Materials Research Express

ACCEPTED MANUSCRIPT • OPEN ACCESS

Effect of hydrogen embrittlement and non-metallic inclusions on tensile

fracture properties of 55CrSi spring steel
To cite this article before publication: Na Li et al 2020 Mater. Res. Express in press https://doi.org/10.1088/2053-1591/ab8b19

Manuscript version: Accepted Manuscript

Accepted Manuscript is “the version of the article accepted for publication including all changes made as a result of the peer review process,
and which may also include the addition to the article by IOP Publishing of a header, an article ID, a cover sheet and/or an ‘Accepted
Manuscript’ watermark, but excluding any other editing, typesetting or other changes made by IOP Publishing and/or its licensors”

This Accepted Manuscript is © 2020 The Author(s). Published by IOP Publishing Ltd.

As the Version of Record of this article is going to be / has been published on a gold open access basis under a CC BY 3.0 licence, this Accepted
Manuscript is available for reuse under a CC BY 3.0 licence immediately.

Everyone is permitted to use all or part of the original content in this article, provided that they adhere to all the terms of the licence
https://creativecommons.org/licences/by/3.0

Although reasonable endeavours have been taken to obtain all necessary permissions from third parties to include their copyrighted content
within this article, their full citation and copyright line may not be present in this Accepted Manuscript version. Before using any content from this
article, please refer to the Version of Record on IOPscience once published for full citation and copyright details, as permissions may be required.
All third party content is fully copyright protected and is not published on a gold open access basis under a CC BY licence, unless that is
specifically stated in the figure caption in the Version of Record.

View the article online for updates and enhancements.

This content was downloaded from IP address 178.171.20.254 on 21/04/2020 at 02:28

Page 1 of 22 AUTHOR SUBMITTED MANUSCRIPT - MRX-121064.R1

1
2
3
4 Effect of Hydrogen Embrittlement and Non-metallic Inclusions on
5
6
7 Tensile Fracture Properties of 55CrSi Spring Steel
8

pt
9 Na Li1, Wei Wang1,2, Qimin Liang1
10
11
12 Abstract
13

cri
14
The tensile fracture behavior of 55CrSi spring steels were investigated. The results
15
16
17 demonstrate that interior inclusion and hydrogen level has a significant effect on
18
19
20 ductility and a minimal effect on tensile strength of the spring steel. It was due to the

us
21
22 effect of cracking from MgO-Al2O3 spinel inclusion or the inclusions with a mixture
23
24
25 of CaO, SiO2 and part of Al2O3 due to hydrogen. The results of SEM showed that the
26
27
28
29
an
ductility reduction is connected with the formation of ‘‘fisheye’’ which formed under

30 the influence of mobile hydrogen. For the specimen containing MgO-Al2O3 spinel
31
32
inclusion, the fracture surface in the “fisheye” area is mainly composed of three
dM
33
34
35 regions including typical quasi-cleavage mixed intergranular fracture, dimple mixed
36
37
38 transgranular fractures and ductile fracture from the interior to the edge, whereas there
39
40
is no obvious transition zone from brittle fracture to dimple fracture in the “fisheye”
41
42
43 area of the specimen containing inclusions with a mixture of CaO, SiO2 and part of
pte

44
45
46 Al2O3.
47
48 Keywords: Hydrogen embrittlement; Non-metallic inclusion; Fisheye; Quasi-cleavage
49
50
51
ce

52 1. Introduction
53
54
55 55CrSi is an alloyed spring steel used to produce quality springs for suspension,
56
Ac

57
58
59
1
School of Materials Engineering, Shanghai University of Engineering Science, Shanghai 201620, China
2
Department of Nuclear Engineering and Radiological Science, University of Michigan, Ann Arbor, MI
60
48109-2104, USA
 AUTHOR SUBMITTED MANUSCRIPT - MRX-121064.R1 Page 2 of 22

1
2
3
4 valve and coil products [1]. Due to the complex stress load during service, ductility
5
6
7 and yield strength are required for the spring steel [2-4]. however, the ductility and
8

pt
9 yield strength are affected with the cleanliness to the spring steel [5-7]. The main
10
11
12 factors to tensile failure are the local stress caused by nonmetallic inclusions, which
13

cri
14
contains the size, type and composition of inclusions, and difference in thermal
15
16
17 coefficients of expansion between the steel and inclusions [8,9]. Inclusions with poor
18
19
20 deformability can induce cracks easily at the interfaces of the inclusions/steel matrix

us
21
22 during hot rolling [10-12].
23
24
25 The major types of nonmetallic inclusions are oxides, sulfides and nitrides in the
26
27
28
29
an
spring steel. In most cases, Al2O3, Al2O3-CaO or Al2O3-SiO2-CaO-MgO, MnS, TiN or

30 CaS inclusions will probably be present in commercial spring steel products [5,13-15].
31
32
The entirely removal of such inclusions is often impossible due to ordinary
dM
33
34
35 manufacturing processes, it is desirable to control the kinds of nonmetallic inclusions
36
37
38 and amount within trace limits that mechanism properties of the steel are nor
39
40
relatively affected by nonmetallic inclusions.
41
42
43 As for the spring steel, the researches are focused mainly on increasing the
pte

44
45
46 strength and fatigue properties, and developing the ductility. In general, the
47
48 susceptibility to hydrogen embrittlement increases with strength level increasing. It
49
50
51 has been recognized that trace hydrogen in the steels may lead to hydrogen
ce

