1. INTRODUCTION

Austenitic stainless steels are one of the candidate structural materials used for the liquid hydrogen transportation and storage systems. There are many investigations about the mechanical properties of austenitic stainless steels at low temperature[1-6], however, the mechanical properties of those weld metals have not been investigated enough[7-11]. Moreover, susceptibility to hydrogen embrittlement of those weld metals hardly has been investigated.

So the purpose of this work was to clarify the effect of the amount of d-ferrite phase and hydrogen charging on the mechanical properties of the weld metals of commercial grade austenitic stainless steels.

2. EXPERIMENTAL PROCEDURE

Four kinds of welding electrodes for each stainless steel were newly designed and prepared by varying the balance of Cr and Ni contents so that the strength level of weld metals would be almost the same and d-ferrite content of weld metals would change between 0% and 10%. The effect of d-ferrite content was considered to be cleared by using these welding electrodes. The chemical composition of welding electrodes for both steels are also shown in TABLE I. The welding electrodes for SUS304L and SUS316L is called 304L series and 316L series, respectively. Welded joints used in this work were made by means of Tungsten Inert-Gas arc welding. d-ferrite content of those weld metals measured by Ferrite-Meter are shown in TABLE II.

Hydrogen charging of testing specimen was carried out under hydrogen gas with a pressure of 1.67MPa in an autoclave kept at 573K to make hydrogen content of the specimen about 10ppm. Hydrogen contents of weld metals with and without hydrogen charging are also shown in TABLE II.

Tensile test was performed at room temperature (R.T.), in liquid nitrogen (77K) and in liquid helium (4K) by using specimen without hydrogen charging. The strain rate of tensile test at R.T. was 1.67x10 ^-4 s^-1 up to 0.2% yield strength and then 1.67x10^-3 s^-1 to failure, whereas, 2.38x10^-4 s^-1 up to 0.2% yield strength and then 9.52x10^-4 s^-1 to failure for 77K and 4K. Charpy impact test and fracture toughness test were performed at room temperature (R.T.), in liquid nitrogen (77K) and in liquid helium (4K) by using specimens with and without hydrogen charging. KIC value was evaluated from JIC value in J-integral characterization test for the evaluation of fracture toughness. The dimension and sampling position of each specimen are shown in FIGURE 1.

3. RESULTS AND DISCUSSION

FIGURE 2 and FIGURE 3 shows the effect of temperature on the tensile properties of weld metals of 304L series and 316L series, respectively. The tensile strength of all weld metals in both series remarkably increases linearly with decrease in temperature, and the rate of this increase is larger in 304L series than in 316L series. The tensile strength of weld metals in each series are found to be approximately equivalent at each testing temperature as was expected. In the case of 4K, 304L series and 316L series have a remarkable high tensile strength of about 1500MPa and about 1400MPa, respectively. The yield strength of weld metals in each series are also found to be approximately equivalent at each testing temperature. They are about 400MPa in 304L series regardless of testing temperature, whereas in 316L series they gradually increase with decrease in temperature and reach to about 600MPa at 4K. The elongation and the reduction of area of all weld metals in both series drop as the temperature decreases. However, their values remain above 30% even at 4K in both series except WM-E containing 10% d-ferrite in 316L series. The fractography of the fracture surface of all specimens show dimple fracturing regardless of d-ferrite content and testing temperature.

3.2. Charpy Impact Properties

FIGURE 4 shows the effect of d-ferrite content and hydrogen charging on the Charpy absorbed energy of weld metals in both series. There is slight effect of d-ferrite content on Charpy absorbed energy at R.T. and 77K in 304L series. However, a decrease of about 40J is found at 4K when d-ferrite content increases from 0% to 2%, and then there is almost no change even if d-ferrite content increases to 11%. While in 316L series there is also slight effect of d-ferrite content on Charpy absorbed energy at R.T.. However, a remarkable decrease of about 80J is found at both 77K and 4K when d-ferrite content increases from 0% to 1%, and then absorbed energy gradually decreases as d-ferrite content increases to 10%. These results revealed that the impact properties of weld metals at low temperature depend not on the amount of d-ferrite but on whether d-ferrite exists or not. And this dependence was particularly strong in 316L series.

The Charpy absorbed energy of all weld metals in both series significantly drop when testing temperature decreases from R.T. to 77K, and then there are slight decrease as the testing temperature decreases to 4K. The degree of this drop increases with d-ferrite content increment, and it is smaller in 304L series than in 316L series. WM-A containing 11% d-ferrite in 304L series has a high absorbed energy of 90J even at 4K, whereas WM-E containing 10% d-ferrite in 316L series has a low absorbed energy of 35J at 4K.

Regardless of d-ferrite content and testing temperature, due to hydrogen charging, the absorbed energy decreases by 15-60J in 304L series, which is explained as hydrogen embrittlement. While in 316L series the hydrogen embrittlement is found in WM-H containing no d-ferrite at all testing temperatures only, and the decrease of absorbed energy is about 25J. However, the hydrogen embrittlement is not found in WM-E and WM-F containing 10% and 5% d-ferrite, respectively.

