j m a t e r r e s t e c h n o l .

2 0 1 9;8(2):2408–2422

Available online at www.sciencedirect.com

www.jmrt.com.br

Review Article

The effects of non-metallic inclusions on

properties relevant to the performance of steel in
structural and mechanical applications

André Luiz Vasconcellos da Costa e Silva

EEIMVR-UFF, Volta Redonda, RJ, Brazil

a r t i c l e i n f o a b s t r a c t

Article history: Non-metallic inclusions (NMIs) occur typically in low or very low volume fractions (from
Received 5 May 2018 10−2 in a high oxygen weld deposit to 10−5 in very clean bearing steels) but play an impor-
Accepted 15 January 2019 tant role in many properties of steel. NMIs play a decisive role in processes involving ductile
Available online 22 February 2019 fracture, fatigue and corrosion, for instance. These are some of the properties more rele-
vant to the performance of steel in structural and mechanical applications. Furthermore,
Keywords: NMIs may influence nucleation during phase transformations of steel. In this work, the rela-
Non-metallic inclusions tion of these properties to NMIs is reviewed, highlighting progress and difficulties in each
Fracture area. Perhaps because of their very low volume fraction, NMIs are sometimes overlooked
Fatigue in the basic physical metallurgy education and their study is left to the realm of those
Ductility interested in steelmaking. In the last decades a dramatic evolution in the understanding
Steel of their relationship to properties, however, has led to significant improvements in many
Steelmaking steel products: the outstanding increase of fatigue life in automotive springs and in bear-
ings is one of many such examples. It is concluded that steel improvement in many cases
requires “inclusion engineering” and this can only be achieved through close collaboration
between physical metallurgy, process metallurgy and steelmaking. Those who realized this
have made significant progress in steel development in recent decades as highlighted in
this short review.
© 2019 The Authors. Published by Elsevier B.V. This is an open access article under the
CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).

