PSI - Issue 54

Mihaela Iordachescu et al. / Procedia Structural Integrity 54 (2024) 52–58 Mihaela Iordachescu et. al / Structural Integrity Procedia 00 (2023) 000–000

56

5

Eq (3) particularized for K = 31 MPa·m ½ and combined with Eq (1) determines the maximum crack depth a ssy in the SSY regime for given values of COD. As a consequence, the crack depths values found from Eq (1) and the values of F and COD measured along the tests are valid predictions if they are below the a ssy value corresponding to the measured COD. Fig. 2b shows the a ssy and K values resulting from the F and COD values of the loading process in FIP medium shown in Fig.2a. As stated, only the K values below 31 MPa·m ½ validly describe the stress singularity at the crack front. The measured values of the crack size, both on the fracture surface and during testing through the VIC-2D captured images on the lateral face of the specimen, have also been plotted in Fig. 2b and indicate that the transition from fatigue cracking to environmentally assisted cracking occurs in the SSY regime. The measurements of the crack depth and the recorded loads show that the bearing capacity of the specimen yields and stabilizes at the end of the SSY regime (Fig. 2a) with a subsequent decrease as assisted cracking progresses. This process is consistent with the fractographic characteristics of the assisted cracking surface which can be seen in Fig. 3. The irregular morphology of the transition zone blurs its boundaries and complicates the assignment of a crack depth different from that of the fatigue pre-cracking (Fig. 3a).

Fig. 3. Fracture morphology of a SENT specimen broken in SSRT-FIP: a) general view; b) in the yielding stage; c) during stable loading

According to previous published research performed by Iordachescu et al. (2022), the size of the microstructural components whose failure leads to cracking growth is similar in both the yield and stable loading stages of tests in air and in the SSRT-FIP tests continued until failure. This micro-structural failure is due to hydrogen action and occurs by the weakening and detachment of the boundaries between adjacent martensitic packages. The fracture morphology exhibits a significant difference in the package boundaries through which the crack has propagated in the yielding (Fig. 3c) and stable loading stages (Fig. 3b). This difference reflects whether the boundary originated in that of a previous austenitic grain or not, since the singularities contained in the first increase their capacity as sinks for the hydrogen attracted towards the crack front because of the local stress concentration. The difference ceases to be determinant as the crack grows and the effect of the singularities on hydrogen uptake is dominated by that of the stress concentration. The stable loading stage takes place when assisted cracking occurs at a rate such that it provides all the elongation applied to the specimen by the machine actuator, with no additional load (Fig. 2a and Fig. 3b). 4. Damage tolerance analysis The results of the assisted cracking and of the fracture tests carried out in FIP environment, as well as those of the fracture tests in air of SENT specimens with distinct initial damage, have been represented in the damage tolerance diagram given in Fig. 4, whose ordinate F/F 0 and abscissa A f /A 0 are relative values of the failure load, F and of the cracked area, A f , respectively. The corresponding reference values F 0 = R m BW and A 0 = BW are the bearing capacity and the cross-section of specimen (Fig. 1b) in the absence of damage, R m being the tensile strength R m of