52
53
embrittlement [15]. Hydrogen embrittlement can reduce simultaneously strength,
54
55
56 toughness and ductility of the steels. Alternatively, hydrogen embrittlement can also
Ac

57
58
59 cause a significant reduction of ductility without decreasing in yield and tensile
60
Page 3 of 22 AUTHOR SUBMITTED MANUSCRIPT - MRX-121064.R1

1
2
3
4 strength [16].
5
6
7 The solubility of hydrogen in the steels can be classified in two different
8

pt
9 categories. In generally, hydrogen in steel has a low solubility. However, hydrogen in
10
11
12 traps has very high solubility [17]. There are two kinds of hydrogen traps in the steel.
13

cri
14
The first kind is weak traps. The binding energy between weak traps and hydrogen is
15
16
17 smaller. Therefore, hydrogen can escape from the steel matrix. Another kind is strong
18
19
20 traps. Internal inclusions belong to strong traps. The binding energy between strong

us
21
22 traps and hydrogen is higher. Hydrogen trapped by these strong traps can hardly
23
24
25 escape from the steel matrix [6].
26
27
28
29
an
Lovicu et al. [18] reported that nonmetallic inclusions has a greater contribution

30 to hydrogen sensitivity in advanced high-strength steels. Studies show that the

31
32
hydrogen cracking crack initiated at inclusions inside the tensile specimen [19,20].
dM
33
34
35 Furuya et al. [21] proposed that hydrogen is segregated at voids which formed around
36
37
38 inclusions. High concentration of hydrogen can induce voids growing and initiate
39
40
cracking under tensile loading.
41
42
43 Different inclusion rating methods have been suggested. Cleanness of the steels
pte

44
45
46 can be remarkably improved by hydrogen embrittlemant of a tensile test specimen
47
48 than the methods by fatigue testing or optical microscopy [22].
49
50
51 Generally speaking, nonmetallic inclusions play the major role in hydrogen
ce

52
53
embrittlement, however, the interaction of nonmetallic inclusions and hydrogen is not
54
55
56 yet clear in detail. The aim of this study is to further investigate the characterization of
Ac

57
58
59 nonmetallic inclusions in spring steel based on a detailed microscope of the tensile
60
 AUTHOR SUBMITTED MANUSCRIPT - MRX-121064.R1 Page 4 of 22

1
2
3
4 fracture surfaces. Accordingly, clarify the effect of the type and composition of
5
6
7 nonmetallic inclusion on hydrogen embrittlement of spring steels. The fundamental
8

pt
9 understanding of this question is essential to improve the final properties of the
10
11
12 springs.
13

cri
14
2. Experiments and materials
15
16
17 The investigated material is a high strength spring steel 55CrSi. The specimens
18
19
20 were supplied by a steel company as wires. The chemical composition of the spring

us
21
22 steel is given in Table 1. The spring steel 55CrSi used here was hot-rolled wire rods.
23
24
25 After acid pickling in 40 vol% NaCl for 10 min, phosphating and drying, the wire
26
27
28
29
an
rods were cold drawn from 12 mm to 4.0 mm and 4.7 mm in diameter (Φ4.0 and

30 Φ4.7), respectively. The specimens were heated up to 860 °C for 30 min, and then oil
31
32
quenched and tempered at 430 °C for 60 min followed by air cooling. A ONH-836
dM
33
34
35 analyzer (LECO, America) was used to measure the O and H contents of specimens
36
37
38 Φ4.0 and Φ4.7. Prior to the measurement, all samples were polished to remove the
39
40
oxides on the surface of the sample. Tensile tests were operated at room temperature
41
42
43 using a CMT5105 universal materials tester machine, operating at a speed of 4
pte

44
45
46 mm/min.
47
48 Table 1. Chemical composition of 55CrSi spring steel (wt%)
49
50 Elements C Si Cr Mn P S Cu
51
ce

52 Composition 0.55 1.5 0.7 0.7 ≤0.03 ≤0.03 ≤0.2

53
54 The microstructure and tensile fracture surface of the specimens were
55
56 investigated using a scanning electron microscope (SEM) on a S3400-N instrument to
Ac

57
58
59 identify the morphologies of the fractures. Samples for SEM were cut from the
60
Page 5 of 22 AUTHOR SUBMITTED MANUSCRIPT - MRX-121064.R1

1
2
3
4 exposed core of the tensile test specimens and prepared by mechanical polishing and
5
6
7 further ultrasonically cleaned in 99% alcohol for 30 min, then dried by the inert
8

pt
9 atmosphere. The sizes and composition of nonmetallic inclusions were measured
10
11
12 using SEM and energy dispersive spectroscopy (EDS), respectively.
13

cri
14
15
16 3. Results and discussion
17
18 3.1 Analysis of SEM microstructure
19
20

us
21 SEM micrographs in Figs. 1a and 1b show the microstructure of both the
22
23
24 heat-treated specimens. Typical tempered martensitic structure (sorbite structure) can
25
26 be seen in both specimens. As shown in Fig. 1, the microstructure of both the
27
28
29
an
specimens consists of lath martensite with prior austenite grains，prior austenite grain
30
31 boundaries are shown as the red dashed line in Fig. 1a and 1b. The substructure
32
dM
33
34 consists of spherical cementite particles. From Fig. 1, it can obviously seen that there
35
36
37 is no significant difference among both the specimens. It is known that prior austenite
38
39 grain boundary and lath martensite is susceptible to hydrogen embrittlement because
40
41
42 they can act as the path of crack propagation, and accelerate the rate of crack
43
pte

