall height, thickness, vertical load, seismic input, and assumptions regarding global building behaviour. Initially, linear and nonlinear dynamic analyses were performed on SDOF oscillators to simulate the global behaviour of masonry buildings. These analyses were used to generate floor-level accelerograms for multi-storey configurations, based on the assumption that the building responds in its fundamental vibration mode. The resulting accelerograms were employed to assess the accuracy of code-based formulations for floor response spectra (Section 5.1), under the assumption of global elastic response. In this case, NTC 2018 slightly underestimates, while FprEN 1998 slightly overestimates the peak spectral ordinates. When the global response is nonlinear, both code formulations tend to overestimate floor spectral demands, especially near the global elastic fundamental period. Notably, for periods