PSI - Issue 42

Lisa Claeys et al. / Procedia Structural Integrity 42 (2022) 390–397 Claeys et al./ Structural Integrity Procedia 00 (2019) 000 – 000

395

6

appeared resistant to damage initiation. However, this observation could also be related to the high cracking sensitivity of the grain boundaries leading to stress relaxation once cracks have formed. The significant amount of grain boundary cracking can be related to segregation of manganese and carbon to the grain boundaries leading to local hydrogen accumulation and crack initiation , aided by the impingement of deformation twins and ε -martensite laths on the grain boundaries introducing stress concentrations (Dieudonné et al. (2014); Guo et al. (2014); Koyama et al. (2011)). The depth of the hydrogen-influenced fracture zone was measured on different locations as well and added to Table 2. Multiple influencing factors should be considered. First, the hydrogen concentration charged to the TWIP steel was lower than the value detected in the 304L ASS. Moreover, the hydrogen distribution through the thickness was most likely dissimilar due to a difference in hydrogen diffusivity caused by the altered alloying strategy to stabilize austenite as well as the slightly increased charging temperature for the TWIP steel. (Ismer et al. (2010)) stated, for example, that large Mn fractions led to long-range percolating Mn chains presenting efficient hydrogen diffusion . Moreover, the transformation to α’ -martensite continuously increased the average hydrogen diffusivity during the tensile test for 304L ASS enabling larger distances to be overcome. Finally, to alter the fracture type from ductile microvoid coalescence to either quasi-cleavage or intergranular fracture, a different critical amount of hydrogen might be required. Due to the mentioned differences in initial hydrogen content and distribution before and during the tensile test, this critical hydrogen concentration is, therefore, reached at a specific though different point below the surface for both materials.

a

b

ND

TD

100 µm

100 µm

Fig. 4: SE images of the fracture surface after hydrogen charging for 7 days: (a) hydrogen-induced quasi-cleavage fracture in 304L ASS; (b) hydrogen-induced intergranular fracture in 18Mn-0.6C TWIP steel, where the arrows indicate quasi-cleavage fracture

Table 2. Relevant data for interpretation of constant extension rate tensile tests Material Hydrogen content [wppm] Strain at fracture reference [%] Strain at fracture hydrogen [%]

Depth of H-affected zone [µm]

304L ASS TWIP steel

50.0 ± 2.2 27.5 ± 1.7

87.1 ± 1.5 106.1 ± 1.6

38.5 ± 2.1 61.9 ± 3.7

135 ± 5 87 ± 5

As already observed on the EBSD measurements performed on intermediately tensile tested specimens, many HACs formed on the ND surface. Fig. 5 and Fig. 6 show SE images at different magnifications of the ND surface of fractured specimens for 304L ASS and TWIP steel, respectively. The 304L ASS showed multiple large cracks near the fracture surface due to the strain concentration in the necked region. Significantly smaller cracks were observed over the remaining part of the gauge section. The fracture TWIP steel specimen, on the contrary, showed large cracks over the entire section. Moreover, Fig. 6c clearly shows crack initiation along grain boundaries for the TWIP steel, which can be related to the intergranular fracture surface. Finally, the drop in strength level of the hydrogen-charged TWIP steel compared to the reference condition can be explained by the abundant presence of these large cracks over the entire gauge section reducing the cross-sectional area of the specimen, i.e. a geometrical effect. As these cracks