E-mail: andre.costaesilva@gmail.com
https://doi.org/10.1016/j.jmrt.2019.01.009
2238-7854/© 2019 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license
(http://creativecommons.org/licenses/by-nc-nd/4.0/).
 j m a t e r r e s t e c h n o l . 2 0 1 9;8(2):2408–2422
Andre Luiz Vasconcellos da Costa e Silva
graduated as Metallurgical Engineer from
the Military Institute of Engineering (IME),
Rio de Janeiro, in 1976. Obtained his
M.A.Sc. from the University of British
Columbia, Vancouver, in 1979 and his
Ph.D. from the University of Florida,
Gainesvile, in 1994. His carreer started as
an engineer in Eletrometal Aços Finos S.A.
(currently Villares Metals S.A.). Has been
the Materials Engineering Manager at CBV
Industria Mecanica S.A. (Currently TechnipFMC plc) and has
taken part on pioneer projects on deepwater oilfield comple-
tion and production. Has been the Technical Director of the
Brazilian Institute for Nuclear Quality, where he has been cer-
tified both as Level III Inspector and Level III Expert in Materials
Engineering. He is Professor at the Metallurgical Engineering
School in Volta Redonda, of Universidade Federal Fluminense.
Andre was the Chairman of the Alloy Phase Diagram Inter-
national Commission (APDIC) from 2008 to 2013. In 2014 he
received from IOM3 , England the Hume Rothery Award, for dis-
tinguished achievements in relation to phase transformations
and in 2017 became the 15th recipient of the triennial ABM,
Brazil “Hubertus Colpaert” Silver Medal for his contribution
in metallography and physical metallurgy. He is a member of
ABM, ASM and TMS.
1. Introduction
While a large emphasis is given to the importance of steel
microstructure in the education of metallurgists and materi-
als scientists, the influence of non-metallic inclusions (NMIs)
on steel properties is comparatively neglected. In industry, on
the other hand, the attention to the importance of NMIs on
steel performance is pervasive and the focus of constant stud-
ies and improvements. This has become especially important
as steel has been challenged by different alternative materi-
als and by more demanding applications. The improvement of
many properties only became possible with the understand-
ing of their relationship to the type, size and distribution of
NMIs present in the matrix. In the last decades, the steel-
making industry developed significant process improvements
that lead to much better control of inclusion volume fraction,
size and composition (e.g. Refs. [1–6]). Tailoring inclusions to
improve properties and performance is an important feature
of steelmaking and the term “inclusion engineering”, coined
in the 1980s [7] is widely used. Inclusion engineering starts
with the definition of the desirable properties the inclusions
should have. Then, through the definition of adequate pro-
cessing conditions, a product is made where these desirable
inclusions are predominantly formed. Controlling inclusion
distribution in the final product, in special those NMIs formed
after the beginning of solidification (secondary inclusions, see
item 2) remains a significant challenge. The tools for model-
ing and understanding this issue have been recently reviewed
[8]. The origin and control of NMIs have been reviewed in a
previous publication [1]. In the present work, the status of
understanding NMIs properties and behavior during process-
ing is first considered. Then, the current understanding of the
relationship between NMIs properties, size, distribution, and
properties that are critical for the performance of steels in
structural and mechanical applications are summarized. As
2409
literature in the field is growing at a fast pace, the reader is
directed, whenever possible, to recent reviews and important
publications on specific topics.
2. Classification of non-metallic inclusions
– a brief overview
Many commercially available elements that are acceptable in
steel composition have a high affinity for oxygen and can
thus be used as deoxidizers, forming non-metallic deoxida-
tion products when added to the liquid steel [9]. Examples are
silicon, manganese and aluminum. Deoxidation products can
become important oxide NMIs. In the case of sulfur, on the
other hand, only elements with low solubility in iron (such
as Ca and Mg) or rare-earths, have sufficiently high affinity
for sulfur to form non-metallic sulfides at the liquid metal
temperatures [9]. Thus, most of the sulfur in steel must be
removed from solution by slag refining and the rest, by precip-
itation reactions occurring mostly during solidification. The
most common sulfide precipitating during solidification is
MnS. Based on these observations two possible classifications
for NMIs emerge: (a) using their chemical composition (oxides,
sulfides, etc.) or (b) considering the moment they form with
respect to the start of the solidification: primary, before solid-
ification starts, and secondary, after solid steel starts forming
in the mold. Furthermore, inclusions originating from the
steelmaking process are classified as “endogenous” and those
appearing from “external” sources (fragments of refractories,
entrapped slag, etc.) are classified as “exogenous”. Rarely how-
ever, will an “exogenous” volume of material survive long
enough in the steel without suffering extensive reaction with
the melt. These reactions produce changes in the NMIs. Thus,
this classification can, sometimes, be confusing [1].
Finally, a common way of classification of NMIs is related
to the inclusion size: inclusions can be classified as macro and
micro inclusions. A sensible cutoff between sizes should fol-
low the proposal by Kiessling [10,11]: an inclusion is a macro
inclusion if it is large enough to cause immediate failure of
the product either during processing or use. All other inclu-
sions are classified as micro inclusions. Thus, while important,
this is a difficult classification to apply. Frequently, sizes are
defined arbitrarily to separate macro and micro inclusions and
the classification becomes difficult to justify.
3. Properties of NMIs relevant to their
effects on steel properties and performance
Some properties of NMIs have a key importance on how they
influence the behavior of steels. These include plasticity or
hardness as a function of temperature, coefficient of ther-
mal expansion (CTE), crystallization behavior (in the case of
glassy inclusions) and to a lesser extent, solubility of metallic
solutes. NMIs have ionic, covalent or mixed bonding charac-
ter. As such, they are in general brittle at room temperature
and have no strong bonding to the metallic matrix. As temper-
ature increases, some inclusions become more plastic. Many
measurements of hardness and plasticity of inclusions have
been performed (e.g. Refs. [12–15]). The observed changes are,
2410 j m a t e r r e s t e c h n o l . 2 0 1 9;8(2):2408–2422

0.6
Before hot After
FeO working hot working
ν 0.4
Void
0.2 Mn in (Fe.Mn)O “hard’’
inclusion

ν 0.1 Ca-aluminates
0

0.2 3 MgO. AI2O3 MgO . AI2O3

Spinels
0.1
ν +MgO.AI2O3
0

1.0 Fe-silicate Complex

Silicates +SiO2 inclusion
+CaFe
ν 0.5 pure SiO2
+Mn Plastic
inclusion
0

1.0 MnS
Fig. 2 – Schematic illustration of the behavior of NMIs after
Me, X = Te X=Se thermomechanical treatment, depending on their relative
(Mn, Me)(S, X)
0.5 plasticity at the working temperatures.
800 1000 1200 1400 1600

Temperature (ºC)

Fig. 1 – Effect of temperature on relative plasticity of NMIs,

depending on their chemical composition. The curves are
of semi-quantitative nature.
Adapted from Ref. [11].

