Optimizing Measurement of Fatigue Crack Growth Relationships for Cr-Mo Pressure Vessel Steels in Hydrogen Gas

The objective of this study was to explore an approach for measuring fatigue crack growth rates (da/dN) for Cr-Mo pressure vessel steels in high-pressure hydrogen gas over a broad cyclic stress intensity factor (ΔK) range while limiting test duration, which could serve as an alternative to the method prescribed in ASME BPVC VIII-3, Article KD-10. Fatigue crack growth rates were measured for SA-372 Grade J and 34CrMo4 steels in hydrogen gas as a function of ΔK, load-cycle frequency (f), and gas pressure. The da/dN vs. ΔK relationships measured for the Cr-Mo steels in hydrogen gas at 10 Hz indicate that capturing data at lower ΔK is valuable when these relationships serve as inputs into design-life analyses of hydrogen pressure vessels, since in this ΔK range crack growth rates in hydrogen gas approach rates in air. The da/dN vs. f data measured for the Cr-Mo steels in hydrogen gas at selected constant-ΔK levels demonstrate that crack growth rates at 10 Hz do not represent upper-bound behavior, since da/dN generally increases as f decreases. Consequently, although fatigue crack growth testing at 10 Hz can efficiently measure da/dN over a wide ΔK range, these da/dN vs. ΔK relationships at 10 Hz cannot be considered reliable inputs into design-life analyses. A possible hybrid approach to efficiently establishing the fatigue crack growth rate relationship in hydrogen gas without compromising data quality is to measure the da/dN vs. ΔK relationship at 10 Hz and then apply a correction based on the da/dN vs. f data. The reliability of such a hybrid approach depends on adequacy of the da/dN vs. f data, i.e., the data are measured at appropriate constant-ΔK levels and the data include upper-bound crack growth rates.