HYDROGEN ENVIRONMENT EMBRITTLEMENT OF AUSTENITIC
STAINLESS STEELS AT LOW TEMPERATURES

(Study on low temperature materials used in WE-NET 7)

G. Han1, J. He2, S. Fukuyama and K. Yokogawa
Chugoku National Industrial Research Institute, AIST, MITI, 2-2-2 Hiro-suehiro,
Kure, Hiroshima, 737-0197, Japan
1: JST Postdoctoral Researcher, 2: NEDO Industrial Researcher


1. INTRODUCTION

Austenitic stainless steels have been expected to be used as the structural materials of storage tanks, containers and line pipes for liquid hydrogen supply because of the excellent performance of the materials at low temperatures. However, it is well known that the austenitic stainless steels suffer from hydrogen embrittlement. Since Whiteman and Troiano [1] found, in 1965, that stable austenitic stainless steel of type 310 hydrogen-charged cathodically showed hydrogen embrittlement, many studies on hydrogen embrittlement of the austenitic stainless steels have been conducted [1-11].

It was found that austenitic stainless steels sensitized at temperature range of 773 to 1123 K showed severer hydrogen embrittlement than the solution-annealed ones[12-17]. Two mechanisms of hydrogen embrittlement of sensitized austenitic stainless steels have been proposed; one is hydrogen embrittlement by carbides [13,14] and the other is that by strain-induced martensite [16,17], both existing along the grain boundaries. However, it has not yet been clarified whether the carbides or the strain-induced martensite, both existing along the grain boundaries, is responsible for hydrogen embrittlement of the sensitized metastable austenitic stainless steels, because the carbides and the martensite coexist along the grain boundaries during deformation of the materials.

In this study, the austenitic stainless steels were tested in the high-pressure hydrogen at the temperature range from room temperature to 80 K. The effects of the testing temperature and the stain-induced martensite which was distinguished from that of the carbides in the sensitized steels on HEE at low temperatures is discussed.

2. EXPERIMENTAL

Commercially available types 304L, 304, 316, 310S and JJ1 austenitic stainless steels were used, whose chemical compositions are shown in Table 1. The steels were machined into cylindrical smooth tensile specimens with a gauge length of 20 mm and diameter of 4 mm. Three types of heat treatment, solution-annealed, sensitized (marked with (S)) and desensitized (SD) heat treatments, were applied to the specimens, as shown in Table 2.

The surfaces of all the specimens were ground using sandpaper, polished with paste and finally washed in an ultrasonic cleaner in an acetone bath before performing tensile tests. The tensile tests of the steels were conducted in hydrogen and helium of 1 MPa in the temperature range from 295 to 80 K using the equipment [18] specially developed in our laboratory. The purity of the testing gases was 99.9999% for hydrogen and 99.999% for helium, respectively. All the tests were conducted with an initial strain rate of 4.17x10-5 s-1 to ensure a reasonable effect of hydrogen on the tensile properties. The fracture surfaces of the specimens were observed using a scanning electron microscope (SEM) after the tensile tests. The carbides and the strain-induced martensite were examined by a transmission electron microscope (TEM) of JEOL 2000FX.

3. RESULT

Effects of testing temperature and heat treatment on HEE

The load-elongation curves for solution-annealed type 304 in hydrogen and helium of 1 MPa at 295, 220 and 80 K are shown in Fig. 1. Hydrogen shows a marked effect of reducing both the ultimate tensile strength (UTS) and the total elongation to fracture, but no effect was observed on the 0.2 % proof stress. The influence of hydrogen on tensile properties depends on the testing temperature. Hydrogen shows a slight reduction in UTS and large reduction in the total elongation at room temperature, large reduction in both UTS and total elongation at 220 K, and no effect at 80 K.

The effect of testing temperature on elongation and reduction of area of type 304L in hydrogen and helium of 1 MPa is shown in Fig. 2. Hydrogen shows considerable effect at temperature range from 165 K to room temperature. The ductilities decrease gradually and arrive at a maximum at 220 K and increase rapidly in hydrogen, while those decrease gradually in helium with decreasing the temperature. The ductility curves in hydrogen and helium join each other at 165 K.

HEE of a material is quantitatively represented as the relative reduction of area, namely, the ratio of the reduction of area in hydrogen divided by that in helium (j Hydrogen / j Helium) [8]. The value of the relative reduction of area of 1.0 indicates no hydrogen influence. The more the value decreases, the more the hydrogen influence increases. The effect of testing temperature on the relative reduction of area of the austenitic stainless steels in hydrogen and helium of 1 MPa is shown in Fig. 3. Types 304L, 304 and 316 are susceptible to HEE and HEE of the steels reaches a maximum at about 220 K, while types 310S and JJ1 are not susceptible to HEE in the temperature range from 295 to 80K. It should be noted that HEE of types 304 and 316 depended on the type of heat treatment and the sensitization caused a considerable increase in HEE, while the desensitization caused recovery from HEE as much as that of the solution-annealed specimens. The results shown above indicate that hydrogen influence is maximum at around 220 K. Therefore, the observations of the fracture surfaces and the microstructures described in sections 3.2 and 3.3 were primarily made at 220 K.