44 propagation [22,23].
45
46
47
48
49
50
51
ce

52
53
54
55
56
Ac

57
58
59 Fig. 1. SEM micrographs of 55CrSi spring steel after heat treatment
60
 AUTHOR SUBMITTED MANUSCRIPT - MRX-121064.R1 Page 6 of 22

1
2
3 (a) Φ4.0 specimen and (b) Φ4.7 mm specimen
4
5 3.2 O, N and H contents of specimens Φ4.0 and Φ4.7 and mechanical properties
6
7
8 The H and O contents of specimens Φ4.0 and Φ4.7 are shown in Table 2. It can

pt
9
10
11 be seen that the H and O contents of specimens Φ4.0 are 2.86 ± 0.239 ppm and 96.7 ±
12
13 4.88 ppm, and the H and O contents of specimens Φ4.7 are 6.19± 0.239 ppm and 161

cri
14
15
16 ± 4.88 ppm, respectively.
17
18 Table 2. O, N and H contents of specimens Φ4.0 and Φ4.7
19
20
Sample Weight H content O content N content

us
21
22 Blank testing 0.239 ppm 4.88 ppm 5.23 ppm
23 Φ4.0mm 0.2960g 2.86 ppm 96.7 ppm 112 ppm
24
Φ4.7mm 0.3298g 6.19 ppm 131 ppm 242 ppm
25
26
The mechanical properties of specimens Φ4.0, Φ4.7 and normal specimens are
27
28
an
29 shown in Table 3. It can be seen that Φ4.0 and Φ4.7 specimens exhibit relatively high
30
31
32 tensile strength, 1900 MPa for Φ4.0 specimen, and 1830 MPa for Φ4.7 specimen,
dM
33
34 respectively. However, the average values of elongation in specimens Φ4.0 and Φ4.7
35
36
37 are 27.6% and 20.7%, respectively. It is obvious that elongation of both the specimens
38
39
40
is reduced, as compared to its normal specimen (≥40%). Meanwhile, tables 2 and 3
41
42 show the changes in strength and ductility of specimens Φ4.0 and Φ4.7 charged with
43
pte

44
45 various H contents. It is obvious that as hydrogen concentration increases, the tensile
46
47 strength and ductility of the specimens decreases.
48
49
50 The content of oxygen really represents the level of oxide inclusions in the steel,
51
ce

52
53
so total oxygen is a very important and common index of steel cleanliness [24]. As
54
55 shown in Table 2, it can be seen that Φ4.0 and Φ4.7 specimens contain high content of
56
Ac

57
58 oxygen. The excessive oxygen content resulted in the increase of inclusion quantity,
59
60 which would deteriorate the cleanliness of the steel.
Page 7 of 22 AUTHOR SUBMITTED MANUSCRIPT - MRX-121064.R1

1
2
3
4 Based on the experimental data, it should be noted that hydrogen concentration
5
6
7 and the impurity have a significant effect on ductility and a minimal effect on tensile
8

pt
9 strength of the spring steel.
10
11
12 The Vickers hardness obtained for Φ4.0 and Φ4.7 specimens are HV 563 and HV
13

cri
14
537, respectively. The Vickers hardness has an important role in deciding the
15
16
17 susceptibility to hydrogen embrittlement [25]. It is proposed that Vickers hardness of
18
19
20 high strength steels is over HV400, hydrogen embrittlement occurs more markedly

us
21
22 [21].
23
24 Table 3 Mechanical properties of 55CrSi spring steel and the specimens investigated after heat
25
treatment
26
27
28
29
Symbol
Tensile Strength
σB (MPa)
an Elongation
Ψ （Ptc）
Vickers
Hardness (HV)
Indicators 1841~1961 ≥40 400
30
31 Φ4.0mm 1900 27.6 563
32 Φ4.7mm 1830 20.7 537
dM
33
34 3.3 Analysis of SEM micrographs and composition of nonmetallic inclusions
35
36
37 Fig. 2a shows the morphologies of the tensile fracture surfaces of the Φ4.0
38
39
specimen. A specific defect, known as “fisheye” can be observed. The ‘‘fisheye’’ is
40
41
42 seen in SEM as elliptic shape. In a backscattered electron contrast, the “fisheye”
43
pte

44
45 appears as bright areas, as shown in Fig. 2b. Many researchers have reported that
46
47 “fisheye” was often found on the tensile fracture specimens after hydrogen
48
49
50 pre-charging, and implying that the hydrogen atoms has diffused to the center of the
51
ce

52
cross-sectional area [26-28].
53
54
55
56
Ac

57
58
59
60
 AUTHOR SUBMITTED MANUSCRIPT - MRX-121064.R1 Page 8 of 22

1
2
3
4
5
6
7
8

pt
9
10
11
12
13

cri
14
15
16
17
18
19
20

us
21
22
23
24 Fig. 2. SEM image of the tensile fracture surface with fisheye and inclusion for the Φ4.0 specimen
25
(a) fracture surface morphology of the internal “fisheye”; (b) back scatter electron (BSE) image
26
27
28
29
an
of fracture surface of the “fisheye”; (c) interior hole in the center of the “fisheye”; (d) interior
inclusion in the center of the “fisheye” (counterpart on the matching fracture side of Fig. 2(c)).