however, too complex to describe in a simple way. A significant

amount of work in simulation of the deformation of imbedded
phases and associated voids is also available (e.g. [14,16–18]).
Nonetheless, one of the most widely used concepts to describe
deformation behavior continues to be inclusion “relative plas-
ticity”. This concept is especially useful when there is a lack
of accurate knowledge of NMIs properties. The concept of rel-
ative plasticity was introduced in the 1960s as the ratio of the
true deformation of inclusions to the true deformation of the
steel [19]. Kiessling [11] summarized the relative plasticity of
inclusions in steel in a classical graph (Fig. 1).
The graph in Fig. 1 gives an overview of the complexity of
the behavior of inclusions when steels are deformed. Depend-
ing on temperature and inclusion composition, their relative
plasticities vary, and these particles may deform, crack or have
mixed behavior, as illustrated in Fig. 2.
In many cases, the combination of low bonding strength to
the matrix and matrix deformation leads to void creation and
separation (or debonding, as commonly used in the composite
literature). Furthermore, hard inclusions will typically break
and redistribute in the steel under these conditions. This has
been discussed by Robinson and Pickering [20] in the case of
alumina, for instance. Understanding the breakage and redis-
tribution of alumina is further complicated by the fact that
alumina inclusions frequently cluster during the processing in
the liquid state, as shown in Fig. 3 [21]. Predicting the behav-
ior of hard inclusions or inclusions that become less plastic
at lower temperatures has been a challenge to modeling the
behavior of steel. Attempts to correlate NMIs behavior with
their elastic moduli or hardness, in drawing tire cord wire at
room temperature, for instance, have not been successful [22].
Fig. 3 – Alumina cluster in steel after deep etching with
nital. Scanning electron microscope (SEM), Secondary
electrons (SE) [21] Courtesy of C. Cicutti, Tenaris-Siderca.
Nonetheless, these are the basic parameters needed to quan-
titatively describe the mechanical behavior of inclusions and
hence their effects on steel properties via modeling (e.g. [23]).
Inclusions that are plastic at the working temperature will
deform when steel is worked. This will result in elongation
of the inclusions along the major working directions. This
introduces, in many cases, shape anisotropy in the inclusions.
This results in anisotropy of the properties influenced by NMIs
[24]. Bernard and co-workers did the most comprehensive
experimental study of oxide plasticity to date [25]. Recently
Zhang and co-workers [22] have confirmed the relationship
between high temperature plasticity and melting point of
oxide inclusions, by correlating the calculated liquidus tem-
perature of the inclusions with their aspect ratio measured
in rolled and drawn tire cord wire. In some modeling work
of NMIs deformation during hot working, the plastic defor-
mation of the inclusions is associated with its viscous flow
 j m a t e r r e s t e c h n o l . 2 0 1 9;8(2):2408–2422
1400
T liquidus
1300
1200
T (ºC)
1100
Party crystalline
1000
Glassy
900
800
1 10 100
Time (minutes)
Fig. 4 – TTT diagram for the composition 41.3% SiO2 , 33.7%
CaO and 24.5% Al2 O3 . The line is qualitative and separates
open point (glassy samples) from filled points (crystals
observed).
Adapted from Ref. [30].
[18] and reasonable prediction of anisotropy is achieved. In
other cases, elastic-plastic behavior is assumed [23], using
data determined some decades ago [25]. On a more quantita-
tive study, Matsuoka [26] and co-workers demonstrated that
the ratio of flow stress of the inclusion to that of the matrix will
define the elongation of the inclusions. Interestingly, the prop-
erties of MnS used in modeling were those determined in the
1960s [27]. They also modeled the behavior of composite inclu-
sions, having alumina surrounded by MnS. The behavior is as
sketched for “complex” inclusions in Fig. 2. The results con-
firmed the experimental observations of [28] who showed that
as sulfur content falls below 60 ppm, the aspect ratio of these
Al2 O3 /MnS inclusions decreases substantially. These results
are of special importance for modern high purity, clean steels.
The anisotropy introduced by NMIs shape change cannot
be eliminated by further heat treatment. The deformation of
NMIs and of segregates is normally responsible for the “fiber”
appearance observed during the macrographic examination
of steels [29]. An eventual crystallization of the NMIs during
steel processing can complicate the prediction of the extent
of their deformation and the variation of plasticity with tem-
perature. Inclusions that are initially “glassy” or amorphous
have been shown to crystallize when subjected to treatments
at temperatures in the range of hot working temperatures of
steels [30,31] (Fig. 4).
The crystallized inclusions have different rheological
behavior (or relative plasticity) when compared to the “glassy”
ones [30]. This can be especially important when NMI plastic-
ity is critical, as in the case of tire cord and spring steel, as will