Fracture surfaces

Fracture surfaces, obtained by SEM, of solution-annealed, sensitized and desensitized type 304 fractured in hydrogen of 1 MPa at 220 K are shown in Fig. 4. All the fracture surfaces clearly show brittle fracture modes regardless of the type of heat treatment. But transgranular fracture along the strain-induced martensite laths and the twin boundaries are found in 304, intergranular fracture is seen in 304(S), and transgranular fracture mainly occurs in 304(SD). It is clear that the sensitization changed the hydrogen induced fracture mode of the steel from transgranular to intergranular, and the desensitization caused the fracture mode to recover from intergranular fracture to transgranular fracture. The same fracture behavior is also observed in type 316. Fracture surfaces of sensitized type 310S fractured in hydrogen and helium of 1 MPa at 220 K are shown in Fig. 5. Dimple rupture is observed on the fracture surfaces tested both in hydrogen and helium.

Strain-induced martensitic transformation

The microstructure of sensitized type 304 deformed 10 % at 220 K is shown in Fig. 6. M23C6 carbides are sporadically observed along the grain boundary, and the strain-induced martensite is also observed as a band along the grain boundary in deformed 304(S). The microstructure of desensitized type 304 deformed 10 % at 220 K is shown in Fig.7. Transformation of a' martensite, occurs predominantly in the depleted zone along the grain boundaries formed during sensitization, and also in grains with deformation. Discontinuous carbides along the grain boundary are still observed in 304(SD), but no martensite is identified along the grain boundaries. It is evident that desensitization prevents the depletion of chromium and carbon, and therefore the transformation of strain-induced martensite does not occur along the grain boundaries.

The microstructure of sensitized type 310S deformed 10 % at 220 K is shown in Fig. 8. Continuous carbides along the grain boundary and fine carbides in the grains are observed while neither a' martensite nor e martensite is identified. The transformation of strain-induced martensite did not occur in type 310S, even in the sensitized one.

4.DISCUSSION

Effect of strain-induced martensite

Tensile properties of metastable austenitic stainless steels of types 304L, 304 and 316 were degraded considerably in 1 MPa hydrogen environment, independent of the type of heat treatment, whereas those of stable austenitic stainless steel of types 310S and JJ1 were not affected in hydrogen environment. It is considered that fracture occurred along the strain-induced martensite in the grains in hydrogen environment, as described previously [2-4]. In sensitized types 304 and 316, the sensitization markedly enhanced the influence of hydrogen on the tensile properties and changed the fracture mode from transgranular to intergranular in hydrogen, as shown in (Fig. 4b). M23C6 carbides precipitated along the grain boundaries during the sensitization and at the same time, a zone depleted of chromium and carbon was formed in the region adjacent to the carbides [24]. Transformation of the strain-induced martensite preferentially occurred in the depleted zone along the grain boundaries, as shown in (Fig. 6), for type 304, because of the loss of chromium and carbon which stabilize the austenite in the zone.

The desensitizing heat treatment was applied to sensitized types 304 and 316 to distinguish the effect of the strain-induced martensite from that of the carbides, both existing along the grain boundaries in the materials on HEE. The desensitization considerably improved the resistance to HEE (Figs. 3) and changed the fracture mode in hydrogen from intergranular to transgranular (Fig.4c). TEM observation showed that the transformation of strain-induced martensite along the grain boundaries did not occur upon desensitization. These results indicated that the improvement of HEE was due to the absence of strain-induced martensite along the grain boundaries, caused by the elimination of the depleted zone by homogeneous diffusion of chromium and carbon during desensitization. The effect of sensitization on hydrogen embrittlement in austenitic stainless steels hydrogen-charged thermally have been attributed to the precipitation of carbides along the grain boundaries [16,17], but this mechanism could not explain the present results. If the effect of sensitization on HEE were due to carbides rather than martensite, intergranular fracture would be observed in the desensitized materials. It is clear that intergranular fracture of the sensitized materials is caused not by the carbides, but by the strain-induced martensite along the grain boundaries.

In stable austenitic stainless steel of type 310S, transformation of the strain-induced martensite was not observed, as shown in Fig. 8. It is considered that the amount of elements which stabilize the austenite is sufficient to prevent martensitic transformation, even though sensitization caused the precipitation of carbides along the grain boundaries. Hydrogen showed no effect in these materials. This also indicates that HEE of the material is not related to the carbides along the grain boundaries.

