PSI - Issue 62

Luca Comegna et al. / Procedia Structural Integrity 62 (2024) 484–491 Luca Comegna/ Structural Integrity Procedia 00 (2019) 000 – 000

489

6

0.1 and 2.5 mm/month with acceleration/deceleration during the wet/dry seasons, that seems again consistent with the CWB 5 trend (Fig. 4b).

a

0.0 0.5 1.0 1.5 2.0 2.5 3.0 0 50 100 150 200 250 300 350 400 450 500 550 600 v av [mm/month] CWB 5,av [mm] b

Fig. 4. Monitoring results: (a) fluctuation of the water pressure head at piezometer P3, h w,P3 , derived through equation (4) from 1991 to 2020, and measured from 2105 to 2019; (b) mean displacement velocity at the ground surface, v av , measured by inclinometer G1 from 1998 to 2003, compared to the Cumulative Water Balance over five months values averaged over the corresponding monitored period, CWB 5,av (modified after Comegna et al., 2023). 4. Analyses of the weather-induced kinematics A simplified numerical model was developed to reconstruct the post-failure kinematics as a function of the weather induced piezometric regime of the slope. The analysis was launched using the 2D finite element code PLAXIS, taking into account the real geometry of the problem and employing a mesh consisting of 1,309 15-noded triangular elements (Fig. 5a). The three lithostratigraphic units were modelled with a linear elastic-perfectly plastic constitutive law, assigning the soil properties reported in Table 2. More specifically, the mechanical properties of the embankment and of the parent formation have been derived from in situ and laboratory tests, while those assigned to the debris have been calibrated through the analyses. The sliding surfaces were simulated using interface elements. The mechanical properties of the interface elements within the embankment and the debris are equal to those of the corresponding crossed soils; a residual friction angle of 20.4° was instead assigned to the 20° inclined planar sliding surface within the substrate. The piezometric regime was simulated assigning the monthly varying positions of the water table from September 1, 1991, to August 31, 2020, accounting for the h w,P3 oscillation (Fig. 4a) determined by equation (4) and accounting for the monitoring evidences described in section 3. To simulate the vehicle load, a constant vertical stress q = 20 kPa, has also been uniformly distributed along the road width.

Table 2. Properties assigned in the FEM numerical analysis to the different lithostratigraphic units: soil unit weight,  ; cohesion, c’ ; friction angle, φ ’ ; Young’s modulus, E’ ; Poisson’s ratio,  .

γ [kN/m 3 ]

 ’ [-]

Lithostratigraphic Unit

φ' [°]

E' [MPa]

c' [kPa]

Embankment Debris cover

18

0 35 0 28 14 30

40 40

0.3 0.3 0.3

20.5

Parent formation Parent formation (interface)

21

200

-

0 20.4 200

0.3

The analysis enabled the assessment of cumulative horizontal displacements over the reference period, which were mostly controlled by the activation of plastic failure points along the interfaces due to the mobilization of the available shear strength associated to the rising water table positions occurring during the wet season. It was therefore possible to reproduce with good approximation the inclinometric profiles (Fig. 5a) and the horizontal surface displacements observed during the different monitoring campaigns (Fig. 5b), consequently filling the gaps in the monitoring data. In