be discussed later.
The coefficients of thermal expansion (CTE) of NMIs are
different from that of steel. Brooksbank and Andrews [32]
evaluated how these differences in CTE could influence resid-
ual stresses around inclusions, in what is now a classical
work. They indicated inclusions that could be more deleteri-
ous by being surrounded by a tensile stress field associated
with these tessellated stresses. In their view, this could be
especially important under fatigue conditions [32]. These
2411
stresses are also considered relevant to machinability [33]. It
is believed that the formation of stress fields, cavities and
pores in the steel matrix around the inclusions has a favor-
able effect on machinability. Inclusions having higher CTE
than the steel separate from the matrix on cooling from steel
processing temperatures. This may also cause problems in
metallographic sample preparation and difficulties with size
determination. When automatic methods are used, a gray-
level threshold is set to differentiate between oxides and
sulfides [34]. Depending on the selected threshold, the dark
region between matrix and inclusion can contribute differ-
ently to the measured inclusion size.
Significant differences exist in the use of the expression
“inclusions” in the discussion of fracture modeling. Many
authors include carbides, carbonitrides and other second-
phase particles together with NMIs in the definition of
“inclusions” (e.g. [35,36]). When considering the ductile frac-
ture process, it seems important to take in account both type
of particles [37,38]. Special attention must be paid to the dif-
ference of the strength of the matrix–inclusion interface, and
the size and distribution of the different types of particles. The
interface between NMIs and steel has, in general, very low
or no strength [13,14,39,40]. Conversely, the strength of the
interface between carbides and steel, for instance, has been
estimated in the range of 1200–2000 MPa [41]. This difference
has great importance on the effect of NMIs on the steel proper-
ties, particularly fracture. The distinction between NMIs and
second phase particles can become blurred. In some steels,
titanium is used as a nitrogen getter, and TiN inclusions may
be formed in the liquid state [29]. On the other hand, in elec-
tric steels MnS is formed as a fine precipitate to control grain
boundary movement [29]. Gladman [42], using Al2 O3 particles
in steel, demonstrated the controlling role of particle size and
volume fraction in affecting grain boundary movement.
4. Effects of NMIs on some important steel
properties
4.1. Effects on processing – hot and cold working and
forming
It is sometimes convenient to separate the influence of NMIs
during processing from that during application, as the condi-
tions in processing are normally not the ones envisaged for
the steel application. This is true even considering that most
of the problems associated with NMIs discussed in this sec-
tion are related to their relative plasticity and their influence
on steel ductile fracture, the subject of Section 4.2.
Inclusions that occupy a significant portion of the mate-
rial cross section during hot or cold work or that are in regions
where processing deformation is high may cause fracture dur-
ing processing. Fig. 5 illustrates the fracture of a tire cord wire
during cold drawing, caused by the presence of a large inclu-
sion [29]. Controlling the volume fraction, size and distribution
of NMIs is thus important. Furthermore, inclusion engineering
is of paramount importance [43] on the wire making opera-
tions of tire cord and springs. Inclusion engineering is also
important for fatigue properties [44] and will be discussed
later.
2412 j m a t e r r e s t e c h n o l . 2 0 1 9;8(2):2408–2422
48,5
μm
Acc.V Spot magn Det WD 100 μm
15.0 kV 6.0 343x BSE 8.9
Fig. 5 – Fracture caused by a large NMI during cold drawing
of steel wire. The approximate size of the inclusion is
indicated. SEM, Backscattered Electrons (BE) [29].
Fig. 6 – (a) Through thickness rupture of cold rolled steel
foil caused by alumina clusters. (b) Cross section of the end
region of a crack in (a) showing cluster of alumina NMIs.
SEM, SE.
Inclusion distribution in steel products greatly influences
how they affect steel properties. A cluster of relatively large
alumina inclusions has caused the defects shown in Fig. 6 in
a cold rolled foil [45]. Thus, in the case of tin foil, for instance,
it is necessary to perform 100% inspection for perforations
caused by NMIs [45]. Similarly, NMIs that could be acceptable
in a foil for drawing and ironing forming operations (D&I) for
can manufacture may cause failure if they are located in the
highly deformed can flange region [46,47] (Fig. 7). The behav-
ior of flange cracking in D&I steels for cans has been known to
be critical for some time but still requires proper processing
Fig. 7 – Schematic representation of a flange crack in a D&I
can, associated with an NMI.
Adapted from Refs. [46,48].
134 μm
Acc.V Spot magn Det WD Exp 100 μm
20.0 kV 5.6 250x BSE10.0 1
Fig. 8 – Voids formed at the ends of a hard, large,
non-deformable NMI during cold drawing of steel wire (see
