PSI - Issue 66

Yamato Abiru et al. / Procedia Structural Integrity 66 (2024) 525–534 Author name / Structural Integrity Procedia 00 (2025) 000–000

527

3

The universal testing machine used in this study operated at a maximum load of 40 kN and a testing speed of 75 kN/min. A Shimadzu servo pulser was used to produce specimens with varying crack lengths. For uncharged specimens, the servo pulser applied a maximum load of 20 kN at a frequency of 10 Hz, while for hydrogen-precharged specimens, a frequency of 1 Hz was used. Compressive loads were applied using blocks measuring 20 mm in length and 10 mm in width. These blocks featured R-machined ends, and a 0.6-mm thick copper plate was placed between each specimen and block to alleviate stress concentration at the contact edges, thus minimizing surface indentations on the specimens. Hydrogen precharging was achieved through immersion in a 40 °C, 20 mass% ammonium thiocyanate solution for 72 h, a process confirmed to yield a saturated hydrogen ingress rate under these conditions, reaching a concentration of approximately 6 ppm. Both ends of the piping specimens were sealed to prevent solution penetration, thus preventing internal corrosion. Hydrogen-precharged specimens were stored at -80 °C until testing to prevent hydrogen loss. Prior to testing, specimens were soaked in ethanol and brought to room temperature. Hydrogen ingress measurements were obtained using a gas chromatography-based thermal desorption analyzer (TDA). Fig . 2 shows the ECT system setup. This test followed the same experimental conditions as in Reference (Abiru et al. 2024). Fig. 2a shows an overview of the nondestructive testing system, and Fig. 2b shows the ECT system. The system used a self-comparison method, placing an excitation coil between two detection coils. The test steel pipe sample was positioned centrally within the coil, with the current set to 0.1 A. During testing, the specimen was moved from 50 mm to 250 mm from its left end along the 300 mm total length, recording the frequency of the excitation coil at each position. Fig. 2c shows the function generator (DSO-X 2002A, Agilent Technologies, Tokyo, Japan) used to convert the current from the ECT system into alternating current for the excitation coil, while Fig. 2d shows the bipolar power supply (PBA20-12, TEXIO, Yokohama, Japan) powering the ECT system. A lock-in amplifier (LI5640, NF Corporation, Yokohama, Japan) applied a sinusoidal reference signal to the input, allowing only signals synchronized with the reference to be detected via a low-pass filter. Both uncharged and hydrogen-precharged specimens, each with a 2 mm crack length (half the wall thickness), were prepared for the ECT testing. Fig. 3 shows the impact inspection system (Abiru et al. 2024). Inspections were performed at both ends of the fixed beam. Fig. 3a shows the impact positions for the hammering test, which were performed under 10 specific conditions: specimens prior to cyclic compression testing, uncharged specimens (at cycles N = 70,000, 120,000, 230,000, and

Fig. 2. ECT device.