PSI - Issue 75

Hannes Schwarz et al. / Procedia Structural Integrity 75 (2025) 625–632 Schwarz, Fliegener, Rennert / Structural Integrity Procedia (2025)

629

5

Fig. 3. Fatigue dimensioning scheme acc. to FKM guideline (Rennert et al. , 2020) modified for hydrogen influence based on pressure dependent HEI values.

2.3. Validation use case: fatigue strength assessment of demonstrator components As a validation scenario, tube sections from AISI 321 austenitic stainless steel were tested in reference and hydrogen pre-charged conditions under pulsating internal oil pressure (1140 bar maximum pressure, load ratio R = 0.05, frequency f = 1 Hz). The geometry of the tube sections, the test setup and the failure position are shown in Figure 4. Similar as for the specimens presented above, the tube sections were pre-charged with gaseous hydrogen at T = 300 °C and p = 345 bar for 14 days and stored in liquid nitrogen prior to testing. The hydrogen concentration was validated by TDS measurements and a value of c H2 = 55 wppm was obtained (theoretical value acc. to Duportal et al. , 2020 c S = 61 wppm). Eight valid tests (four in reference, four in pre-charged condition) were performed at the same load level of 1140 bar. As pressurization medium, mineral based hydraulic oil was used. From our understanding and as discussed in the literature (Ratoi et al. , 2020), hydrocarbon lubricants act as a physical barrier for hydrogen permeation which leads to a reduced hydrogen effusion and thus to conservative results, since the hydrogen content from pre-charging is preserved. In the fatigue tests, the mean number of cycles up to failure is indeed significantly reduced from 69 ’ 436 (reference samples) to 19 ’ 284 (hydrogen pre-charged samples) at the same load level. Now we apply the modified fatigue strength assessment scheme acc. to FKM guideline (see Section 2.2) to evaluate the applicability at component level under hydrogen conditions. The results are shown in Figure 5 both for as received parts (left diagram) and hydrogen pre-charged parts (right diagram). In both cases, the experimental results (yellow circles) are plotted together with the calculated SN curves from the FKM guideline, considering three different material parameter sets from the manufacturers certificate, literature data (Jürgensen et al. , 2024) and our test data at specimen level, which result in a minor offset of the SN curves. To account for the hydrogen influence, in two cases the HEI from our test campaign of 304L specimens was applied. In one case the HEI was taken directly from literature data (see Table 1). Since direct implementation of the fatigue ratio  H2 /  ref as HEI (as discussed in Section 2.1) would result in an equal or slightly increased fatigue strength under H 2 (  H2 /  ref ≥ 1 at N ref = 10 5 ) for our experimental