Hydrogen Embrittlement Properties of Stainless and Low Alloy Steels in High Pressure Gaseous Hydrogen Environment

Home , Hydrogen embrittlement

ISIJ International, Vol. 52 (2012), No. 2, pp. 234–239 Review Hydrogen Embrittlement Properties of Stainless and Low Alloy Steels in High Pressure Gaseous Hydrogen Environment

Tomohiko OMURA and Jun NAKAMURA

Corporate Research and Development Laboratories, Sumitomo Metal Industries, Ltd., Amagasaki, 660-0891 Japan. E-mail: [email protected] (Received on June 30, 2011; accepted on September 28, 2011)

Recent research on Hydrogen Environment Embrittlement (HEE) susceptibility of stainless and low alloy steels in highly pressurized gaseous hydrogen environments was reviewed from the viewpoint of tensile properties, hydrogen absorption and fatigue properties. HEE susceptibility evaluated by Slow Strain Rate Test (SSRT) in high pressure hydrogen environments strongly depended on steel chemical compositions. Austenitic stainless steels such as type 316L or iron- based superalloy as A286 showed sufficient resistance to HEE, while stainless steels with low levels of alloying elements such as type 304L showed a remarkable ductility loss in high pressure gaseous hydrogen due to martensitic transformation. Martensitic stainless or low alloy steels also showed a remarkable ductility loss in gaseous hydrogen. Relationship between HEE susceptibility and an amount of hydrogen absorption was investigated. HEE susceptibility and hydrogen embrittlement under cathodic charging in aqueous solution showed the same dependence on the amount of hydrogen absorption, which implies HEE occurs by hydrogen absorption from external gaseous hydrogen environments. Fatigue properties in high pressure gaseous hydrogen environments were evaluated by means of internal or external pressurization tests. Austenitic stainless steels such as type 316L showed little decrease in fatigue life by hydrogen, while metastable stainless steel as type 304 or precipitation hardened superalloy as A286 showed degradation in fatigue life by hydrogen gas. Low alloy steel also showed a decrease in fatigue life in hydrogen, while high strength low alloy steel with much Mo and V showed longer fatigue life than conventional steel.

KEY WORDS: hydrogen environment embrittlement; hydrogen gas; SSRT; fatigue; austenitic stainless steel; low alloy steel.

In this paper, recent research on HEE is reviewed from 1. Introduction the viewpoint of tensile properties, effect of hydrogen For rapid commercialization of fuel cell vehicles, high absorption and fatigue properties of stainless and low alloy pressure hydrogen systems for storage and transportation of steels. Tensile properties were evaluated by means of Slow hydrogen fuel must be developed in the near future. For the Strain Rate Test (SSRT) in an autoclave pressurized with safety and public acceptance of using compressed high pres- gaseous hydrogen. Relationship between tensile properties sure hydrogen gas, investigating the effect of gaseous and sub-surface hydrogen concentration were discussed. hydrogen on mechanical properties of structural materials Fatigue properties were evaluated by means of internal and used for the hydrogen system is one major subject. It is external pressure fatigue tests to assess fatigue life of tubes widely recognized that gaseous hydrogen decreases mechan- or cylinders used for transportation and storage of hydrogen. ical properties of steels or other metals.1–8) The environmental degradation in gaseous hydrogen is called Hydrogen 2. Tensile Properties Evaluated by SSRT Environment Embrittlement (HEE), or called Hydrogen Gas Embrittlement (HGE) recently, which occurs when a hydro- Tensile tests in gaseous hydrogen are valuable as rapid gen free material is mechanically tested in gaseous hydro- and economical screening tests, which provide a good indi- gen near room temperatures. NASA has carried out many cation of HEE.5,6) Recently, SSRT under extremely low studies on HEE in the 1960’s for the development of fuel strain rates is widely used for assessment of HEE as fol- systems for space shuttles.1–3) Many studies on HEE have lows,10–20) because susceptibility to hydrogen embrittlement been also carried out in Japan.5–8) Additionally, several increases with a decrease in the strain rate. national projects have started and are now continuing to Figure 1 is a schematic illustration of the SSRT appara- select and develop appropriate materials for the high pres- tus. Prior to a test, the autoclave was evacuated and replaced sure hydrogen systems.9–29) with hydrogen gas several times to remove air completely,

