PSI - Issue 84

Luca Vené et al. / Procedia Structural Integrity 84 (2026) 529–535

531

The geometry of the local scour was defined using a simplified configuration consistent with idealized models available in the literature while still reproducing the essential morphological characteristics (Diab et al., 2010; Yang et al., 2020). The upstream slope was assigned a gradient equal to the soil friction angle of 36°, whereas the downstream slope was taken as half that value. For the lateral sides, an intermediate slope equal to three-quarters of the friction angle was adopted. The bottom of the scour cavity was assumed to be flat, with the beginning of the inclined surfaces placed 0.5 m from the pile cap edge. This simplification ensures a regular geometry while realistically representing the transition between the cavity floor and the sloped surfaces 2.2. Hydraulic Characteristics of the Malone Torrent The hydraulic characteristics of the Malone Torrent were obtained from documentation provided by the infrastructure manager. The equivalent roughness of the riverbed was determined using the formula proposed by Müller, which allows deriving the D 90 of the sediment from the Gauckler–Strickler coefficient reported in the available documentation. The simulations were carried out considering hydraulic conditions corresponding to a 200-year return period. The parameters adopted in the hydraulic simulations are as follows: • v : 3.05 m/s; • h : 3.35 m; • k r : 0.227 m; • k s : 0.002 m. The bridge superstructure was excluded from the simulations to isolate the hydraulic effect induced solely by the presence of the pier on the pressure distribution along the riverbed. 3. Methodology The adopted methodology is articulated into three main phases: the hydrodynamic analysis through CFD simulations, the transfer of the obtained pressures to the geotechnical model, and the subsequent simulation of seepage flow within the soil. Each phase is described in the following sections. 3.1. Hydraulic Simulations (CFD) The hydrodynamic simulations were carried out in transient mode using ANSYS Fluent, based on the Finite Volume Method (FVM). The computations were run until a quasi-steady condition was reached, in which the free surface profile no longer exhibited significant temporal variations and the hydraulic variables stabilized in time. The analyses were therefore interpreted under steady-state conditions of the flow field. A free-surface approach was adopted, explicitly accounting for water–air interaction in order to obtain a realistic representation of the surface flow. Turbulence was modeled using the k– ω SST model, which combines the advantages of the k– ω formulation in the near-wall regions with those of the k– ε model in the outer flow. For the geometric configuration, only the central pier of the bridge was represented, including half-span segments on both the left- and right-hand sides. It was assumed that each pier produces the same hydraulic effect; therefore, laterally smooth surfaces were imposed as symmetry boundary conditions. 3.2. Transfer of Pressures to the Geotechnical Model The total pressure values at the riverbed, obtained from ANSYS Fluent, were extracted and interpolated onto the nodes of the geotechnical model grid developed in FLAC3D, using a Python script. This procedure enabled the accurate transposition of hydraulic pressures to the geotechnical domain, where they were applied as boundary conditions for the simulation of seepage flow within the soil.