PSI - Issue 84

Alessandro Nettis et al. / Procedia Structural Integrity 84 (2026) 653–660

658

( ) = = 2 − { + ( ) = = √ 2 − 2 − , − ≤ 1 ( ) = 2 ( )<0 − ( ) = = −√ 2 − 2 − , − 2 < 1 ( ) = 2 ( ) < −

{

(8)

It is worth noting that in Equation (8) the circumference is represented alternatively by its upper or lower half, depending on the location of the intersections along the y-axis. The area resulting from the intersection between the parabolas bundle and the wire circle can be computed through Equation (9). ∩ ( ) = |∫ ( ( ) − ( , )) 2 ( ) 1 ( ) | (9) The area computed following this equation does not always represent the pitting area. In case ( ) represent the lower half of the wire circle, ∩ directly computes the wire remaining area. Hence, the area loss due to corrosion pitting can be computed according to Equation (10). { ( ) = ∩ ( ), ( ) = + ( ) ( ) = − ∩ ( ), ( ) = − ( ) (10) Finally, for each wire in the database, the parameter of the parabolic pit can be computed by Equation (11). ( ) = , (11) 4. Application The methodology explained in the previous section is herein applied to the abovementioned database on naturally corroded strands (Vecchi et al., 2021). The database consists of wire-by-wire data relating to 24 7-wires prestressing strands. For each strand in the database the central wire is assumed as uncorroded, as experimental evidence shows that the corrosion process cannot penetrate deep into the strand and attack the inner wire. In addition, no corrosion or negligible corrosion were found on five strands. As a result, 114 wires out of 144 tested wires provided valuable data. Following the methodology described in Section 3, a wire-by-wire dataset of parameters is obtained. A validation of the obtained dataset has been performed, by classifying the obtained parabolic pits according to the discrete classification proposed by Jeon et al. (Jeon et al., 2019). For validation, a generic pit presenting a parabola opening ranging from -0.25 to 0.25 has been classified as a Type III pit; values outside this range have been classified as Type I and Type II, for greater and lower values, respectively. Each pit has been classified in the three types and the final results coincide with the classification provided by Vecchi et al. (Vecchi et al., 2021) for the 98% of the detected pits. Hence, the obtained parabolic pits almost perfectly follow the pit classification currently available in literature. In Figure 3, there are three examples of parabolas corresponding to the three different types.

Figure 3. Corrosion pit morphologies obtained with the proposed model: a) Type I, b) Type II, c) Type III. As for the statistical distributions that best fit the obtained data, different options have been tried. First, statistical distributions like the gamma or the lognormal have been excluded since the domain of the parameter includes both positive and negative values. Given the asymmetry of the obtained dataset, symmetric distributions like the normal or the generalised normal have been excluded as well. Among all, two statistical distributions have provided the best fit