© 2012 ISIJ 234 ISIJ International, Vol. 52 (2012), No. 2 then filled with hydrogen and pressurized. The purity of es relative effects of alloying elements to Ni on the auste- 30) hydrogen gas was 99.99999%. SSRT was carried out in an nitic stability. Md30 indicates the temperature at which a autoclave pressurized with 45 to 90 MPa gaseous hydrogen 50% martensite structure is produced by a strain of 30%.31) at the temperature range from –40 to 85°C. Many tests were Higher Nieq and lower Md30 values mean higher stability of carried out under the strain rate of 3×10–6 s–1, as there was austenitic phase. Type 420 is a quenched-tempered marten- no remarkable effect of strain rates on HEE susceptibility in sitic stainless steel with the tensile strength of 900 MPa. the range from 3×10–7 to 8×10–5 s–1.7) Fracture elongation SCM435 is a quenched-tempered low alloy steel specified (El.), tensile strength and reduction of area (R.A.) in gas- in JIS (Japanese Industrial Standards) G4053 with tempered eous hydrogen were compared to the values in air or in martensitic microstructure at tensile strength of 800 MPa. nitrogen at the same temperatures. Figure 2 shows examples of fracture surfaces after SSRT Several steels were used as listed in Table 1. They at room temperature in 45 MPa gaseous hydrogen. The frac- include steel sheets or cylinders on the market or modified ture surface of type 304L clearly showed a transgranular laboratory melt steels. Solution-annealed 300 series auste- morphology indicating embrittlement by hydrogen as shown nitic stainless steels, type 304, 304L, 316 and 316L with in Figs. 2(a) and 2(b). In contrast, the fracture surface of various chemical compositions within AISI specifications type 316L revealed a ductile tearing by void formation as were prepared. Tensile cold-works from 30 to 40% were shown in Figs. 2(c) and 2(d). applied to the 316L sheet after the solution heat treatment. Figure 3 summarizes relative fracture El. and relative A286 is a precipitation hardened iron-based superalloy with R.A. evaluated by SSRT in 45 MPa hydrogen at room tem- the tensile strength more than 1 000 MPa. A286 was solu- perature. “Relative” means the ratio of fracture El. and R.A. tion-annealed at 900°C for 1 hr, followed by aging heat in high pressure hydrogen to those in air or in nitrogen at treatment at 720°C for 16 hrs. Nieq and Md30 are parameters ambient pressure. Type 316L showed no degradation. A286 empirically estimated from chemical compositions of steels, and cold worked 316L showed a slight decrease in fracture expressing the stability of austenitic structure. Nieq express- Table 1. Chemical compositions of investigated steels.

Chemical compositions (mass%) Nieq Md30 Material ° CSiMnNiCrMoN Ti(%) ( C) 304 0.04 0.40 1.61 9.34 18.27 0.25 0.038 – 23.8 16.6 304L 0.02 0.35 1.36 9.08 18.15 0.23 0.050 – 23.3 28.6 316 (A) 0.06 0.48 0.86 10.43 16.07 2.17 0.029 – 24.4 1.9 316 (B) 0.05 0.48 0.78 11.09 17.34 2.00 0.023 – 25.6 –9.1 316 (C) 0.05 0.57 0.82 11.27 17.49 2.07 0.028 – 26.1 –19.7 316L 0.02 0.53 0.88 12.04 17.82 2.09 0.041 – 27.3 –23.5 A286 0.05 0.38 1.01 25.16 14.75 1.28 0.005 2.18 37.3 –89.0 420 0.20 0.28 0.92 0.12 12.90 0.01 0.027 – 9.9 – SCM435 0.37 0.25 0.77 0.01 1.10 0.26 0.004 – 1.9 –

Nieq (%)=Ni+0.65Cr+0.98Mo+1.05Mn+0.35Si+12.6C Fig. 1. SSRT apparatus. Md30 (°C)=413–462(C+N)–9.2Si–8.1Mn–13.7Cr–9.5Ni–18.5Mo

–6 –1 Fig. 2. Fracture surface after SSRT. (a) and (b) type 304L, (c) and (d) type 316L, strain rate 3×10 s in 45 MPa H2 at R.T.

