ADVANCED MATERIALS & PROCESSES | OCTOBER 2024 24 FATIGUE PROPERTY DEGRADATION Although under static tension stress, hydrogen embrittlement rarely led to structure failure in austenitic stainless steels, it must not be interpreted that hydrogen embrittlement can therefore be ignored in these materials. Component/structure failures because of hydrogen embrittlement do happen, even in austenitic stainless steels. The predominant mode of these failures has been premature fatigue fracture. Increased attention has been devoted to understanding the hydrogen effects on fatigue behaviors of austenitic stainless steels[7-15]. Results indicated that hydrogen impact on fatigue behavior of materials is more concerning when compared to its direct effect on ductility under consistent tension. Fatigue behavior can be characterized using different methods. For structures with relatively thick walls, such as hydrogen storage devices and pipelines, crack growth rates are often used to monitor and project hydrogen embrittlement progression. This is because after fatigue crack initiation, crack propagation usually must traverse through a considerable amount of material and may take a substantial length of time. Crack growth estimation can be effective and practical. On the other hand, for thin-wall components such as those in tube systems used in fueling stations and onboard hydrogen vehicles, after fatigue cracks are initiated, there is not much material for cracks to grow and failure usually follows rather quickly. For those applications, correlations of stress and number of cycles to failure, S-N curves, are often used. Similar to hydrogen embrittlement manifested in tension testing, nickel is proved to be beneficial for resisting the adverse effect of hydrogen on fatigue behavior, whether it is fatigue crack growth or cycle life preservation. By varying the contents of chromium and nickel in a group of SUS316L stainless steels, Ohmiya and Fujii[8] studied the effects of both elements on fatigue crack growth in hydrogen. It was observed that crack growth in 90 MPa hydrogen gas slowed down as the nickel content increased. Ultimately, when nickel content reached 11.7%, hydrogen appeared to have very limited effect on fatigue crack growth rate, as shown in Fig. 3. In that same study, it was also observed that although increased chromium showed some benefit on slowing fatigue crack growth, the effect was much less as compared to nickel. Gibbs et al.[12] studied the stress-cycle life behavior (S-N curve) of a 316L stainless steel with nickel slightly above 12 wt%. Results of a strain hardened 316 stainless steel from that research are shown in Fig. 4. Specimens with hydrogen pre-charging to saturation were fatigue tested in air and compared with those not pre-charged but fatigue tested in hydrogen gas of 103 and 10 MPa, respectively. Fatigue cycle number reductions of all three group of specimens were very similar, no matter how the hydrogen effect was introduced to the specimens. Interestingly, the cycle number reduction seemed insensitive to hydrogen gas pressure in 316 stainless steel. Overall, accelerated fatigue failure in hydrogen should not be overlooked. MATERIAL SELECTION As mentioned earlier, austenitic stainless steels in general have very good resistance to hydrogen embrittlement as compared to other materials. In this group, Type 316 stands out as the best. Furthermore, the resistance of 316 to hydrogen embrittlement varies depending on the specific chemical composition, mainly alloy contents. Appropriate material selection can minimize the embrittlement effect of hydrogen. Questions have been raised as to whether certain criteria or methods should be set to help material selection for hydrogen services. Several industrial standards contain different test methods for this purpose. The best known includes ISO 11114-4[16], Fig. 3 — Nickel content effect on fatigue crack growth rates of 316 stainless steels (Source: Ohmiya and Fujii[8]). Fig. 4 — Hydrogen effect on fatigue cycle life of strain hardened 316 stainless steel.
RkJQdWJsaXNoZXIy MTYyMzk3NQ==