This article was downloaded by: [Texas A & M International University]

On: 08 September 2015, At: 01:39

Publisher: Taylor & Francis
Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered office: 5 Howick
Place, London, SW1P 1WG

Welding International
Publication details, including instructions for authors and subscription information:
http://www.tandfonline.com/loi/twld20

Effects of chemical compositions and microstructure

on hydrogen embrittlement of austenitic stainless
steel weld metal in high-pressure gaseous hydrogen
environment
a a a a a
Hiroyuki Hirata , Tomohiko Omura , Kana Jotoku , Jun Nakamura & Takahiro Osuki
a
Nippon Steel & Sumitomo Metal Corporation, Tokyo, Japan
Published online: 24 Mar 2015.

Click for updates

To cite this article: Hiroyuki Hirata, Tomohiko Omura, Kana Jotoku, Jun Nakamura & Takahiro Osuki (2015) Effects of
chemical compositions and microstructure on hydrogen embrittlement of austenitic stainless steel weld metal in high-
pressure gaseous hydrogen environment, Welding International, 29:12, 922-927, DOI: 10.1080/09507116.2014.921084

To link to this article: http://dx.doi.org/10.1080/09507116.2014.921084

PLEASE SCROLL DOWN FOR ARTICLE

Taylor & Francis makes every effort to ensure the accuracy of all the information (the “Content”) contained
in the publications on our platform. However, Taylor & Francis, our agents, and our licensors make no
representations or warranties whatsoever as to the accuracy, completeness, or suitability for any purpose of
the Content. Any opinions and views expressed in this publication are the opinions and views of the authors,
and are not the views of or endorsed by Taylor & Francis. The accuracy of the Content should not be relied
upon and should be independently verified with primary sources of information. Taylor and Francis shall
not be liable for any losses, actions, claims, proceedings, demands, costs, expenses, damages, and other
liabilities whatsoever or howsoever caused arising directly or indirectly in connection with, in relation to or
arising out of the use of the Content.

This article may be used for research, teaching, and private study purposes. Any substantial or systematic
reproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in any
form to anyone is expressly forbidden. Terms & Conditions of access and use can be found at http://
www.tandfonline.com/page/terms-and-conditions
 Welding International, 2015
Vol. 29, No. 12, 922–927, http://dx.doi.org/10.1080/09507116.2014.921084
Selected from Quarterly Journal of the Japan Welding Society 2013 31(4) 246– 251

Effects of chemical compositions and microstructure on hydrogen embrittlement of austenitic

stainless steel weld metal in high-pressure gaseous hydrogen environment
Hiroyuki Hirata, Tomohiko Omura, Kana Jotoku, Jun Nakamura and Takahiro Osuki
Nippon Steel & Sumitomo Metal Corporation, Tokyo, Japan
(Received 26 August 2013; final version received 1 November 2013)

The effect of chemical compositions and microstructures on hydrogen embrittlement of austenitic stainless steel weld metals
in high-pressure hydrogen gas was surveyed by using the slow strain rate test (SSRT). As a result, hydrogen embrittlement of
the weld metal was hardly influenced by d ferrite in the weld metal, but by stability of austenite phase, which was estimated
by Md30 value or Ni equivalent. In the weld metal with poor stability of austenite, a0 -martensite was formed near a crack
induced by SSRT. In addition, although the crystal structure of a0 -martensite is as same as d ferrite, susceptibility of
Downloaded by [Texas A & M International University] at 01:39 08 September 2015

hydrogen embrittlement became higher with the increase in a0 -martensite. The mechanism to explain the difference between
d ferrite and a0 -martensite was considered as following. The hardness, which increases the hydrogen embrittlement
susceptibility in bcc structure, is higher in a0 -martensite than in d ferrite. In addition, a0 -martensite might be formed
continuously with propagation of a crack. Therefore, the effect of a0 -martensite on hydrogen embrittlement could be larger
compared with d ferrite.
Keywords: weld metal; hydrogen embrittlement; austenitic stainless steel; d ferrite; martensite