Fig. 4. Effect of chemical compositions on HEE susceptibilities with strain rate of 3×10–6 s–1 in 45 MPa H . Fig. 3. HEE susceptibilities of investigated steels with strain rate of 2 –6 –1 3×10 s in 45 MPa H2 at R.T. “Relative” means the ratio of fracture El. and R.A. in high pressure hydrogen to those in air or in nitrogen at ambient pressure.

El. and R.A. by hydrogen, implying a little effect of the precipitation hardening or the cold work on HEE susceptibility. Ductility losses of type 316 series strongly depended upon their chemical compositions. 316(C) with rich alloying elements such as Cr and Ni, showed no evidence of degradation, while 316(A) and 316(B) showed a remarkable decrease in fracture El. and in R.A. in gaseous hydrogen. The tendency suggests that the chemical composition is one of the important factors affecting HEE, because these steels had quite different HEE susceptibility regardless of the small difference in chemical compositions. Type 304L, type Fig. 5. Effect of the volume fraction of martensite on HEE suscep- 420 and low alloy steel SCM435 showed a remarkable –6 –1 tibilities with strain rate of 3×10 s in 45 MPa H2. decrease in fracture El. and R.A. in gaseous hydrogen. Figure 4 summarizes HEE susceptibilities of solution- annealed 300 series austenitic stainless steels from the view- attributable to the localization of planar dislocations.27,28) point of their chemical compositions and testing tempera- These results suggest that a little difference in chemical tures. Steels with higher Md30 values, with poorer alloying compositions strongly affects HEE susceptibility. Effects of elements such as Cr or Ni, showed low relative R.A., indi- chemical compositions of stainless steels on their HEE sus- cating high HEE susceptibilities, while steels with much ceptibilities have been widely investigated in previous stud- alloying elements showed higher relative R.A.. Relative ies.4,7,8) NASA reported that type 316 stainless steel was R.A. increased with an increase in the testing temperature, negligibly embrittled in highly pressurized gaseous hydro- which indicates increasing the testing temperature decreased gen at room temperature,1–3) while Fukuyama and Yokogawa HEE susceptibility. reported that type 316 showed a ductility loss in hydrogen Microstructural alteration during SSRT was investigated environments.5) The difference would be attributed to a little by an X-ray diffraction technique to confirm whether or not difference in chemical compositions between investigated strain induced martensitic phase had been produced by materials. Therefore, special attention must be paid to the deformation. Figure 5 shows the effect of the volume frac- stability of austenitic structure expressed by Nieq or Md30 tion of strain-induced martensite phase after SSRT near the from a practical point of view. fractured portion on HEE susceptibilities. The results mean that HEE susceptibilities were strongly affected by marten- 3. Effect of Sub-surface Hydrogen Concentration on sitic transformation during deformation. The effects of Hydrogen Embrittlement chemical compositions and testing temperatures on HEE susceptibilities in Fig. 4 were explainable based on marten- It is important to investigate effects of an amount of sitic transformation during SSRT, as higher Md30 (poorer absorbed hydrogen on HEE susceptibility to understand the alloying elements) and lower temperatures accelerated mar- environmental severity determining HEE. Figure 6 shows tensitic transformation. It was also observed that weld met- the effect of hydrogen gas pressure on HEE susceptibilities als containing a little amount of delta ferrite phase (less than of types 316L and 304L. Type 316L showed good resistance 20 volume %) showed same tendency on Figs. 4 and 5.15,16) to HEE up to 90 MPa hydrogen. Relative fracture El. and It is generally recognized that a body-centered cubic (bcc) relative R.A. of type 304L decreased with an increase in structure such as martensite or ferrite phase has greater sus- hydrogen pressure. Absorbed hydrogen concentration into ceptibility to hydrogen embrittlement than a fcc structure. austenitic stainless steels was measured by means of expo- Furthermore, it was recently reported that the poor resis- sure tests in high pressure gaseous hydrogen.10–14) It was tance to HEE of metastable austenitic stainless steels is confirmed that absorbed hydrogen concentration into auste-

