PSI - Issue 77

A. Hell et al. / Procedia Structural Integrity 77 (2026) 41–48 Author name / Structural Integrity Procedia 00 (2026) 000–000

43 3

samples with lever scissors, cleaned with 2-propanol, dried and melted in a graphite crucible in an inert carrier gas atmosphere. The total hydrogen content of the specimens was measured. As IGF is sensitive to surface contamination, only samples without visible corrosion attack after charging were investigated. 2.3. Diffusion calculations From the hydrogen uptake measurements, the time needed for charging of the reference samples was extrapolated and the estimation of diffusivity was correlated with typical values in ferritic-pearlitic microstructures. A hydrogen concentration profile c H , based on Fick’s diffusion laws was calculated for the compact tension specimens for the 96 hour duration of charging. Numerical solving of the equations was performed in COMSOL multiphysics software with the idealized boundary condition of full saturation at the specimen surface ( c surface = const.). The simulation didn’t take into account any electrochemical or microstructural effects leading to local fluctuations in hydrogen ingress. Furthermore stress-assisted diffusion towards the crack tip during mechanical testing as well as hydrogen desorption weren’t considered in this first simple modelling approach. 2.4. Precracking and J- ∆a -curve measurement The crack growth resistance is evaluated using J- ∆a -curve determination. Sinusoidal fatigue cycling of hydrogen charged and uncharged CT50 samples is performed with R = 0.1 in a Schenck PC160M servohydraulic testing machine. The targeted fatigue crack length was 4.5 mm + starter notch, corresponding to 0.5* W . To minimize hydrogen effusion before the J - ∆ a -measurement, the aim was to conduct rapid precracking. Two different strategies were applied, both leading to defined fatigue cracks after approximately 30-45 min of cyclic loading. The first experiments were conducted with a constant maximum force of 27 kN and a testing frequency of 5 Hz, before a second methodology with F max = 19 kN and f = 10 Hz was developed. Subsequently, a Kappa100 DS testing machine from ZwickRoell was used for quasi-static loading of the cracked samples with a crosshead speed of 0.1 mm/min. The load line displacement was monitored using videoextensometry with an accuracy of 0.1 µm and allowed the calculation of stable crack extension using the compliance method. ∆ a and J-integral values were obtained via an approach based on ASTM E1820, which was implemented in a Matlab code. 2.5. Optical Microscopy and SEM fractography Crack fronts resulting from quasi-static fracture mechanical testing were investigated in a LEXT Laser Scanning Microscope OLS 4100 from Olympus. The microscopy samples were machined from the center section of CT50 specimens using wire-cut Electric Discharge Machining (EDM), hot mounted, wet grinded and polished to a grit size of 1 µm. Fracture surfaces from ductile and fatigue fracture were characterized in a JEOL JSM 7900F/XRM-II nanoCT device in order to search for characteristic hydrogen-related microstructural damage features like promoted intergranular failure, quasi-cleavage or a change in dimple size and shape. The fatigue crack surfaces were investigated at 0.5, 2.0 and 4.0 mm crack length. Ductile fracture was examined at the region of maximum crack extension in J - ∆ a -testing over the whole width of the sample. 3. Results 3.1. Characterization of hydrogen uptake The hydrogen uptake of reference sheet samples is shown in Fig. 1. The plot shows the measurement of total hydrogen concentrations c H,total in weight-ppm (wt.-ppm) using inert gas fusion and an exponential fit. A quick increase in hydrogen concentration is visible in the first charging hours, although the hydrogen uptake of the different samples varies considerably, possibly due to charging process fluctuations. In uncharged samples, a residual hydrogen content of 0.39 ppm was measured, either residing in the material or due to remaining surface contamination. Taking into account the residual hydrogen, the maximum H-concentrations obtained with the charging setup are in the order of