PSI - Issue 54

Robin Depraetere et al. / Procedia Structural Integrity 54 (2024) 172–179 R. Depraetere et al. / Structural Integrity Procedia 00 (2023) 000–000

178

7

The presence of a hydrogen concentration of 0 . 33 wppm in an API 5L X56 steel clearly results in embrittlement, exhibited through a reduction of the fracture strain (up to 25%, see Figure 3a). Despite the observed embrittlement, no visual di ff erences in the void distribution could be observed in the X-ray micro-CT data. This is in contrast with the results previously reported by the authors on an API 5L X70 steel where a hydrogen concentration of 1 . 09wppm resulted in reductions up to 50% of the fracture strain, together with a significantly a ff ected void distribution [De praetere et al. (2023)]. An exact comparison between both materials’ embrittlement indices is hampered by their di ff erent levels of hydrogen charging. Note that the previous work reported a hydrogen concentration of 0 . 89wppm, since it was measured using hot extraction at 300 ◦ C, whereas 1 . 09 wppm is the equivalent hydrogen concentration when measured using hot extraction at 900 ◦ C. It is important to note that while the void distribution in the fractured specimens lacks noticeable visual or quan tifiable di ff erences, there are variations in the experienced strains between the uncharged and hydrogen charged spec imens (Figure 3a). This indicates that the presence of hydrogen does accelerate the fracture micromechanisms in the API 5L X56 steel, yet it does not alter the overall nature of the mechanisms. Thedi ff erence in fracture behavior in hydrogen between the two pipeline steels can be explained by two hypotheses. First, it is possible that by charging the API 5L X56 steel with an increased hydrogen concentration, the fracture micromechanisms would significantly change in a way which is similar to the results reported for the API 5L X70 steel [Depraetere et al. (2023)]. The concept of a ‘critical hydrogen concentration’ has been postulated, above which the mechanical characteristics are significantly decreased accompanied by a clear change in fracture micromechanisms [Lin et al. (2022); Djukic et al. (2019)]. Second, di ff erences in material characteristics result in di ff erent susceptibilities to hydrogen embrittlement. Even though both pipeline steels feature a characteristic banded microstructure, they have a di ff erent strength and microstructural features, which are reported to influence the behavior of steel in the presence of hydrogen [Laureys et al. (2022)]. The influence of hydrogen on the fracture behavior of a vintage API 5L X56 pipeline steel (produced in 1965) has been investigated through tensile testing of uncharged and hydrogen charged (0 . 33 wppm) samples with various geometries. The damage evolution in a subset of the samples was evaluated after fracture using X-ray micro-CT. This allowed to quantify damage characteristics including void size distribution and void shapes. The relatively old manufacturing method of the pipeline steel resulted in a significant amount of voids upon frac ture. The phenomenon of hydrogen embrittlement was observed, manifested through a loss in ductility for all ge ometries. On the other hand, the void distribution exhibited no significant di ff erences, both in terms of void size and void shape. This indicates that hydrogen slightly accelerates the regular fracture micromechanisms, rather than chang ing the type of mechanisms. These observations are in contrast with results previously reported by the authors for a di ff erent pipeline steel, where increased hydrogen concentrations led to significant alterations in void distribution. These observations may support the concept of a critical hydrogen concentration above which the fracture mech anisms are fundamentally changed. Alternatively, they might also indicate the importance of certain material charac teristics. It is likely that both factors play a role, yet a deeper investigation into the hydrogen-microvoid interactions for various materials and hydrogen concentrations is necessary. 5. Conclusions

Acknowledgements

The authors acknowledge the support from Research Foundation - Flanders (FWO) via grant G056519N. Further more, the authors acknowledge the assistance of Dante Delanote and Viktor Noppe for the design of the tensile test setup, and Iva´n Josipovic and Matthieu Boone from UGCT for their assistance with performing the X-ray micro-CT scans.

References

American Petroleum Institute, 2018. API Specification 5L - Line Pipe. 46.