1. Introduction of hydrogen embrittlement susceptibility of stainless steels

Hydrogen, the most abundant element in the universe and
a clean source of energy, is expected to replace oil, coal
and other fossil fuels as a core energy source for society in
the near future. The fuel cell is being actively developed as
a technology using this. Because hydrogen is a gas at room
temperature, it is less easy to use as a liquid fuel than
petrol, which it will replace. Known methods of safely
storing and transporting hydrogen include liquid hydrogen
and the use of hydrogen-occluding alloys, and the practical
application of compressed hydrogen gas is the technology
nearest to realization. A 35-MPa high-pressure tank for use
as a storage fuel cell has been developed using resin
materials and a similar tank with the pressure increased to
70 MPa is under development. However, the provision of
hydrogen supply stations is a necessity for the develop-
ment of a hydrogen-powered society and the use of metal
materials, which are low cost and recyclable, for the
storage facilities and transportation pipes, and is under
consideration [1].
On the other hand, it is also known that metal materials
become embrittled in a hydrogen gas atmosphere at room
temperature [2] and the embrittlement characteristics of a
range of metals have been assessed systematically for many
years [2,3]. Ferrous materials, metallic materials which are
more economical and recyclable than aluminium alloys and
titanium, are used in various industrial fields. Of these,
austenitic stainless steel, which has an fcc structure, is
reputedly less susceptible to metallic materials with a bcc
structure [3] and there have been various studies of
embrittlement under high pressure in a hydrogen gas
environment [4 – 7]. For example, a report that SUS316L
stainless steel does not suffer from embrittlement in a
hydrogen gas environment from room temperature to 858C
[4]. It has been noted, on the other hand, in the comparison
q 2015 Taylor & Francis
with an austenitic structure at room temperature, there is a
difference in hydrogen embrittlement susceptibility of the
same type of 316 stainless steels depending on their
chemical compositions [6].
However, welding is considered essential for assembly
to make steel materials into structures. It is usual to use
weld metal unmodified in the solidified form with base
metal which has been conditioned by working and heat
treatment. In particular, in order to reduce the suscepti-
bility to high-temperature cracking in welding filler for
austenitic stainless steel, in many cases the components
are adjusted so that the solidified structure contains several
percentages of d ferrite with a bcc structure at room
temperature [8]. Previous reports of the effects of d ferrite
contained in weld metal on the hydrogen embrittlement
susceptibility of austenitic stainless steel weld metals
include that of Brooks and West [9], that a d ferrite content
up to 10% has no effect on hydrogen embrittlement
susceptibility. On the other hand, Luppo et al. [10]
examined the hydrogen embrittlement susceptibility of
weld metal with the d ferrite content varied up to 30% and
showed that hydrogen embrittlement was particularly
marked when the d ferrite content was high [10]. Thus,
there is no settled view on the effect of d ferrite in
austenitic stainless steel and, in particular, the under-
standing of embrittlement behaviour in a high-pressure
hydrogen gas environment is still inadequate.
In this report, an examination was made of the effect of
chemical composition and microstructure on hydrogen
embrittlement susceptibility with the aim of clarifying
the embrittlement behaviour of austenitic stainless steel
weld metal in a high-pressure hydrogen gas environment,
which is considered important in practical applications.
 Welding International 923

Table 1. Chemical compositions of materials used (mass %).

Heat C Si Mn P S Ni Cr Mo N Nieq Md30 (8C)

308A 0.018 0.46 1.44 0.004 0.001 11.0 18.7 0.01 0.0048 25.0 25.66
308B 0.018 0.50 1.45 0.003 0.001 11.0 20.4 0.01 0.0110 26.0 4.06
308C 0.019 0.47 1.44 0.003 0.001 10.3 19.4 0.01 0.0061 24.9 21.08
308D 0.018 0.49 1.46 0.004 0.001 11.7 18.5 0.01 0.0050 25.7 20.81
308E 0.019 0.50 1.43 0.003 0.001 9.1 19.4 0.01 0.0120 23.6 30.03
316A 0.018 0.50 1.49 0.002 0.001 13.8 18.5 2.51 0.0056 30.2 245.06
316B 0.018 0.52 1.45 0.001 0.001 14.0 19.7 2.53 0.0050 31.2 263.26
316C 0.019 0.50 1.46 0.002 0.001 12.0 18.4 2.52 0.0077 28.4 229.01
316D 0.018 0.49 1.49 0.002 0.001 13.4 17.5 2.52 0.0060 29.2 228.35
316E 0.018 0.52 1.45 0.001 0.001 11.0 18.5 2.52 0.0047 27.4 218.53

