PSI - Issue 84

Lorenzo Brezzi et al. / Procedia Structural Integrity 84 (2026) 1159–1166

1165

broader interaction zone, extending over more than 40 m, thereby distributing dynamic actions over a larger portion of the structure. This widening of the interaction area, rather than an increase in local peak velocities, represents the primary effect of the multiple-source configuration on the dynamic interaction with the viaduct.

Fig. 6. Average values of the impact forces, with focus on the viaduct area, considering all the 200 simulations and the two scenarios.

The flow thicknesses and velocities extracted at the instant of maximum interaction were combined to estimate the potential impact forces acting on the viaduct. Fig. 6 summarises the spatial distribution of the estimated impact forces for the two considered scenarios, expressed in terms of mean values and associated variability across the ensemble of Monte Carlo simulations. In the first scenario, impact forces reach values ranging from 493.2 ± 6.8 kN to 739.1 ± 10.9 kN and are concentrated over a relatively limited portion of the structure. This spatially confined distribution is consistent with the more localised runout pattern and the reduced variability observed for both thickness and velocity fields. These magnitudes indicate that even a single-source rapid landslide may generate actions of the order of several hundreds of kilonewtons on viaduct substructures. In the double-landslide scenario, force values span a wider range, from 99.6 ± 21.2 kN up to 763.3 ± 22.8 kN, distributed over a substantially larger portion of the infrastructure. The highest values appear more spatially concentrated and shifted in a preferential direction, reflecting the interaction and partial superposition of the two landslide masses prior to impact. A more direct quantitative comparison can be obtained by considering the locations of maximum impact force in the two scenarios, identified as points A and B in Fig. 6. In the single-landslide scenario, the mean impact force at point A reaches 663.5 ± 10.3 kN, while in the double-landslide scenario the corresponding maximum at point B increases to 763.3 ± 22.8 kN. This represents an average increase of approximately 15% in peak impact force associated with the multiple-source configuration. Despite this moderate increase in local peak values, Fig. 6 clearly shows that the most significant effect of the double-landslide scenario lies in the spatial extension of the impact zone. While the single-landslide scenario produces a concentrated interaction over a limited length of the viaduct, the double landslide scenario enlarges the zone of impact, distributing dynamic actions over a substantially wider portion of the structure. This increased spatial extent of interaction implies a higher level of event–infrastructure interaction, even in the absence of a pronounced amplification of local peak forces. 5. Discussion and conclusion The results presented in this study provide quantitative insight into the indirect interaction between rapid debris flow phenomena and viaduct infrastructure under conditions characterised by large deformations and significant parametric uncertainty. The analyses show that impact forces of the order of several hundreds of kilonewtons may act on the structure even under relatively simple single-source scenarios, confirming that rapid landslides represent a non negligible hazard for existing viaduct infrastructure.