PSI - Issue 78

Giulia Giuliani et al. / Procedia Structural Integrity 78 (2026) 952–959

957

4. Nonlinear Time-History Analyses To validate the linear design, nonlinear time-history analyses are performed following the methodology prescribed in Eurocode 8 (EN 1998-1, 2005). Nonlinear models are developed for each case using detailed representations of both the isolation system and the superstructure, based on the geometry obtained from the linear design. The modelling approach follows the strategies proposed by Micozzi et al. (2022). Input ground motions are those listed in Tables 1 – 4, with three-component records selected for each fault distance. This procedure accounts for the nonlinear behaviour of isolation devices and the influence of higher structural modes, resulting in significantly different displacement demands compared to the linear estimates. As a result, the preliminary design was revised to accommodate the observed response trends, and the updated design configurations are summarized in Tables 5 and 6.

Table 5. Results of the isolation system design for the 3-storey building. R JB [Km] D [mm] Hr [mm] S2 n HDRB n FSB 0 928 253 3.67 12 16 5 766 200 3.84 14 14 15 489 132 3.70 22 6 30 355 81 4.39 26 2

Table 6. Results of the isolation system design for the 9-storey building R JB [Km] D [mm] Hr [mm] S2 n HDRB n FSB 0 1136 308 3.69 16 12 5 875 231 3.79 20 8 15 663 179 3.71 28 0 30 467 90 5.19 28 0

For each case study, the mean structural responses obtained from the seven selected ground motion records are computed and utilized to perform the necessary verifications in accordance with Eurocode 8 (EN 1998-1:2005) and the European Standard for Anti-Seismic Devices (EN 15129:2009). Specifically, Figure 5 illustrates the maximum and minimum vertical displacements. Buckling verifications are presented in Figures 6 and 7, considering two critical loading conditions: i) the maximum lateral displacement combined with the corresponding axial force, and ii) the maximum axial force coupled with the associated displacement. The results confirm the adequacy of the isolation system design, as the average axial forces remain below the corresponding buckling capacities. However, uplift phenomena were observed in the flat sliding bearings for scenarios located at 0 km and 5 km from the fault, primarily due to the significant vertical ground motion component. Despite this, such occurrences are deemed acceptable within the context of the design methodology, which is based on mean response values rather than conservative envelopes. This approach aligns with the probabilistic framework adopted in Eurocode 8 and reflects current engineering practice when using realistic ground motion ensembles. Analysis of the stability domains further reveals that, for fault distances of 15 km and 30 km, the governing design criterion is not the maximum horizontal displacement, as observed in the 0 km and 5 km cases, but rather the stability requirements of the isolation system. This highlights a critical shift in the design governing parameters as the fault distance decreases: at closer proximities, the dominant effect of large horizontal displacement demands reduces the relative influence of vertical compressive loads on the design process. Consequently, the isolation system must be primarily dimensioned to accommodate the increased lateral demands, while vertical loads play a secondary role in the verification checks.