© 2012 ISIJ 236 ISIJ International, Vol. 52 (2012), No. 2 nitic stainless steels increased with an increase in hydrogen small hydrogen diffusion coefficient of austenitic phase at pressure and temperature according to Sievert’s law. There- room temperature. Type 304L showed a remarkable degra- fore, the effect of hydrogen pressure in Fig. 6 was attribut- dation in all test conditions, and the relative R. A. decreased able to an increase in absorbed hydrogen concentration. with an increase in sub-surface hydrogen concentration. Figure 7 summarizes the effect of hydrogen concentra- Type 316L showed no degradation by SSRT in hydrogen, tion near the material surface (sub-surface hydrogen con- and showed a little ductility loss by means of pre-charged centration) on HEE susceptibilities of types 316L and 304L. technique. However, type 316L also showed ductility loss Figure 7 also contains relative R.A. data measured by using by cathodic charging SSRT under much absorbed hydrogen pre-hydrogen charged specimens in high temperature gas- concentration, although type 316L showed greater relative eous hydrogen (45 MPa H2 at 85°C for 1 000 hrs and R. A. than type 304L. Figure 7 suggests that all types of 70 MPa H2 at 250°C for 72 hrs), then followed by SSRT in degradation – SSRT in gaseous hydrogen (HEE), pre- air at room temperature. This type of degradation is called charged hydrogen embrittlement (IRHE) and hydrogen Internal Reversible Hydrogen Embrittlement (IRHE)1–3) embrittlement by cathodic charging in aqueous solution – caused by absorbed internal hydrogen within the material. were well explainable based on sub-surface hydrogen. Additionally, SSRT under cathodic charging in aqueous Furthermore, surface conditions can affect hydrogen solution was also carried out for comparison purpose.10,11) entry from gaseous hydrogen environments because oxide Cathodic hydrogen charging was conducted in 0.1 N film formed on the surface of stainless steels prevented H2SO4 + 0.01 M (NH2)2SC under the cathodic current den- hydrogen entry. It was confirmed that SSRT in gaseous sity in the range from 0.001 to 30 mA/cm2. Relative R.A. hydrogen broke the oxide film and promoted hydrogen entry values measured by SSRT in gaseous hydrogen and under into steel.21) cathodic charging were correlated to sub-surface hydrogen concentration (horizontal axis) derived from exposure tests 4. Fatigue Properties Evaluated by Internal and Exter- using coupon specimens. The sub-surface hydrogen concen- nal Pressure Fatigue Tests tration was calculated based on hydrogen distribution within a specimen determined by diffusion coefficient of austenitic Tubes and cylinders used for transportation and storage of stainless steel at room temperature, the specimen size and hydrogen are periodically pressurized through the long-term exposure time.13,14) The reason behind the calculation was an service life. Thus, evaluating fatigue properties of these experimental difficulty. The calculation indicated that components is important for the adequate design and the hydrogen localized near the specimen surface due to very safe operation of high pressure gaseous hydrogen systems. Fatigue properties of candidate stainless and low alloy steels were evaluated by means of internal and external pressure fatigue tests as shown in Fig. 8.17–20,22–24) The tubular specimen was filled with hydrogen gas or inert gas. The inner gas pressure was cyclically varied in the internal pressure fatigue test as shown in Fig. 8(a). The internal gas pressure was kept constant and the outside of the specimen is cyclically pressurized by water in the external pressure fatigue test as shown in Fig. 8(b). Actual service conditions of tubes are well simulated by the internal pressure fatigue test, while the external pressure fatigue test can be conducted in shorter cycle time. The maximum pressure of the internal gas was about 90 MPa, and the testing temperature was room temperature. An example of tubular specimens for the pressure Fig. 6. Effect of hydrogen pressure on HEE susceptibilities with fatigue tests is shown in Fig. 9. Stress was mainly applied strain rate of 3×10–6 s–1 at R.T. in the circumferential direction because both ends of the specimen were fixed in the test machine. A semi-elliptical