Note: Nicq ¼ Ni þ 0:65Cr þ 0:98Mo þ 1:05Mn þ 0:35Si þ 12:6C; Md30 ¼ 413 2 462ðC þ NÞ 2 9:2Si 2 8:1Mn 2 13:7Cr 2 9:5Ni 2 18:5Mo.
Downloaded by [Texas A & M International University] at 01:39 08 September 2015

2. Test specimens and experimental methods with a bcc structure in the weld metal was measured
2.1 Test specimens by X-ray diffraction. Furthermore, the fracture surface
was inspected using a scanning electron microscope
Ten-millimetre thick steel sheets based on SUS308L and
(SEM).
SUS316L, with varied contents of Cr and Ni, as shown in
Table 1, were used as test specimens. In the laboratory,
these were prepared by the metal being melted in vacuo on 3. Experimental results
a 17-kg scale, forged and rolled before being solid-solution 3.1 Microstructure of the weld metal
heat-treated at 1323 K £ 1.8 ks. Table 1 shows the Ni
An example of the microstructure of the weld metal is
equivalence [11], expressing the effect of alloy elements
shown in Figure 2. Regardless of their chemical
on their respective austenite phase stability found by
compositions, in all weld metals there is a phase other
Equation (1), and the Md30 value, expressing the
than the matrix found at the dendrite boundaries.
temperature (8C) at which 50% of martensite from
Differences in the quantity of this produced due to their
austenite by 30% working found from Equation (2) [12].
chemical composition are also observed. Table 2 shows
Nicq ¼ Ni þ 0:65Cr þ 0:98Mo þ 1:05Mn þ 0:35Si

þ 12:6C; ð1Þ

Md30 ¼ 413 2 462ðC þ NÞ 2 9:2Si 2 8:1Mn

2 13:7Cr 2 9:5Ni 2 18:5Mo: ð2Þ

2.2 Experimental method

Figure 1. Schematic illustration of test specimen for SSRT.
A V-shaped groove, depth 2 mm and angle 908, was
formed by a machine working in the top surface of the
specimen, and bead-on-plate welded without the use of
filler by plasma welding. The heat input was approxi-
mately 1.5 kJ/mm and pure argon was used for both shield
gas and centre gas.
A horizontal cross-section specimen, with weld metal
at its centre, was taken from the welded joint thus
obtained, the microstructure of the weld metal was
exposed using grinding and etching and this was inspected
by optical microscopy, and the d ferrite ratio in the weld
metal was found by image analysis. A plate-shaped tensile
test specimen with a parallel central part taken
mechanically from the welded joint, as shown in Figure 1,
was used in a slow strain rate test (SSRT). Tests were
performed at room temperature in a 45-MPa hydrogen gas
environment and the strain rate was 3 £ 1026 s21.
After SSRT, the microstructure of the weld metal was
also similarly exposed and inspected and the proportion Figure 2. Microstructures of weld metals.
 924 H. Hirata et al.

Table 2. Results of EPMA analysis (mass %).

Heat Analysis location Ni Cr Mo

308A Matrix 9.75 17.73 –
Phase observed at dendrite boundary 5.14 22.69 –
308B Matrix 13.46 19.86 –
Phase observed at dendrite boundary 7.59 25.38 –
308C Matrix 11.61 19.11 –
Phase observed at dendrite boundary 7.66 23.01 –
308D Matrix 12.74 18.29 –
Phase observed at dendrite boundary 9.11 22.32 –
308E Matrix 9.43 19.31 –
Phase observed at dendrite boundary 8.78 20.34 –
316A Matrix 15.24 18.45 1.92
Phase observed at dendrite boundary 10.56 24.59 5.37
316B Matrix 15.04 20.29 2.43
Phase observed at dendrite boundary 9.78 24.38 2.18
316C Matrix 13.23 18.24 2.18
Downloaded by [Texas A & M International University] at 01:39 08 September 2015

