PSI - Issue 68

Eduard Navalles et al. / Procedia Structural Integrity 68 (2025) 1105–1114 Eduard Navalles / Structural Integrity Procedia 00 (2025) 000–000

1109

5

the observation that the ductility and reduction in area were more affected for the ferritic-pearlitic steel than for the bainitic steel. The TDMS curves shown in Figure 4 indicate that there are multiple hydrogen peaks. Based on at which temperature these peaks appear one could obtain information if the absorbed hydrogen is weakly or strongly trapped into the steel. Once hydrogen has diffused into the material, it tends to migrate to various defects or within the crystal lattice accumulating at specific sites known as trapping sites. These sites are categorized into two types, depending on the energy required for the hydrogen atom to escape the defect. Weak or low energy trapping sites are defined by an energy level of less than 60 kJ/mol (Verbeken, 2012; Yu et al., 2024), while sites with higher energy levels are identified as deep, strong or high energy trapping sites. They can also be qualitatively distinguished by temperature. A peak is considered a weak trapping site when the hydrogen desorption peak is located below 400 °C, while if the peak is above this temperature, then might be considered as a strong trapping site (Ryu et al., 2012; Yu et al., 2024; Zhao et al., 2022). The hydrogen-tested specimens in Figure 4 have different TDMS curve shapes between the two steels, pointing to variations in the distribution of trapping sites. Considering that the steels have undergone significant deformation during SSRT, dislocations and dislocation cores are also potentially attractive trapping sites for diffusible hydrogen.

Figure 4. TDMS curves to analyse the hydrogen distribution inside of the hollow specimen after test for both steels.

The ferritic-pearlitic steel appears to have at least three hydrogen peaks. The first peak is in the low temperature range, around 250 °C, indicates presence of a weak trapping site such as dislocations, grain boundaries or interstitial sites. The largest peak observed in ferritic-pearlitic steel is located at elevated temperature, around 480 °C. The deep trapping sites in this type of steel are at the interfaces between ferrite and cementite in the pearlite (Li et al., 2024). Under tensile loading, these interfaces may undergo decohesion resulting in voids where hydrogen can accumulate and potentially recombine. An additional smaller peak, observed at the end of the curve, around 600-700°C, may correspond to the presence of different carbides in the microstructure, such as titanium carbides (TiC) (Liu et al., 2024). The bainitic steel has multiple trapping sites that could explain the observed shape of the TDMS curve. In the case of the low temperature peak, similar to those in the ferritic-pearlitic steel, they are likely due to hydrogen trapped in dislocations, grain boundaries, and interstitial sites. Bainitic steel has higher amount of grain boundaries compared to ferritic-pearlitic steel due to smaller grain size (1.8 µm average grain size compared to 5 µm for the ferritic-pearlitic steel). The high temperature trapping sites could be due to the complex microstructure of the bainitic steel. The matrix of globular bainite that typically contains MA islands, retained austenite and high-carbon martensite (H.K.D.H. Bhadeshia, 2015). To reveal the significance and evaluate which microconstituents in both steels that are mainly responsible for the hydrogen/steel interaction seen from Figure 4 further in-depth study is in progress. Up to now the results indicate that hydrogen deteriorate the ductility of the investigated steels, but not the other tensile properties. The reduction in ductility seen for both steels is supported by the TDMS results, which clearly evidence that hydrogen uptake occurred. To investigate further whether hydrogen embrittlement took place a fractographic analysis using SEM was performed. Figure 5 and Figure 6, show overviews of the fracture surfaces as well as higher magnification of the fracture surfaces giving a better view of the difference between argon-tested and hydrogen-tested specimens. In all the high magnification micrographs in Figure 5 a ductile failure can be observed. Figure 5a and Figure 5b reveal a significant