Fig. 7. Effect of estimated sub-surface hydrogen concentration on Fig. 8. Internal and external pressure fatigue tests. (a) internal pres- hydrogen embrittlement. sure fatigue test and (b) external pressure fatigue test.

Fig. 9. Tubular specimen for pressure fatigue tests. All dimension in mm.

Fig. 12. Effect of cycle time on fatigue properties of type 316L in H2 of Pi=10–85 MPa and type 304 in H2 of Pi=10– 70 MPa.

Fig. 10. Appearance of a specimen and fracture surface after an external pressure fatigue test of type 304 in H2 of Pi=56 MPa with 20 s/cycle.

Fig. 13. Fatigue properties of high strength low alloy steels in H2 of Pi=85 MPa with 20 s/cycle.

although the difference was small. Type 304 showed the degradation of the fatigue life in the hydrogen environment especially at low stress amplitudes, due to local martensitic transformation during straining.23,24) The fatigue life of cold worked 316L in hydrogen showed the same tendency as that in argon. The fatigue life of A286 in hydrogen was longer than that of type 316L, although it was decreased by hydrogen. Precipitation hardening by gamma prime phase or coarse secondary inter-metallic phases at grain boundaries

Fig. 11. Fatigue properties of stainless steels in H2 of Pi=55– would decrease the resistance to fatigue crack growth in the 85 MPa with 20 s/cycle. case of A286.22–24) Figure 12 shows the effect of cycle time on the fatigue initial notch was machined at the inner surface by electric lives of type 316L and 304. The fatigue life of type 316L discharge. During the fatigue tests, a fatigue crack initiated decreased with increase in cycle time, and reached a certain at the inner notch and it propagated to the outside of the value at longer cycle time. Although the fatigue life of type specimen. The fatigue life was determined as the cycle when 304 was shorter than that of type 316L, it showed same ten- the internal pressure started to decrease due to the leakage dency. The result implies that the fatigue fracture needs the of the internal gas. The fatigue life in hydrogen was com- time for hydrogen to diffuse and accumulate at the crack tip. pared with that in argon to investigate the effect of gaseous Additionally, the data are important for the fatigue life pre- hydrogen on fatigue properties. diction of components used for the hydrogen systems from Figure 10 shows a specimen after an external pressure a practical point of view. fatigue test. A crack with a semi-elliptical shape initiated at Figure 13 shows fatigue lives of low alloy steels – con- the notch and propagated to the outside of the specimen by ventional SCM435 and 0.4%C–1%Cr–0.7%Mo–0.25%V the fracture surface observation as shown in Fig. 10(b). containing martensitic steel (V steel).23,24,29) “Relative” Relationships between the circumferential stress ampli- fatigue lives in the vertical axis means the ratio of cycles to tude at the inner surface and the fatigue life of stainless failure between in hydrogen and in inert environment. The steels are shown in Fig. 11. The fatigue life in hydrogen was relative cycles to failure of SCM435 decreased with increas- slightly shorter than that in argon in the case of type 316L, ing the strength levels, indicating an increase in strength lev-