30 Two magnified images of the white spot in Fig. 2a and dark spot in Fig. 2b are
31
32
shown in Figs. 2c and 2d, respectively. Detailed views of a “fisheye” region and of its
dM
33
34
35 counterpart on the matching side can be seen in Figs. 2c and 2d, respectively. From
36
37
38 the Fig. 2c, it can be observed that a rectangular hole is surrounded by a mixture of
39
40
“flat” and “quasi-cleavage” fracture features. On the fracture surface of counterpart,
41
42
43 an inclusion is observed in the center of the internal “fisheye”. The shape of inclusion
pte

44
45
46 is rectangular and the diagonal length is about 29 µm, as shown in Fig. 2d. Meanwhile,
47
48 Fig. 2c shows brittle fracture with four deep cracks. It is clear that these cracks
49
50
51 originate at inclusions and propagate in a transgranular manner as evidenced by Fig.
ce

52
53
2d. The EDS of the inclusion analysis was carried out and the result is shown in Fig. 3.
54
55
56 It reveals that the inclusion contains Al, Mg and O elements, designated as
Ac

57
58
59 MgO-Al2O3 spinel inclusion.
60
Page 9 of 22 AUTHOR SUBMITTED MANUSCRIPT - MRX-121064.R1

1
2
3
4
5
6
7
8

pt
9
10
11
12
13

cri
14
15
16
17 Fig. 3. Composition of the inclusion in the center of the “fisheye” in Fig. 2(d)
18
19
20 Further observation shows that “fisheye” is mainly composed of three regions

us
21
22 including crack initiation(region I), crack propagation (region II) and final fracture
23
24
25 (region III), as shown in Fig.4a. Fig. 4b presents some details of the fracture initiation
26
27
28
29
an
region at higher resolution than Fig. 2d. Details on fracture surface in the region I

30 show a mixture feature of “flat” and “quasi-cleavage”. Fine straight and Y-shaped tear
31
32
ridges can be observed on the “flat” region and “quasi-cleavage” facets, as illustrated
dM
33
34
35 by arrows in Fig. 4b. The magnified image of a white box “A” in Fig. 4a is shown in
36
37
38 Fig. 4c which indicating the formation of serrated markings. It can be observed that
39
40
the “quasi-cleavage” fracture surface has lath microstructure within serrated markings,
41
42
43 as shown by arrows in Fig. 4c. In addition, Fig. 4c shows the secondary cracks
pte

44
45
46 (shown with the red dashed line), which propagate along prior austenite grain
47
48 boundaries. It was reported [29-31] that secondary crack is closely related to
49
50
51 hydrogen segregation along the prior austenite grain boundaries. These features
ce

52
53
basically indicate the occurrence of mixed fracture surface of intergranular and
54
55
56 quasi-cleavage in the region I. The intergranular fracture is explained by the
Ac

57
58
59 hydrogen-enhanced decohesion mechanism that dissolved hydrogen weakens the
60
 AUTHOR SUBMITTED MANUSCRIPT - MRX-121064.R1 Page 10 of 22

1
2
3
4 bonding of Fe lattice atoms [32]. Du proposed that a critical strain demanded for
5
6
7 intergranular fracture is decreased due to hydrogen reduces the bonding of Fe atoms at
8

pt
9 grain boundaries [33].
10
11
12 The fracture surface outside of region I (region II) appears transgranular ductile
13

cri
14
fracture with tear ridges, as shown in the bottom right of Fig. 4c and Fig. 4d. The
15
16
17 region between region I and II exhibits the step change of roughness. This transition
18
19
20 border is marked as dotted line in Fig. 4c. As seen in Fig. 4d, some voids are

us
21
22 elongated which indicates ductile shear fracture has occurred in this region. Obvious
23
24
25 shear and tear dimples together with a few equiaxed dimples are observed. These
26
27
28
29
an
features basically indicate the occurrence of mixed fracture of ductile shear and

30 transgranular fracture in the crack propagation region (region II).

31
32
dM
33
34
35
36
37
38
39
40
41
42
43
pte

44
45
46
47
48
49
50
51
ce

52
53 Fig. 4. The detail microstructure and morphology of the tensile fracture for the Φ4.0 specimen
54
55 (a) microscopic morphology of circular pits and fibers; (b) enlarged view of Fig. 2c; (c) magnified
56
Ac

57 image of "A" area in Fig. 4a; (d) magnified image of "C" area in Fig. 4c.
58
59 The area outside of the crack propagation region (region III) corresponds to the
60
Page 11 of 22 AUTHOR SUBMITTED MANUSCRIPT - MRX-121064.R1

1
2
3
4 final fracture surface. The high magnification image of a white box “B” in Fig. 4a is
5
6
7 shown in Fig. 5a which indicating the formation of waves on the region “B”, which
8

pt
9 demonstrates that ductile fracture occurred in the region “B”. The fractographic
10
11
12 features indicate that fracture occurs through micro-void initiation and ductile tearing.
13

cri
14
It was suggested that micro-void initiation originates from very closely spaced nuclei,
15
16
17 and that strain localization controls subsequent void growth [34].
18
19
20 Meanwhile, intergranular cracks along prior austenite grain boundaries can be

us
21
22 seen on the fracture surface of the specimen, as shown by arrows “1” and “2” in Fig.
23
24
25 4a. A vicinity of a crack (arrow “1” region in Fig. 4a) is shown in Fig. 5b. An enlarged
26
27
28
29
an
image of the bottom right in Fig. 5b is shown in Fig. 5c. Detailed observations near