Phase observed at dendrite boundary 8.71 22.23
316E Matrix 11.86 18.59
Phase observed at dendrite boundary 10.55 20.62
the results of a quantitative analysis made using an
electron-probe micro-analyser (EPMA) of the quantities of
Cr, Ni and Mo in the matrix and at the dendrite boundaries.
Cr and Mo were at a higher density and Ni at a lower
density, in the phases observed at the dendrite boundaries
than in the matrix. It was judged from these findings that
this was d ferrite produced during solidification [13].
3.2 Hydrogen embrittlement susceptibility of weld
metals
The relationship between the ratio of reduction of area in
atmosphere to the reduction of area in a high-pressure
hydrogen gas environment (hereafter relative reduction of
area) found by SSRT and the d ferrite ratio in the weld
metal is shown in Figure 3. When the d ferrite ratio
increases, almost no decrease in relative reduction of area
was noted in SUS316L weld metal whereas a decrease,
albeit very slight, in relative reduction of area was noted in
SUS308L. Also when the content of d ferrite ratio is the
same, the relative reduction of area in SUS308 weld metal
was smaller than that in SUS316. Figure 4 shows the
Figure 3. Effect of d ferrite fraction in weld metal on relative
reduction of area evaluated by SSRT.
3.4 Figure 4. SEM images of fractured surface of test specimens
2.53 after SSRT in high-pressure H2 gas.
3.09
results of inspection of fractured surfaces after SSRTs in a
high-pressure hydrogen gas environment of SUS308L
weld metal and SUS316L weld metal, with the same d
ferrite ratio of approximately 10%. Figure 5 shows the
results of a microscopic examination of fracture surfaces
in the SSRT test fracture surfaces of SUS308L weld metal
and SUS316L weld metal with the same d ferrite ratio of
approximately 7%. Inspection of the fracture surfaces
showed that SUS316L weld metal, in which almost no
relative reduction of area was noted, underwent full
deformation before fracture and showed a dimpled
appearance over the entire fracture surface. Fracture
surface inspection also showed that the principal cracks
did not selectively propagate in ferrite but also penetrated
austenite and no secondary cracking propagation was
observed. By contrast, although the d ferrite content was
the same, SUS308L weld metal, in which there was
marked relative reduction of area, quasi-cleavage-type
fracture surface, unaccompanied by deformation, was
observed. However, although inspection of the fracture
surface showed initiation of many secondary cracks, and
the propagated cracks penetrated into both d ferrite and
austenite, and no clear difference with SUS316L weld
metal was noted. Figure 6 shows Ni equivalence and Md30
values, as found in Equations (1) and (2), plotted against
relative reduction of area. A good correlation was
observed between these parameters, which show the
stability of austenite, and relative reduction of area. When
Ni equivalence exceeded 28% or the Md30 value was
lower than 2 308C, there was almost no reduction in the
fracture area reduction value in the high-pressure
hydrogen gas environment.
These results indicate that, within the scope of this
study, hydrogen embrittlement susceptibility is more
strongly affected by the stability of austenite during
working than by the d ferrite content in weld metal during
solidification and the basic composition may also be
thought to affect this.
Figure 7 shows the results of hardness values obtained
by microVickers tests of the fracture surface region for the
fracture of a test specimen during SSRTs of SUS308L
 Welding International
Downloaded by [Texas A & M International University] at 01:39 08 September 2015
Figure 5. Optical microstructures of cross section near
fractured surface of test specimens after SSRT in high-pressure
H2 gas.
weld metal in a high-pressure hydrogen gas environment.
Whereas the d ferrite in the unmodified solidified weld
metal had a hardness of around 230 HV, the substrate
around it had a higher hardness of around 280 HV.
Figure 8 shows the proportion of a phase with a bcc
structure contained in the weld metal after SSRT, found by
the X-ray diffraction method, plotted against the Md30
values as found by Equation (2). The figure also shows the
d ferrite ratio contained in the unmodified solidified weld
metal before the SSRT test, as shown in Figure 3. The
results of measuring the d ferrite ratio by X-ray diffraction
(a)
(b)
Figure 6. Effects of Ni equivalent and Md30 value of weld
metal on relative reduction of area evaluated by SSRT.
925
Figure 7. Vickers hardness value measured at locations near
fractured surface of test specimens after SSRT in high-pressure
H2 gas.
of one part of the weld metal before SSRT broadly