The observations of fracture surface shows that fracture mode on each testing condition was the same in both series. The difference in fractography with or without hydrogen charging was not found. FIGURE 5 shows the fractography and the optical micrograph of vertical section to the notch and the fracture surface of WM-A containing 11% d-ferrite in 304L series which was tested after hydrogen charging. It can be seen that a crack propagates through the network of d-ferrite or d-ferrite/austenite interface. Such an appearance as in FIGURE 5 were observed on the fracture surface of specimens containing d-ferrite tested at 77K and 4K with and without hydrogen charging. On the other hand, dimple fracturing were observed for the other conditions.

3.3. Fracture Toughness Properties

FIGURE 6 shows the effect of temperature on the KIC value of weld metals in both series. The KIC value of all weld metals drops with decrease in temperature. WM-A containing 11% d-ferrite in 304L series has a high KIC value of above 400MPa(m)^1/2 at R.T.. However, this linear drops with decrease in temperature, and its value is around 100MPa(m)^1/2 at 4K. Similarly, WM-E containing 10% d-ferrite in 316L series has a high KIC value of above 500MPa(m)^1/2 at R.T.. However, this also linearly drops with decrease in temperature, and its value is around 120MPa(m)^1/2 at 4K. Furthermore, unstable fracturing was occurred on some test conditions in 316L series. WM-D and WM-H containing no d-ferrite in both series have small rate of drop in KIC value with decrease in temperature and have a good KIC value of about 300 MPa(m)^1/2 even at 4K. The change in the KIC values of WM-C and WM-G containing 2% d-ferrite in 304L series and 1% d-ferrite in 316L series, respectively, with decrease in temperature are not similar to those of WM-D and WM-H containing no d-ferrite but similar to those of WM-A and WM-E containing 11% d-ferrite in 304L series and 10% d-ferrite in 316L series, respectively. Such a tendency like this one is similar to the result of Charpy impact test.

Moreover, the KIC value decreases due to hydrogen charging. The degree of this decreases is larger in weld metals with d-ferrite than without d-ferrite.

The observed fracture modes were similar to those observed in Charpy impact test, and a crack propagates through the network of d-ferrite or d-ferrite/austenite interface on the specimen with d-ferrite tested at low temperature. Though hydrogen embrittlement was occurred, similarly to the result of Charpy impact test, the difference in fractography with or without hydrogen charging was not found. There were good proportional relationship between the Charpy absorbed energies and the KIC values in both series.

3.4 Effect of d-Ferrite Content

The effect of d-ferrite and hydrogen charging on the mechanical properties of the weld metals of commercial grade austenitic stainless steels, SUS304L and SUS316L, were investigated in this work. As a result, it was clarified that the toughness of weld metals at low temperature depends not on the amount of d-ferrite but on whether d-ferrite exists or not, and the toughness dropped with only a few percent d-ferrite content. Moreover, hydrogen embrittlement occurred on some conditions with about 10ppm hydrogen charging. However, the reason whether hydrogen embrittlement occurs or not was not clear.

On the other hand, it is well known that the high temperature crack sensitivity of austenitic stainless steels in welding is remarkably dependent on the amount of d-ferrite. It is practically difficult to weld austenitic stainless steels without d-ferrite in weld metals.

The mechanical properties obtained in this work will serve as a reference for the design and construction of liquid hydrogen transportation and storage systems. And we will continue further investigations about the effect of hydrogen embrittlement and about the mechanical properties of various welded joints at low temperature.

4. CONCLUSION

1. The tensile strength of all weld metals remarkably increases linearly with decrease in temperature. It reaches to about 1500MPa in 304L series and about 1400MPa in 316L series at 4K. The elongation and reduction of area remain above 30% for almost all conditions even at 4K.
2. The results of Charpy impact test and fracture toughness test reveal that the toughness of weld metals at low temperature depends not on the amount of d-ferrite but on whether d-ferrite exists or not. And this dependence is particularly strong in 316L series.
3. Hydrogen embrittlement occurs on some conditions with about 10ppm hydrogen charging. However, the reason whether hydrogen embrittlement occurs or not is not clear, and further investigation is necessary.

Acknowledgments

REFERENCES

1. K. A. Warren and R. P. Reed : NBS Monograph 63. U.S. Government Printing Office, Washington, D.C. (1963)
2. D. T. Reed and R. P. Reed : Cryogenics, July (1981) 415.
3. H. I. McHenry : Austenitic Steels at Low Temperature, ed. by R. P. Reed and T. Horiuchi, Plenum Press, New York (1983)1.
4. R. M. McClintock and H. P. Gibbons : NBS Monograph 13. National Institute of Standerds and Technology, Boulder, Colorado (1960)
5. R. Ogata, M. Shimada and T. Horiuchi : Trans. Jap. Inst. Met., 27 (1986) 5.
6. H. I. McHenry : Materials at Low Temperature, ed. by R. P. Reed and A. F. Clark, ASM, Metals Park, Ohio (1983) 371.
7. W. J. Mills : International Materials Reviews, 42 (1997) 45.
8. K. S. Lee and D. D.-hughes : Austenitic Steels at Low Temperature, ed. by R. P. Reed and T. Horiuchi, Plenum Press, New York (1983) 211.
9. D. T. Reed, H. I. McHenry, P. A. Steinmeyer and R. O. Thomas,Jr : Weld. Jour., 59 (1980) 104s.
10. E. R. Szumachowski and H. F. Reid Weld. Jour., 57 (1978) 325s.
11. C. E. Witherell : Weld. Jour., 59 (1980) 326s.