30 the arrow “1” region reveal ductile fracture accompanied with a large number of
31
32
dimples at the edges of intergranular cracks. Dimples in different size in ridges were
dM
33
34
35 also observed in Fig. 5c. It indicates that a large number of ductile fracture occurred
36
37
38 before complete rupture by localized shear.
39
40
Moreover, an enlarged image of region (A) in Fig. 2a is shown in Fig. 5d.
41
42
43 Detailed observations near region (A) outside of the “fisheye” exhibit dimple fracture.
pte

44
45
46 The region outside of the “fisheye” corresponds to the final fast fracture surface.
47
48
49
50
51
ce

52
53
54
55
56
Ac

57
58
59
60
 AUTHOR SUBMITTED MANUSCRIPT - MRX-121064.R1 Page 12 of 22

1
2
3
4
5
6
7
8

pt
9
10
11
12
13

cri
14
15
16
17
18
19
20

us
21
22
23
24
25
26 Fig. 5. Enlarged SEM microstructure morphology of tensile fracture fibers of Φ4.0 specimen
27
28
29
(a)
an
high magnification image of a white box “B” in Fig. 4a; (b) micromorphology of cracks

30 indicated by arrow 1 in Fig. 4a; (c) enlarged image of the bottom right in Fig. 5b; (d) enlarged
31
32 image of region (A) in Fig. 2a .
dM
33
34 From the above results, it can be concluded that, for the specimen containing
35
36
37 MgO-Al2O3 spinel inclusion, the fracture surface in the “fisheye” area is mainly
38
39 composed of three regions including typical quasi-cleavage mixed intergranular
40
41
42 fracture (region near the crack origin), dimple mixed transgranular fractures and
43
pte

44 ductile fracture from the interior to the edge. The difference among three fracture
45
46
47 features could be attributed to hydrogen content at the near tip of cracks, the stress
48
49
50 condition and the rate of crack growth [35]. During the period for the crack initiation,
51
ce

52 the specimen is in a approximately plane-strain condition, the hydrogen content is

53
54
55 higher due to concentration of stress at the near tip of cracks, brittle quasi-cleavage
56
Ac

57 fracture would like to induce in the combined effect of hydrogen atom and stress
58
59
60 condition. During the period for the crack propagating, the specimen can nearly be
Page 13 of 22 AUTHOR SUBMITTED MANUSCRIPT - MRX-121064.R1

1
2
3
4 regarded as the state of plane stress. Moreover, because of the rate of the crack
5
6
7 propagation is higher than that of the crack initiation, hydrogen content at the near tip
8

pt
9 of cracks decreases due to hydrogen atom transfer, so ductile fracture prefers to occur
10
11
12 in this condition [35].
13

cri
14
Fig. 6a shows the morphologies of the tensile fracture surfaces of the Φ4.7
15
16
17 specimen. “Fisheye” is also be observed on the fracture surfaces of the Φ4.7 specimen.
18
19
20 The ‘‘fisheye’’ is seen in SEM as elliptic shape.

us
21
22 Two high magnification images of the central region in Fig. 6a are shown in Fig.
23
24
25 6b and 6c. Detailed views of a “fisheye” area and of its counterpart on the matching
26
27
28
29
an
side can be seen in Figs. 6b and 6c. In the Fig. 6b, a similar circular hole with a

30 mixture of “flat” and “quasi-cleavage” fracture features is seen. On the counterpart, an

31
32
inclusion is observed in the center of the internal “fisheye”, as shown in Fig. 6c. The
dM
33
34
35 shape of inclusion is spherical and the diameter of inclusion is about 60 μm.
36
37
38
39
40
41
42
43
pte

44
45
46
47
48
49
50
51
ce

52
53
54
55
56
Ac

57
58
59
60 Fig. 6. SEM image of the tensile fracture surface with fisheye and inclusion for the Φ4.7 specimen
 AUTHOR SUBMITTED MANUSCRIPT - MRX-121064.R1 Page 14 of 22

1
2
3
4 (a) fracture surface morphology of the internal “fisheye”; (b) interior hole in the center of the
5
6 “fisheye”; (c) interior inclusion in the center of the “fisheye” (counterpart on the matching fracture
7
8 side of Fig. 6(b)); (d) magnified image around the inclusion.

pt
9
10 The magnified image around the inclusion in Fig. 6c is shown in Fig. 6d
11
12
13
indicating the formation of serrated markings. It can be seen that the secondary cracks

cri
14
15 (shown with the red dashed line), which propagate along prior austenite grain
16
17
18 boundaries. Clearly, the secondary cracks in the Φ4.0 specimen are short and blunt,
19
20 while those in the Φ4.7 specimen are relatively deep and sharp. In addition, a high

us
21
22
23 magnification image of a white box “A” in Fig. 6a is shown in Fig. 7a, some distinct
24
25
26
embrittled cracks are revealed. An enlarged image of the left in Fig. 7a is shown in
an
27
28 Fig. 7b. Detailed observations reveal a clearly distinguishable embrittled edge along
29
30
31 the fracture surface (Fig. 7a), while the periphery of cracks was rather ductile with
32
dM
33 numerous large and shallow dimples (Fig. 7b). It should be noted that there is no
34
35
36 obvious transition zone from brittle fracture to dimple fracture in the “fisheye” area of
37
38
39
the Φ4.7 specimen in comparison with that of the Φ4.0 specimen.
40
41
42
43
pte

44
45
46
47
48
49
50
51
ce

52
Fig. 7. The detail microstructure and morphology of the tensile fracture for the Φ4.7 specimen
53
54
(a) magnified image of a white box “A” in Fig. 6a; (b) enlarged image of the left in Fig. 7a
55
56
It can be obviously seen from Fig.6c that the interfaces between the inclusion
Ac