matched the results of deriving this by image analysis.
When the Md30 was small, no difference was observed
between the d ferrite ratio and the bcc phase ratio before
and after SSRT. When the Md30 was larger, however, the
bcc phase ratio after SSRT increased more than the d
ferrite ratio before SSRT and the degree of increase was
greater.
Nakamura et al. [14] have shown that a0 -martensite is
produced at the fatigue crack tip in SUS304, a metastable
austenite stainless steel, in a high-pressure hydrogen
environment. It is also known that the martensite has a bcc
structure rather than a bct structure, when the C content is
small [15]. If these reports are considered in conjunction
with the results shown in Figure 8, it is evident that the
differences between the bcc phase ratio and d ferrite ratio
before and after SSRT are due to a0 -martensite produced
in the fracture region rather than strain. It is known that, in
addition to a0 -martensite, 1-martensite, which has an hcp
structure, is also produced in the martensite induced by
working [15], but the results of X-ray diffraction tests of
the fracture region show no peaks indicating its presence.
Accordingly, the difference between the bcc phase
ratio and the d ferrite ratio before and after SSRT was
defined as the a0 -martensite ratio and plotted against the
relative reduction of area by SSRT. The results are shown
in Figure 9. It was judged that a0 -martensite was not
produced when the difference between the bcc phase ratio
Figure 8. Effect of Md30 value on fraction of bcc/d ferrite phase
before and after SSRT.
 926 H. Hirata et al.
Figure 9. Relationship between estimated fraction of strain
induced martensite in weld metal and relative reduction of area
Downloaded by [Texas A & M International University] at 01:39 08 September 2015
evaluated by SSRT.
and the d ferrite ratio before and after SSRT was negative
and the a0 -martensite ratio was plotted as 0. A correlation
was observed between the a0 -martensite ratio and the
relative reduction of area and the greater the a0 -martensite
ratio, the more marked was the embrittlement in a high-
pressure hydrogen gas environment.
From the above findings, it is clear that the hydrogen
embrittlement susceptibility of weld metal was strongly
affected by a0 -martensite. It was also found that the effect
of d ferrite is smaller than that of a0 -martensite. It was also
judged that, because the SUS308L weld metal had poorer
austenite stability than SUS316L weld metal, a0 -
martensite is produced and hydrogen embrittlement
susceptibility increased by working at room temperature.
4. Discussion of the effect of microstructure on the
hydrogen embrittlement susceptibility of weld metal
As stated above, within the scope of this study, the effect
of d ferrite produced during solidification on the hydrogen
embrittlement susceptibility of weld metal is small.
By contrast, when an a0 -martensite phase is produced,
despite the fact that, like d ferrite, it has a bcc structure, the
hydrogen embrittlement susceptibility is increased. The
reasons are considered as described below.
It is known that hydrogen embrittlement susceptibility
in ferrite steel with a bcc structure is dependent on its
hardness, and the hydrogen embrittlement susceptibility is
higher in materials with a greater hardness [16,17].
As shown in Figure 7, the results of hardness measurement
and examination of the microstructure of the region of the
fracture in the specimen fractured during a SSRT showed
that a structure thought to be a0 -martensite produced by
working was harder than d ferrite. Furthermore, although
experimental proof for this remain a matter for the future,
because hydrogen which causes embrittlement is present
in solid solution in greater quantities in austenite than in
ferrite [7], it is thought probable that the a0 -martensite
produced by the working transformation of austenite,
which contains hydrogen in solid solution, contains a
supersaturation of hydrogen in solid solution greater than
the solid solution limit of hydrogen in ferrite in an
equilibrium state. This results in a0 -martensite being a
hard structure, which occludes hydrogen and thus has
greater hydrogen embrittlement susceptibility than d
ferrite despite having a similar bcc structure.
In addition, unlike d ferrite which is dispersed in the
weld metal, it is thought that when cracking initiated by a0 -
martensite propagates within austenite, cracking develops
continuously at the crack tip due to occurrence of strain
concentration. This results in a0 -martensite being control-
ling for hydrogen embrittlement susceptibility in weld
metal, with marked embrittlement occurring due to its
