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

174

3

L

R1.2

R6

R2

⌀ 6

l 0 = 40

100

35

⌀ 6

l 0 / 2

l 0 / 2

l 0 / 2

R ∞

R6

R2 R1.2

(a)

(b)

Fig. 1: (a) The microstructure at mid-thickness in the longitudinal-to-rolling (L), transverse-to-rolling (T) and through-thickness (S) direction. (b) Geometries of the specimens extracted from the pipe.

di ff erent geometries (R ∞ , R6, R2 and R1.2), di ff erent levels of positive stress triaxialities are obtained. By employing two notches, both the states of complete fracture and maximum load can be assessed with a single test [Kim et al. (2018)]. Prior to tensile testing, a portion of the specimens were precharged with hydrogen using the electrochemi cal method. The specimens were submerged for 6 hours in an electrolyte consisting of 0 . 1 M NaOH with 1 g / l NH 4 SCN, and a current density of 0 . 8mA / cm 2 was applied. This resulted in a nominal hydrogen concentration of C L , 0 = 0 . 33 ± 0 . 05 wppm (average ± standard deviation), as measured with hot extraction at 900 ◦ C. The specimens were mounted in a universal test rig after hydrogen charging. A fixed displacement rate of 2 . 5 × 10 − 4 s − 1 was applied, defined as the ratio of the piston displacement to the total length of the reduced section(s) l 0 . During the tensile tests, the lateral contractions were optically monitored using the technique described in Depraetere et al. (2023). This allows to determine the true strain ϵ t as: ϵ t = ln( A 0 / A ) (1) where A 0 is the initial minimal cross-sectional area, and A the instantaneous minimal cross-sectional area. The final cross-sectional area A f was measured post-mortem on the fracture surface and allows to calculate the fracture strain ϵ f as: ϵ f = ln( A 0 / A f ) (2)

2.3. High resolution X-ray computed tomography

A subset of the fractured specimens was scanned using high resolution X-ray computed tomography (X-ray micro CT) to quantify the damage after deformation. Details of the scanning procedure are reported in Depraetere et al. (2023). For each geometry, one uncharged and one hydrogen charged specimen were scanned. Grayscale thresholding was carried out to distinguish voids from the material. The size of each individual void was characterized by its volume V , which can be transformed into an equivalent diameter of a sphere with the same volume as:

3 6 V π

D eq =

(3)