57
58
59 and the matrix are not cohesive. In comparison with the Φ4.0 specimen, the interface
60
Page 15 of 22 AUTHOR SUBMITTED MANUSCRIPT - MRX-121064.R1

1
2
3
4 cohesive of the Φ4.7 specimen is the worst. Moreover, the EDS result in Fig. 8 shows
5
6
7 that the inclusion contains of O, Ca, Si and Al elements. It can be concluded that the
8

pt
9 inclusion was mainly composed of CaO, SiO2 and part of Al2O3.
10
11
12
13

cri
14
15
16
17
18
19
20

us
21
22
23
24
25
26 Fig. 8. The inclusion composition of the center of the “fisheye” in Fig. 6(c)
27
28
29 3.4 Summary
an
30
31 In present study, two specimens have same compositions and heat treatment
32
dM
33
34 processing. It is obvious that tensile strength and elongation of both the specimens are
35
36
37 reduced, as compared to its normal specimen. However, compared with Φ4.7
38
39 specimen, the Φ4.0 specimen exhibit higher tensile strength because of cohesive
40
41
42 MgO-Al2O3 spinel inclusion. Two specimens have different sizes (Φ4.0 mm and Φ4.7
43
pte

44 mm), but, there is least size difference between the two specimens, thus, under the
45
46
47 assumption that the influence of specimen size on mechanical properties and fracture
48
49
50 behavior may be neglected.
51
ce

52 Various kinds of inclusions is a result of the varying mechanism properties of

53
54
55 the specimens. For the inclusion with a mixture of CaO, SiO2 and part of Al2O3, the
56
Ac

57 fracture surface of the specimen showed an evident gap at the inclusion/the matrix
58
59
60 interfaces, as shown in Fig.6c. During the tensile test, the degree of deformation
 AUTHOR SUBMITTED MANUSCRIPT - MRX-121064.R1 Page 16 of 22

1
2
3
4 between the inclusion and matrix is different. It is inevitable that the concentration of
5
6
7 stress at the inclusion/the matrix interfaces will be induced, which results in the
8

pt
9 aggregation of hydrogen at the inclusion/the matrix interfaces due to hydrogen atoms
10
11
12 diffusion driven by the stress gradient-induced. Once hydrogen content exceeds the
13

cri
14
critical value for crack initiation, hydrogen induced cracking occurs immediately.
15
16
17 Inclusion with MgO-Al2O3 spinel inclusion which has a sharp shape and are tightly
18
19
20 bound to the matrix, as shown in Fig. 2d. The cohesive interfaces between the

us
21
22 inclusion and matrix may lead to a lower concentration of stress and reduce the
23
24
25 possibility of crack initiation.
26
27
28
29
an
Compared with the Φ4.0 specimen, the inclusion within the Φ4.7 specimen has

30 an evidently larger size. Furthermore, it is known that the smaller inclusion may lead
31
32
to a lower concentration of stress and reduce the possibility of crack initiation [36].
dM
33
34
35 Furthermore, the secondary cracks that initiated from the prior austenite grain
36
37
38 boundaries, as indicated in Figs. 2c and 6d, it further demonstrates a significant
39
40
aggregation of hydrogen atoms in the prior austenite grain boundaries. Thus, it can be
41
42
43 inferred that hydrogen embrittlement plays an important role in accelerating the
pte

44
45
46 initiation and growth of the secondary cracks.
47
48 4. Conclusions
49
50
51 In this study, the effect of hydrogen on the tensile strength of 55CrSi spring steels and
ce

52
53
fracture behavior were investigated. Based on the experimental results and the
54
55
56 analysis, the conclusions are drawn as follows:
Ac

57
58
59 (1) The tensile strength and elongation of 55CrSi spring steel are reduced due to the
60
Page 17 of 22 AUTHOR SUBMITTED MANUSCRIPT - MRX-121064.R1

1
2
3
4 combined effect of nonmetallic inclusion and hydrogen.
5
6
7 (2) Inclusion with MgO-Al2O3 spinel nonmetallic inclusion which has a sharp shape
8

pt
9 and are tightly bound to the matrix. For the inclusion with a mixture of CaO, SiO2 and
10
11
12 part of Al2O3, the fracture surface of the specimen showed an evident gap at the
13

cri
14
inclusion/matrix interfaces.
15
16
17 (3) For the specimen containing MgO-Al2O3 spinel inclusion, the fracture surface in
18
19
20 the “fisheye” area is mainly composed of three regions including typical

us
21
22 quasi-cleavage mixed intergranular fracture, dimple mixed transgranular fractures and
23
24
25 ductile fracture from the interior to the edge, whereas there is no obvious transition
26
27
28
29
an
zone from brittle fracture to dimple fracture in the “fisheye” area of the specimen

30 containing inclusions with a mixture of CaO, SiO2 and part of Al2O3.

