PSI - Issue 84

Francesco Cannizzaro et al. / Procedia Structural Integrity 84 (2026) 874–881

879

of 0.65 m and rise to a height of 0.52 m above the extrados of the arches at the crown section. Each pier has a rectangular plan measuring 3.2 m × 7.0 m, a height of 3.59 m, and is equipped with two symmetric circular cutwaters with a radius of 1.6 m. The pier foundations are parallelepiped in shape, with a height of h = 4.17 m and a rectangular base measuring B = 10.89 m by 3.9 m. The complete bridge geometry is reported in Scozzese et al. (2023). A DMEM numerical model was developed to simulate the effects of scour on the bridge. Fig. 3a shows a three dimensional view of the model, in which different colors denote different materials. The model consists of a total of 5,704 elements, corresponding to 42,568 degrees of freedom. The abutments are fully restrained at their base as well as on the rear and lateral surfaces. Conversely, the pier foundations are restrained at the base and along all lateral surfaces, while allowing free displacements along the restrained surfaces and exhibiting elastic stiffness in the orthogonal direction, in accordance with the impedance formulation proposed by Gazetas (1991). The interaction between the fill material and the masonry is modelled using frictional interfaces, whereas cohesive interfaces are introduced at the interfaces between piers and arches and between piers and spandrel walls.

(a)

(b)

Fig. 3. (a) Exploded view of the numerical model and (b) restraint groups associated with the scour hole depth y s .

The mechanical properties of the materials and the stiffness values of the pier foundations, consistent with those adopted in Scozzese et al. (2023), are reported in Tables 1 and 2, respectively. The elastic behaviour of the materials is defined by the Young’s modulus E and the shear modulus G . The nonlinear tensile behaviour of masonry is characterized by the tensile strength f t and the fracture energy G t , and is described by an exponential softening law. The backfill material and the friction interfaces are modelled as elastic–perfectly plastic in tension, with negligible tensile strength, and as elastic in compression. Conversely, the compressive behaviour of masonry follows a parabolic stress–strain law defined by the compressive strength f c and the fracture energy G c . The diagonal shear behaviour is governed by the shear strength  o , according to a Turnsek and Cacovic failure domain. Sliding is assumed to be inhibited for the masonry material, whereas a rigid-plastic behaviour governed by a Mohr-Coulomb yield criterion, defined by the cohesion c and the friction coefficient  , is assumed for the remaining materials. Finally, the unit weight is denoted by w . The model is subjected to the self-weight and to an additional vertical load applied to the top surfaces of the bridge equal to 6.6 kN/m 2 corresponding to the pavement weight. A local scour at one of the two piers is considered. The scour hole is assumed to start from the midpoint of the pier foundation corresponding to the upstream top edge. In Fig. 3b the restraint groups associated with the scour hole depth y s are identified with different colors. Each level of the scour hole depth corresponds to a phase of the analysis where the corresponding restraints are removed from the model and the gravity weights are redistributed to the active elements of the model. 12 levels of scour hole depth are investigated corresponding to values of y s ranging from 100 to 800 cm.