PSI - Issue 44

Dario De Domenico et al. / Procedia Structural Integrity 44 (2023) 633–640 Dario De Domenico et al. / Structural Integrity Procedia 00 (2022) 000–000 7 the soil involved by the viaduct piers (Nspt ranging from 45 to 97 and , 30 ranging from 380 m/s to 700 m/s), a ground type B as per Table 3.1 of EC8-1 (2004) was considered. The selected seismic records are used as seismic excitation to perform nonlinear time-history analyses. Elastic flexural and shear stiffness of cracked concrete were assumed one-half of the corresponding values of the uncracked elements (§4.3.1 EC8-1 (2004)), which led to fundamental periods of 1.544 s and 1.186 s in the two principal directions. The resulting peak bending moments are illustrated in Fig. 7 in terms of Bresler domains (bending moment along the two principal directions with fixed axial force) for the single earthquake events and as average values for the two directional combinations ( E x + 0.3 E y and E y + 0.3 E x ). To quantify the modification of the seismic safety margin in the various corrosion scenarios studied in this work, a so-called “bending exploitation ratio” ψ was calculated for the average seismic response from the nonlinear dynamic analyses and reported in Fig. 7: this value represents the geometrical ratio between bending demand and bending capacity points measured along the straight line passing through the origin and having slope defined by the proportion of bending moments in the two principal directions. An estimate of the safety margin can be obtained as the complement to the unity of ψ , as this value indicates the distance between the demand and the capacity threshold. 639

Fig. 7. Results from nonlinear time-history analyses on the bridge piers in the uncorroded and corroded scenarios.

Although an adequate safety margin was noted in an ideal uncorroded scenario (around 35% and 65% in the two principal directions), the corrosion induced degradation of material parameters led to a contraction of the strength interaction domain: the ratios ψ tend to increase from the uncorroded to the corroded case, with increase percentages of around 20% and 38% for corrosion scenario (1) and of around 24% and 38% for corrosion scenario (2) in the two directional combinations, respectively. This implies that the safety margin of the corroded bridge piers decreases of 20-40% when taking into account a corrosion state of steel bars placed along the cortical part of the section with mass losses in percentage of 20-30%, as suggested by experimental findings. 5. Conclusions The seismic vulnerability assessment of RC bridge piers with corroded steel bars has been investigated through an experimental-numerical procedure. The proposed approach has been illustrated in the context of the Zappulla multi-span viaduct, whose RC piers are experiencing chloride-induced corrosion, as confirmed by a comprehensive set of experimental tests performed in situ, including carbonation tests, corrosion potential mapping and extraction of corroded steel bars for tensile tests, in conjunction with SonReb tests and concrete coring. The test results have been critically interpreted to calibrate a numerical fiber-hinge model of the bridge pier, based on which nonlinear time-history analyses under spectrum-compatible natural records have been performed to investigate the influence of corrosion on the seismic behavior of the corroded piers. It has been found that although an adequate safety margin exists in an ideal uncorroded scenario (around 35% and 65% in the two principal directions), the corrosion induced degradation of material parameters leads to a contraction of the strength interaction domain: the seismic safety margin of the pier structures is thus reduced of up to 20-40% by taking into account a mass loss in percentage of the steel bars along the cortical part of the section equal to 20-30%, as suggested by experimental findings. In the