31
32
Acknowledgments
dM
33
34
35 The authors gratefully acknowledge the financial support of the National Natural
36
37
38 Science Foundation of China (No. 51271111).
39
40
References
41
42
43 [1] Y. Hu, W. Chen, C. Wan, F. Wang. Effect of deoxidation process on inclusion and
pte

44
45
46 fatigue performance of spring steel for automobile suspension [J]. Metallurgical
47
48 and Materials Transactions A, 2018, 49B: 569-580.
49
50
51 [2] V.I. Zurnadzhy, V.G. Efremenko, K.M. Wu, A.Yu. Azarkhov, Yu.G. Chabak, V.L.
ce

52
53
Greshta, O.B. Isayev, M.V. Pomazkov. Effects of stress relief tempering on
54
55
56 microstructure and tensile/impact behavior of quenched and partitioned
Ac

57
58
59 commercial spring steel [J]. Materials Science and Engineering A, 2019, 745:
60
 AUTHOR SUBMITTED MANUSCRIPT - MRX-121064.R1 Page 18 of 22

1
2
3
4 307-318.
5
6
7 [3] Z.G. Yang, S.X. Li, Y.D. Li, Y.B. Liu, W.J. Hui, Y.Q. Weng. Relationship among
8

pt
9 fatigue life, inclusion size and hydrogen concentration for high-strength steel in
10
11
12 the VHCF regime [J]. Materials Science and Engineering A, 2010, 527: 559-564.
13

cri
14
[4] Y. Jiang, Y.L. Liang, C.H. Yin, S.L. Long, F. Zhao , Y. Xiao. Influence of
15
16
17 multiphase on the strain hardening behavior of 60Si2CrVAT spring steel treated
18
19
20 by a Q–P–T process [J]. Journal of Materials Science, 2018, 53: 10396-10410.

us
21
22 [5] A. Bytyqi, M. Jenko, M. Godec. Analysis of inclusions in spring steel using
23
24
25 scanning electron microscopy and Auger spectroscopy [J]. Vacuum, 2012, 86:
26
27
28
29
648-651.
an
30 [6] Y.D. Li, Z.G. Yang, Y.B. Liu, S.X. Li, G.Y. Li, W.J. Hui, Y.Q. Weng. The influence
31
32
of hydrogen on very high cycle fatigue properties of high strength spring steel [J].
dM
33
34
35 Materials Science and Engineering A, 2008, 489: 373-379.
36
37
38 [7] U. Karr, B. Schönbauer, M. Fitzka, E. Tamura, Y. Sandaiji, S. Murakami, H.
39
40
Mayer. Inclusion initiated fracture under cyclic torsion very high cycle fatigue at
41
42
43 different load ratios [J]. International Journal of Fatigue, 2019, 122: 199-207.
pte

44
45
46 [8] P.A. Thornton. The influence of nonmetallic inclusions on the mechanical
47
48 properties of steel: A review [J]. Journal of Materials Science, 1971, 6: 347-356.
49
50
51 [9] J. Lankford. Effect of oxide inclusions on fatigue failure [J]. International
ce

52
53
Materials Reviews, 1977, 22(1): 221-228.
54
55
56 [10] H.V. Atkinson, G. Shi. Characterization of inclusions in clean steels: a review
Ac

57
58
59 including the statistics of extremes methods [J]. Progress in Materials Science,
60
Page 19 of 22 AUTHOR SUBMITTED MANUSCRIPT - MRX-121064.R1

1
2
3
4 2003, 48(5): 457-520.
5
6
7 [11] Y.Murakami, M. Endo. Effects of defects, inclusions and inhomogeneities on
8

pt
9 fatigue strength [J]. International Journal of Fatigue, 1994, 16(3): 163-182.
10
11
12 [12] A.I. Gustavsson, A. Melander. Fatigue limit model for hardened steels [J].
13

cri
14
Fatigue and Fracture of Engineering Materials and Structures, 1992, 15(9):
15
16
17 881-894.
18
19
20 [13] Y. Furuya, S. Matsuoka, T. Abe. Inclusion-controlled fatigue properties of 1800

us
21
22 MPA-class spring syeels [J]. Metallurgical and Materials Transactions A, 2004,
23
24
25 35A: 3737-3744.
26
27
28
29
an
[14] W. Shi, S. Yang, A. Dong, J. Li. Understanding the corrosion mechanism of

30 spring steel induced by MnS inclusions with different sizes [J]. JOM, 2018,
31
32
70(11): 2513-2522.
dM
33
34
35 [15] S.K. Dwivedi, M. Vishwakarma. Hydrogen embrittlement in different materials:
36
37
38 A review [J]. International Journal of Hydrogen Energy, 2018, 43: 21603-21616.
39
40
[16] J. Venezuela, Q. Zhou, Q. Liu, H. Li, M. Zhang, M. S. Dargusch, A. Atrens. The
41
42
43 influence of microstructure on the hydrogen embrittlement susceptibility of
pte

44
45
46 martensitic advanced high strength steels [J]. Materials Today Communications,
47
48 2018, 17: 1-14.
49
50
51 [17] M.D. Chapetti, T. Tagawa, T. Miyata. Ultra-long cycle fatigue of high-strength
ce

52
53
carbon steels part I: review and analysis of the mechanism of failure [J].
54
55
56 Materials Science and Engineering A, 2012, 43(11): 4075-4087.
Ac

57
58
59 [18] G. Lovicu, M. Bottazzi, F. D’aiuto, M. De Sanctis, A. Dimatteo, C. Santus, R.
60
 AUTHOR SUBMITTED MANUSCRIPT - MRX-121064.R1 Page 20 of 22

