PSI - Issue 62

Silvia Caprili et al. / Procedia Structural Integrity 62 (2024) 355–360 Caprili et al./ Structural Integrity Procedia 00 (2019) 000 – 000

358

4

satisfy the prescription of aforementioned code about the minimum number of bending/re-bending to exploit. Finally, X-ray tomographies were executed on about 80 strands, selected on the basis of the results of mechanical tests: the selection was necessary due to both economic and practical reasons, being not possible to perform these investigations on the whole sample. Tomographies were executed on a portion of strand of about 20 cm in correspondence of the failure section; information about pits distribution and geometry (depth, length, width) along the wires were achieved (Fig. 3).

(a)

(b)

Fig. 3. Some results of X- Ray tomographies: a) ‘wire by wire’ failure modality; b) ‘bird cage’ modality.

As general consideration, the entity of pits, as depth, length, etc., was not very marked if compared with data reported in literature (just to give an idea, the maximum value of pit depth was equal to 0.40 mm); pits were rather uniformly distributed along the length of the wire. This typology of deterioration is well-corresponding to the HE mechanism, where even the presence of very small pit can be responsible for the activation of the typical brittle performance with crack propagation in the drawing direction after initiation. This is the reason why a direct correlation between the geometry of pit and the corresponding mechanical performance (residual) cannot be established for HE (considering for instance the pit depth or the mass loss as corrosion indicator for predicting a corresponding deterioration of the mechanical performance). More in-depth investigations, regarding e.g the total pit volume related to the overall ability to absorb Hydrogen, are needed and are under execution. 3. Modelling and analysis of uncorroded and corroded strands With the aim of calibrating the experimental behaviour of the strands and to elaborate reliable models to predict the decrease of their mechanical characteristics, numerical analyses are ongoing. In the current literature, different kind of models are available to reproduce the behaviour of the strands; among them, Franceschini et al (2022, 2023) and Jeon et al (2019) modelled the strand as composed by springs representing each wire: this approach, that is part of so-called ‘ spring models ’, allowed to evaluate the behaviour of the strand as the combination of the behaviour of each wire. Other approaches, such as the ‘ equivalent bar models ’ describe the strands as an equivalent bar (Wang et al , 2020) with reduced cross-section as function of pitting corrosion. In the present work detailed three-dimensional solid models of the strands were elaborated; basing on the results of X ray tomography, pits were exactly reproduced in terms of depth, width and length with the aim of calibrating the mechanical performance using the results of the experimental tests performed on the corroded strands. Of course, the general idea finally aims to elaborate a simplified mechanical model helping engineers in predicting corrosion impact on the mechanical performance of damaged strands (and wires). Remember, in this context, that in case of HE the decrease of the bearing performance is not related to the cross-section reduction but to the material embrittlement, that should be opportunely considered.