Since the diffusion coefficient of hydrogen in a' martensite is much higher than that in austenite [25], so a' martensite provides a path for rapid hydrogen transportation in the material. Therefore the susceptibility, of types 304L, 304 and 316 with a' martensite, to hydrogen embrittlement increases but that of types 310S and JJ1 without martensite was negligible. Furthermore, sensitization preferentially causes the transformation of a' martensite along the grain boundaries of types 304 and 316, provides a short-circuit path for hydrogen transportation to the crack tip in the materials, and enhances intergranular fracture and susceptibility to hydrogen embrittlement.

5. CONCLUSION

Tensile behaviors of the austenitic stainless steels of types 304L, 304, 316, 310S and JJ1 were investigated in hydrogen and helium of 1 MPa in the temperature range from 295 to 80 K. The effects of testing temperature and strain-induced a' martensite on HEE were examined. The results obtained were as follows.

1. Metastable austenitic stainless steels of types 304L, 304 and 316 showed considerable HEE but stable austenitic stainless steel of types 310S and JJ1 were not affected by hydrogen environment. HEE of types 304L, 304 and 316 depends on the testing temperature, and the maximum HEE occurred at around 220 K independent of the type of heat treatment.

2. Susceptibility of types 304 and 316 to HEE was enhanced by sensitization and recovered by desensitization. Transgranular fracture was observed in the solution-annealed materials, intergranular fracture in the sensitized materials and transgranular fracture in the desensitized materials in hydrogen environment.

3. The strain-induced martensite was observed together with the carbides, both along the grain boundaries of the sensitized materials. As a result of desensitization, the formation of martensite along the grain boundaries was inhibited. HEE of the sensitized materials was not due to the carbides, but to the strain-induced a' martensite along the grain boundaries

ACKNOWLEDGEMENT

The authors would like to thank the New Energy and Industrial Technology Development Organization and the Japan Science and Technology Corporation for postdoctoral fellowships


REFERENCES

  1. M.B.Whiteman and A.R.Troiano, Corrosion, 21, 53 (1965).
  2. R.B.Benson, Jr., R.K.Dann, and L.W.Roberts, Jr., Trans. Metall. Soc. AIME, 242, 2199 (1968).
  3. S.Singh and C.Altstetter, Metall. Trans. A, 13A, 1799 (1982).
  4. T.P.Perng and C.J.Altstetter, Metall. Trans. A, 18A, 123 (1987).
  5. M.R.Louthan, G.R.Casky, J.A.Donovav, and D.E.Rawl: Mater. Sci. Eng., 10(1972), pp. 357-368.
  6. N.Ohtani, S.Asano, Y.Fujishima, and Y.Yamamasu, J. Jpn Inst. Metals, 37, 746 (1973).
  7. J.H.Holbrook, Hydrogen effect in materials, Ed. by I.M.Bernstein and A.W.Thompson, AIME, Warrendale, 655 (1981).
  8. G. R.Caskey, Jr., Environment Degradation of Engineering Materials in Hydrogen, Ed. by M.R.Louthan, Jr., R.P.McNitt, and R.D.Sisson, Jr., Virginia, 283 (1981).
  9. L.Ma, G.Liang, and Y.Li, Adv. Cry. Eng. (Materials), 38, 77 (1992).
  10. J.R.Buckley and D.Hardie, Mater. Sci. Technol., 9, 259 (1993).
  11. J.R.Buckley and D.Hardie: Corrosion Science, 34, 93 (1993).
  12. J.P.Fidelle, R.Bernardi, R.Broudeur, C.Roux, and M.Rapin, Hydrogen embrittlement testing, ASTM-STP 543, ASTM, 543, 221 (1974).
  13. C. L. Briant, Metall. Trans. A, 10A, 181 (1979).
  14. S. P. Hannula, H. Hanninen, and S. Tahtinen, Metall. Trans. A, 15A, 2205 (1984).
  15. S.Fukuyama, K.Yokogawa, K.Kudo, and M.Araki, Trans. Jpn. Inst. Met., 26, 325 (1985).
  16. A.W.Thompson, Mater. Sci. Eng., 14, 253 (1974).
  17. M.Hasegawa and S.Nomura, J. Iron Steel Inst. Jpn., 59, 1961 (1973).
  18. G.Han, J.He, S.Fukuyama, and K.Yokogawa, Rev. Sci. Instrum., 68, 4232 (1997).
  19. A.R.Troiano, Trans. ASM, 52, 54 (1960).
  20. D. M. Symons and A. W. Thompson, Metall. Trans. A, 27A, 101 (1996).
  21. J.P.Hirth, Metall. Trans. A, 11A, 861 (1980).
  22. M.R.Louthan, Jr., Hydrogen in Metals, Ed. by I.M.Bernstein and A.W.Thompson, ASM, (1974), p. 53.
  23. J.K.Tien, A.W.Thompson, I.M.Bernstein, and R.J.Richards, Metall. Trans. A, 7A(1976), 821-829.
  24. E. C. Bain, R.H. Aborn, and J.B.Rutherford, Trans. Am. Steel Treating Soc., 21, 481 (1933).
  25. H.G.Nelson, Treatise Mater. Sci. Technol., 25, 275 (1983).