1
2
3
4 Valentini. Hydrogen Embrittlement of Automotive Advanced High-Strength
5
6
7 Steels [J]. Materials Science and Engineering A, 2003, 356(1-2): 227-235.
8

pt
9 [19] G. Han and D. Feng. Effects of hydrogen on behaviour of low cycle fatigue of
10
11
12 2.25Cr-1Mo steel [J]. Journal of Materials Science and Technology, 1995, 11 (5):
13

cri
14
353-357.
15
16
17 [20] Y.Tsuchida. T. Watanabe, T. Kato, T. Seto. Effect of hydrogen absorption on
18
19
20 strain-induced low-cycle fatigue of low carbon steel [J]. Procedia Engineering,

us
21
22 2010, 2(1): 555-561.
23
24
25 [21] S. Fujita, Y. Murakami. A New Nonmetallic inclusion rating method by positive
26
27
28
29
an
use of hydrogen embrittlement phenomenon [J]. Metallurgical and Materials

30 Transactions A, 2013, 44A: 303-322.

31
32
[22] R. L. S. Thomas, J. R. Scully, R. P. Gangloff. Internal hydrogen embrittlement of
dM
33
34
35 ultrahigh-strength AERMET 100 steel [J]. Metallurgical and Materials
36
37
38 Transactions A, 2003, 34(2): 327-344.
39
40
[23] A. Nagao, C. D. Smith, M. Dadfarnia, P. Sofronis, I. M. Robertson. The role of
41
42
43 hydrogen in hydrogen embrittlement fracture of lath martensitic steel [J]. Acta
pte

44
45
46 Materialia, 2012, 60(13-14): 5182-5189.
47
48 [24] L.F. Zhang, B. G. Thomas. State of the Art in Evaluation and Control of Steel
49
50
51 cleanliness[J]. ISIJ International, 2003, 43(3): 271-291.
ce

52
53
[25] T. Das, S. K. Rajagopalan, S. V. Brahimi, X. Wang, S. Yue. A study on the
54
55
56 susceptibility of high strength tempered martensite steels to hydrogen
Ac

57
58
59 embrittlement (HE) based on incremental step load (ISL) testing methodology [J].
60
Page 21 of 22 AUTHOR SUBMITTED MANUSCRIPT - MRX-121064.R1

1
2
3
4 Materials Science and Engineering A, 2018, 716: 189-207.
5
6
7 [26] M.J. Santofimia, L. Zhao, J. Sietsma. Microstructural Evolution of a Low-Carbon
8

pt
9 Steel during Application of Quenching and Partitioning Heat Treatments after
10
11
12 Partial Austenitization [J]. Metallurgical and Materials Transactions A, 2009, 40:
13

cri
14
46-57.
15
16
17 [27] A.B. Hadzipasic, J. Malina, S. Niznik. The influence of microstructure on
18
19
20 hydrogen diffusion in dual phase steel[J].Acta Metallurgica Slovaca, 2011,

us
21
22 17(2):129-137.
23
24
25 [28] J. Yang, Y. Song, Y. Lu, J. Gu, Z. Guo. Effect of ferrite on the hydrogen
26
27
28
29
an
embrittlement in quenched-partitioned tempered low carbon steel [J]. Materials

30 Science and Engineering A, 2018, 712: 630-636.

31
32
[29] M.A.Arafin, J.A.Szpunar. Effect of bainitic microstructure on the susceptibility
dM
33
34
35 of pipeline steels to hydrogen induced cracking [J]. Materials Science and
36
37
38 Engineering A, 2011, 528(15): 4927-4940.
39
40
[30] J. Venezuela, Q. Liu, M. Zhang, Q. Zhou, A. Atrens. The influence of hydrogen
41
42
43 on the mechanical and fracture properties of some martensitic advanced high
pte

44
45
46 strength steels studied using the linearly increasing stress test [J]. Corrosion
47
48 Science, 2015, 99: 98-117.
49
50
51 [31] N.E. Nanninga, Y.S. Levy, E.S. Drexler, R.T. Condon, A.E. Stevenson, A.J.
ce

52
53
Slifka. Comparison of hydrogen embrittlement in three pipeline steels in high
54
55
56 pressure gaseous hydrogen environments [J]. Corrosion Science, 2012, 59: 1-9.
Ac

57
58
59 [32] J. Han, J. Nam, Y. Lee. The mechanism of hydrogen embrittlement in
60
 AUTHOR SUBMITTED MANUSCRIPT - MRX-121064.R1 Page 22 of 22

1
2
3
4 intercritically annealed medium Mn TRIP steel [J]. Acta Materialia, 2016, 113:
5
6
7 1-10.
8

pt
9 [33]Y.A. Du, L. Ismer, J. Rogal, T. Hickel, J. Neugebauer, R. Drautz. First-principles
10
11
12 study on the interaction of H interstitials with grain boundaries in a- and g-Fe [J].
13

cri
14
Physical Review B, 2011, 84 : 144121.
15
16
17 [34] J. J. Williams, G. Piotrowski, R. Saha, N. Chawla. Effect of overaging and
18
19
20 particle size on tensile deformation and fracture of particle-reinforced aluminum

us
21
22 matrix composites [J]. Metallurgical and Materials Transactions A, 2002, 33(12):
23
24
25 3861-3869.
26
27
28
29
an
[35] R. A. Oriani. Hydrogen embrittlement of steels [J]. Annual Review of Materials

30 Science, 1978, 8: 327-357.

31
32
[36] S. Shen, X. Song, Q. Li, X. Li, R. Zhu, G. Yang. A study on stress corrosion
dM
33
34
35 cracking and hydrogen embrittlement of Jethete M152 martensitic stainless steel
36
37
38 [J]. Materials Science and Engineering A, 2019, 740-741: 243-251.
39
40
41
42
43
pte

44
45
46
47
48
49
50
51
ce

52
53
54
55
56
Ac

57
58
59
60