production, whereas the effect of d ferrite is negligible.
However, when the production of a0 -martensite at the
crack tip is considered, this means that martensite is
constantly present at the crack tip and it is difficult to
explain the increase in hydrogen embrittlement suscepti-
bility by an increase in the average a0 -martensite ratio as
shown in Figure 9. The reason for this is may be
understood as follows: the quantity of a0 -martensite
produced by working is dependent on the degree of strain
and working temperature [5]. This result may be said to
suggest that when the test temperatures are the same, there
is a0 -martensite transformation at a strain as small as that
for the austenite phase which has a high Md30 value. Due
to this, a0 -martensite is formed with a small strain at the
crack tip in weld metal according to the height of the Md30
value, and this causes embrittlement so that the hydrogen
embrittlement susceptibility of the weld metal increases
with an increase in the average a0 -martensite ratio after
fracture.
5. Conclusion
(1) Within the range of chemical compositions in this
study, d ferrite in weld metal has a slight effect on
hydrogen embrittlement susceptibility.
(2) In the case of weld metal with high hydrogen
embrittlement susceptibility, a0 -martensite is pro-
duced due to strain being induced by SSRT. There is a
correlation between the a0 -martensite ratio and the
hydrogen embrittlement susceptibility of weld metal,
with hydrogen embrittlement susceptibility increas-
ing with increase in the quantity produced.
(3) The reason that the effect of a0 -martensite on
hydrogen embrittlement susceptibility is more
marked than that of d ferrite is that a0 -martensite is
metallurgically more hydrogen embrittlement sus-
ceptible than d ferrite and also that it is thought to be
produced continuously due to strain concentration at
the crack tip.
Acknowledgement
This study contains a part carried out as a Hydrogen Society-
building Joint Basic Development Project of the New Energy and
Industrial Technology Development Organisation (NEDO).
 Welding International
References
1. Yamada T, Kobayashi H. Criteria for materials used in
hydrogen station equipment. J Inst Safety High Pressure Gas
Eng. 2012;49(10):29 – 37 (in Japanese).
2. Gray HR. Hydrogen embrittlement testing. ASTM STP543.
1974.
3. Jewett RP. Hydrogen environment embrittlement of metals.
NASA contractor report 2163. 1973.
4. Omura T, Kobayashi K, Miyahara M, Kudo T. Hydrogen
embrittlement properties of stainless steels in high pressure
gaseous hydrogen environment. Zairyo-to Kankyo.
2006;55(4):139 –145 (in Japanese).
5. Fukuyama S, Zhang L, Wen M, Yokogawa K. Tensile
properties of SUS304 stainless steel in high pressure
hydrogen at room temperature. J Jpn Inst Met.
2003;67(4):157 –160 (in Japanese).
6. Fukuyama S, Sun D, Zhang L, Wen M, Yokogawa K. Effect
of temperature on hydrogen environment embrittlement of
Downloaded by [Texas A & M International University] at 01:39 08 September 2015
type 316 series austenitic stainless steels at low tempera-
tures. J Jpn Inst Met. 2003;67(9):456– 459 (in Japanese).
7. Fukuyama S, Zhang L, Yokogawa K. Development of
materials testing equipment in high pressure hydrogen and
hydrogen environment embrittlement of austenitic stainless
steels. J Jpn Inst Met. 2004;68(2):62 – 65 (in Japanese).
8. Ando S, Kikuta Y, Kimura Y. Modern welding technology,
Series 5. Sanpo Syuppan; 1980 (in Japanese).
927
9. Brooks JA, West AJ. Hydrogen induced ductility losses in
austenitic stainless steel weld. Metall Trans A.
1981;12A(2):213– 223.
10. Luppo MI, Hazarabdian A, Ovejero-Garcia J. Effects of
deltaferrite on hydrogen embrittlement of austenitic stainless
steel welds. Corros Sci. 1999;41:87 – 103.
11. Hirayama T, Ogirima M. Influence of chemical composition
on martensitic transformation in Fe – Cr– Ni stainless steel.
J Jpn Inst Met. 1970;34(5):507 – 510 (in Japanese).
12. Hasegawa M, editor. Stainless steel handbook. Nikkan
Kogyo Shinbunsya; 1974.
13. Nishimoto K, Natsume S, Ogawa K, Matsumura O. Welding
anthology 11: welding of stainless steel. Sanpo Syuppan;
2001 (in Japanese).
14. Nakamura J, Miyahara M, Omura T, Semba H, Wakita M.
Fatigue properties and degradation mechanism for stainless
and high strength steels in high pressure gaseous hydrogen
environment. J Soc Mater Sci. 2011;60(12):1123– 1129 (in
Japanese).
15. Nishiyama Z. Transformable martensite, the basic collec-
tion. 16. Maruzen; 1971 (in Japanese).
16. Hinotani S, Terasaki F, Takahashi K. Hydrogen embrittle-
ment of high strength steels in high pressure hydrogen gas at
ambient temperature. Tetsu-to-Hagane.
1978;64(7):899– 905 (in Japanese).
17. Kushida T. Tokyo University graduation dissertation. (in
Japanese).