© 2012 ISIJ 238 ISIJ International, Vol. 52 (2012), No. 2 els increase the detrimental effect of hydrogen. The V steel life by hydrogen. The fatigue life of type 304L was shorter showed longer fatigue lives than SCM435 at the same than that in argon especially at low stress amplitudes due to strength levels. That means using the V steel enables an local martensitic transformation during straining. The increase in the strength levels or an increase in fatigue lives. fatigue life of A286 in hydrogen was longer than those of The fatigue fracture surface of SCM435 revealed intergran- 316L, although it was decreased by hydrogen. The fatigue ular cracking, although the fracture surface of the V steel life of conventional low alloy steel SCM435 was remark- revealed transgranular cracking.23,24) The difference in frac- ably decreased by hydrogen especially at high strength levels. ture surface morphology was attributed to the difference in V containing low alloy steel showed better fatigue life than precipitated carbides morphologies. In SCM435, film like SCM435 due to improvement in carbides morphologies. cementite preferentially precipitated along prior austenite grain boundaries. Carbides on grain boundaries decreased REFERENCES the resistance to fatigue crack propagation in hydrogen. On 1) H. R. Gray: Hydrogen Embrittlement Testing, ASTM STP 543, ed. the contrary, uniform and spheroidized cementite precipitat- by L. Raymond, ASTM, Philadelphia, (1974), 133. 2) R. P. Jewett, R. J. Walter, W. T. Chandler and R. P. Frohmbert: ed in the V steel which is less harmful to hydrogen embrit- NASA CR-2163, NASA, Washington, (1973). tlement. The results in Fig. 13 mean that the metallurgical 3) W. T. Chandler and R. J. Walter: Hydrogen Embrittlement Testing, ASTM STP 543, ed. by L. Raymond, ASTM, Philadelphia, (1974), 170. improvements in carbides morphologies increased fatigue 4) G. R. Caskey, Jr.: Environment degradation of engineering materials in life in hydrogen in the case of the V steel. It was also con- hydrogen, eds. by M. R. Louthan, Jr., R. P. McNitt and R. D. Sisson, firmed that the V steel showed better resistance to HEE than Jr., Virginia Polytechnic Institute, Blacksburg, VA, (1981), 283. 29) 5) G. Han, J. He, S. Fukuyama and K. Yokogawa: Acta Mater., 46 SCM435 by means of SSRT with a notch in hydrogen. (1998), 4559. Generally, the ratings of fatigue properties in gaseous 6) D. Sun, G. Han, S. Vaodee, S. Fukuyama and K. Yokogawa: Mater. Sci. Technol., 17 (2001), 302. hydrogen show the same tendency to that of SSRT properties. 7) S. Nomura and M. Hasegawa: Bull. Jpn. Inst. Met., 15 (1976), 563 That means the same metallurgical controls are effective to (in Japanese). improve both fatigue and tensile properties in hydrogen. 8) S. Nomura and M. Hasegawa: Tetsu-to-Hagané, 64 (1978), 288 (in Japanese). However, there are some exceptions such as A286. There- 9) M. Tamura and K. Shibata: J. Jpn. Inst. Met., 69 (2005), 1039 (in fore, further investigations are needed to clarify microstruc- Japanese). 10) T. Omura, K. Kobayashi, M. Miyahara and T. Kudo: Proc. of World tural effects on fatigue and tensile properties. Hydrogen Technologies Convention (WHTC) 2005, Singapore, (2005). 11) T. Omura, K. Kobayashi, M. Miyahara and T. Kudo: Zairyo-to-Kankyo, 5. Summary 55 (2006), 139 (in Japanese). 12) T. Omura, K. Kobayashi, M. Miyahara and T. Kudo: Corros. Eng., Recent research on HEE is reviewed from the viewpoint 55 (2006), 169. 