PSI - Issue 78

Andrea Dhima et al. / Procedia Structural Integrity 78 (2026) 1366–1373

1369

are the primary structural elements that absorb most of the seismic energy, whilst the superstructure is subjected to lower stress levels. Following on-site inspections, moderate to severe corrosion was identified affecting the structure. To account for the effect of corrosion-induced degradation on the structural response, two approaches were adopted. For traffic loads, the structural capacity was assessed by considering reductions in both steel and concrete strength due to corrosion, as further detailed in Section 3, in order to estimate the resulting decrease in the moment-axial force (M-N) interaction domains. For the seismic assessment, two different non-linear numerical models were developed, focusing on the structural response of the piers. The first model, consistent with the anticipated brittle shear failure mode, is based on a lumped plasticity approach incorporating shear hinges. The capacity of these hinges was determined from the shear strength of transversely reinforced members, as specified in the Italian standard (NTC 2018). The constitutive law governing their hysteretic behaviour was defined in accordance with (FEMA 440, 2005). The second model, was implemented using a non-linear force-based beam element with distributed plasticity, following the formulation proposed by (Spacone et al., 1996) and hereafter called fibre model. Both models accounted for the effects of corrosion-induced degradation by applying a reduction in the mechanical parameters of the concrete and the reinforcement bars. The key parameters adopted for the two different modelling approaches are reported in Table 2. Table 2. Initial values of the mechanical parameters utilised for the non-linear modelling of the viaduct. Shear hinges model Fibre model FEMA 440 (2005) Park et al. model (1982) Menegotto and Pinto model (1973) Parameter Unit Parameter Unit Unconfined Confined Parameter Unit Fy [kN] 1570.48 K [-] 1.00 1.08 fy [MPa] 230 G [MPa] 14198.75 fc [MPa] 17.42 18.29 E [MPa] 200000 As [mm 2 ] 3433333 0 [%] 0.002 0.0020 b [-] 0.005 1 [%] 0.0034 0.0060 [%] 0.0035 0.0070 3. Modelling corrosion mechanisms A critical aspect in the analysis of structural deterioration processes is the modelling of steel reinforcement corrosion. The onset of the corrosive process initially causes a reduction in the cross-sectional area of the reinforcement bars. The reduction can be uniform, when associated with carbonation-induced degradation, or localised (a phenomenon known as pitting) when caused by chloride attack. One of the most widely used models to describe uniform corrosion is that proposed in the RILEM reports (Sarja et al., 1996), whereas for pitting, a key reference model is that of Val and Melchers (Val et al., 1998). Both models are highly dependent on the corrosion current density i corr , which in turn is a function of environmental conditions, the type of physical-chemical attack, the properties of the steel, and the crack width. Several experimental campaigns have been conducted by various researchers to assess the influence of corrosion on pre-deformed steel bars. The formulation to determine the corroded yield strength of the steel, f y * , is as follows: ( ) * 1 [%] f CR f y s y  = −  (1) where f y denotes the uncorroded yield stress, β s is an experimentally derived coefficient, and CR [%] refers to the percentage of corroded mass. To account for the deterioration of the passive reinforcement bars, the β s factors proposed by (Imperatore et al., 2017) were employed: 0.0151 for uniform corrosion and 0.0199 for pitting. The corrosion of reinforcement bars also affects the mechanical properties of the concrete. Preliminary studies by (Vecchio & Collins, 1986) have shown that the compressive strength undergoes a significant reduction when tensile strains sufficient to induce cracking are present orthogonal to the direction of compression. More recently, (Coronelli & Gambarova, 2004) extended this model to incorporate the direct effects of corrosion, proposing that the loss of concrete strength β stems from the expansive pressure exerted by corrosion products on the reinforcement: