PSI - Issue 84

Marco Barla et al. / Procedia Structural Integrity 84 (2026) 441–448

446

4. Numerical modelling All slope interventions, including transitional phases and the presence of construction equipment, were simulated sequentially using 2D FEM SSR analyses in Rocscience RS2, focusing on the most critical cross-section. The numerical model reproduced both the structural measures and the in-situ stratigraphic conditions. Mechanical properties were assigned to each unit based on design documentation. Stress and deformation states were evaluated by simulating the sequential modifications applied to the slope, from the original natural conditions prior to motorway construction. Primary works, including the motorway embankment, the roadside anchored wall, the rockfill wall, and anchored slabs, were modelled progressively through seven computational stages. Then, the current degraded condition of the existing structures was simulated, accounting for complete failure of the upper row of anchors and an 80% reduction in the prestress of the remaining anchors, as well as slope movement triggered by soil degradation and water table rise. Finally, ten consecutive computational stages simulated the implementation of all planned stabilization interventions. Fig. 5 illustrates the main computational scenarios considered.

Fig. 5. Computational scenarios considered in the numerical analyses.

Slope hydraulic conditions were represented using two piezometric lines corresponding to distinct groundwater levels. The deep piezometer, associated with the underlying aquifer, was kept constant across all analyses, while the superficial piezometer, representing the saturation of the silty sand lens due to rainfall events, was varied in geometry and position to simulate different saturation levels. The geometry of the superficial piezometer is a key factor in the numerical scenarios, as field observations and preliminary analyses indicate that partial saturation of the silty sand lens is likely the main trigger and sustaining factor for slope instability. Two saturation scenarios were considered, 20% and 65% of the silty sand lens thickness, so each computational scenario was analyzed twice (except for the first three). The 65% saturation, identified through back-analysis, represents the high water table condition during the intense rainfall between 21 and 26 October 2024, when slope displacements reached 7-10 mm/day, which was assumed to correspond to a factor of safety (FS) of 0.75. The 20% saturation scenario represents a low water table condition following partial drainage towards the deeper aquifer. In addition, a dedicated set of numerical simulations was performed to assess slope stability during excavation and demolition works carried out behind the anchored concrete slabs, explicitly accounting for phased excavation, demolition sequences, or construction equipment loads. These analyses focused on the potential local instabilities associated with shallow slip surfaces that may develop during construction. The simulations followed the demolition