13) T. Omura, K. Kobayashi, M. Miyahara and T. Kudo: Zairyo-to-Kankyo, of tensile properties, effect of hydrogen absorption and 55 (2006), 537 (in Japanese). fatigue properties of stainless and low alloy steels. Tensile 14) T. Omura, K. Kobayashi, M. Miyahara and T. Kudo: Corros. Eng., 55 (2006), 747. properties were evaluated by means of SSRT in an autoclave 15) T. Omura, M. Miyahara, H. Hirata and T. Kudo: Zairyo-to-Kankyo, pressurized with gaseous hydrogen. Tensile properties were 57 (2008), 30 (in Japanese). correlated to the sub-surface hydrogen concentration. 16) T. Omura, M. Miyahara, H. Hirata and T. Kudo: Corros. Eng., 57 (2008), 35. Fatigue properties of candidate stainless and low alloy steels 17) T. Omura, M. Miyahara, H. Semba, M. Igarashi and H. Hirata: Proc. were evaluated by means of internal and external pressure of ASME PVP 2007, The American Society of Mechanical Engi- neers, New York, (2007), Paper No. PVP2007–26496. fatigue tests. The following conclusions were obtained. 18) T. Omura, M. Miyahaya, H. Semba, M. Igarashi and H. Hirata: J. (1) SSRT results showed that type 316L showed no High Press. Inst. Jpn., 46 (2008), 205 (in Japanese). degradation above room temperatures. A286 and cold 19) T. Omura and J. Nakamura: The Piping Eng., 53 (2011), 51, (in Japanese). worked 316L showed a slight decrease in ductility in hydro- 20) T. Omura and J. Nakamura: Zairyo-to-Kankyo, 60 (2011), 241 (in gen at room temperature. HEE susceptibilities of type 316 Japanese). 21) T. Omura, K. Moriguchi, K. Kobayashi, M. Miyahara and T. Kudo: strongly depended upon their chemical compositions. Meta- Proc. of JSCE Materials and Environments 2006, Japan Society of stable austenitic stainless steel of type 304L, martensitic Corrosion Engineering, Tokyo, (2006), 335 (in Japanese). 22) M. Miyahara, T. Omura, H. Okada, K. Ogawa, M. Igarashi and T. stainless steel of type 420 and low alloy steel of SCM435 Kudo: Proc. of World Hydrogen Technologies Convention (WHTC) showed considerable ductility losses at room temperature. 2005, Singapore, (2005). HEE susceptibilities of austenitic stainless steels were 23) J. Nakamura, M. Miyahara, T. Omura, H. Semba, M. Wakita and Y. Otome: Proc. of ISSS 2009, Kyoto, (2009). strongly affected by martensitic transformation during 24) J. Nakamura, M. Miyahara, T. Omura, H. Semba, M. Wakita and Y. deformation, which is determined by the chemical compo- Otome: Proc. of Fatigue 2010, Prague, (2010). 25) S. Fukuyama, M. Imade and K. Yokogawa: Proc. of ASME PVP sitions and testing temperatures. 2007, The American Society of Mechanical Engineers, New York, (2) Embrittlement properties evaluated by SSRT in high (2007), Paper No. PVP2007–26820. pressure hydrogen, SSRT using pre-charged specimen and 26) H. Nakagawa: Proc. of ASME PVP 2007, The American Society of Mechanical Engineers, New York, (2007), Paper No. PVP2007– SSRT under cathodic charging, were all correlated to the 26462. sub-surface hydrogen concentration. Type 304L showed a 27) S. Ohmiya and H. Fujii: Proc. of ASME PVP 2007, The American Society of Mechanical Engineers, New York, (2007), Paper No. remarkable degradation in all test conditions. Type 316L PVP2007–26492. showed little degradation by SSRT in hydrogen and SSRT 28) H. Fujii and S. Ohmiya: J. High Press. Inst. Jpn., 47 (2009), 85 (in using pre-charged specimen, although it showed remarkable Japanese). 29) T. Omura, J. Nakamura, K. Kobayashi and M. Miyahara: Proc. of ductility loss under cathodic charging with mush sub- JSCE Materials and Environments 2009, Japan Society of Corrosion surface hydrogen concentration. Engineering, Tokyo, (2009), A106 (in Japanese). 30) T. Hirayama and M. Ogirima: J. Jpn. Inst. Met., 34 (1970), 507 (in (3) Internal and external pressure fatigue test results Japanese). showed that type 316L showed little degradation of fatigue 31) T. Angel: J. Iron Steel Inst., 177 (1954), 165.