Fig. 2). SEM, BE.
control to engineer the NMIs, as requirements are ever more
stringent.
Failure associated with NMIs also happens during closed
die forging. In some cases, 100% immersion ultrasonic inspec-
tion must be performed to eliminate sections of bar stock
where these NMIs concentrate. As severity of forging increases
and steels become stronger and less ductile, the detection lim-
its of ultrasonic examination are approached or surpassed.
Even when they do not cause material failure, NMIs may
introduce voids and stress concentrators, depending on their
relative plasticity, as shown in Fig. 8.
Interesting results have been obtained when modeling the
behavior of single, relatively large, inclusions in a steel matrix
during hot working, using finite element modeling. The work
of Ervasti and Stahlberg [17] is a representative example. How-
ever, as difficulties related to the proper characterization of
 j m a t e r r e s t e c h n o l . 2 0 1 9;8(2):2408–2422
Fig. 9 – Section through growing ductile fracture in a
low-carbon steel showing formation of voids around NMIs
but not around pearlite (originally 240×). This figure was
published in [54]. Copyright Elsevier, 1973.
NMIs properties persist, Ervasti and Stahlberg [17] adopted
an arbitrary relative plasticity (expressed as the ratio of flow
stresses,  inclusion / matrix ) of 3 for hard inclusions and of 1/3
for soft inclusions. In addition, one of the key challenges in
modeling the behavior of inclusions in steel is the multi-
particle aspect and considerations of size and distribution,
which greatly affect the results [36,49].
4.2. Influence on ductile fracture of steel
Ductile fracture processes are important in structural and
mechanical applications of steel. They influence ductility,
formability and toughness (at temperatures where the duc-
tile share of the fracture process is relevant). Ductile fracture
of steels generally occurs by the nucleation, growth and coa-
lescence of microvoids. These voids nucleate by debonding or
cracking of NMIs and second phase particles. Srivastava and
co-workers [50] reviewed key work on this process since its
identification in 1949. The process is in most cases dominated
by the nucleation and growth of microvoids preferentially
around NMIs and their coalescence by necking of the metal
between inclusions. Hence, there is a significant effect on
properties related to ductile fracture. This has been well
demonstrated (see e.g. [36,51–53]). This process occurs, in gen-
eral, on two microscopic scales [37,38]: on a larger one, voids
nucleate on NMIs that crack or debond from the steel matrix.
On a finer one, sub-micron precipitates play a role on void
nucleation. The primary voids grow until a certain strain when
they coalesce. The finer voids soften the ligaments and pro-
mote strain localization and the linking of the primary voids.
Fig. 9 [54] is an especially good illustration where the effect
of the NMIs is evident. A tensile specimen was cut longitudi-
nally in the region of necking, before fracture. The process of
nucleation and coalescence of the voids is clearly seen. This
technique has been widely used for understanding fracture
mechanisms. In some cases, it has also been used to generate
data for validation of modeling of fracture in metals (e.g. Refs.
[36,55]).
The relative effects of NMIs and structural constituents,
such as carbides, on the ductile fracture process are fre-
quently discussed. Apparently, the higher bonding strength
of the carbide–metal interface (Section 3, above) makes their
2413
contribution relevant only at higher deformation, normally
leading to a contribution to the smaller microvoids. If, how-
ever, nitrides or carbides are sufficiently large (see Section
2) they may also contribute to the formation of primary or
large voids (e.g. refs [56,57]). In steels that are “dual-phase”
(DP) or “duplex” the nucleation of cracks at interphase bound-
aries may also play a role in ductile fracture [58]. There is
a general lack on information about steel cleanness in the
studies of fracture of DP steels, so it has not been easy to
define the relative role of NMIs and martensite/ferrite inter-
faces (see for instance, [58–60]). Jamwal and co-workers [61],
however, where able to quantitatively show that tensile elon-
gation in their DP steel was related to the density of voids (or
“pullouts”) nucleated at martensite/ferrite interfaces and had
no relationship to the number of pullouts nucleated at NMIs,
which remained essentially constant when the elongation var-
ied.
The importance of interfacial strength on second-phase
effect on the properties of composites has been identified
and discussed [62]. The work of Ashby and co-workers is
especially illustrative of this [63]. Information about these
concepts has not been, however, widespread in the steel com-
munity. So, even if carbides can also cause voids [40,64] they
crack and debond at much higher strains than NMIs [54].
Thus, NMIs normally control the ductile fracture process of
steels. As a result, reduction of area, elongation and upper
shelf energy as well as the anisotropy in these properties
are very frequently controlled by content, size, shape and
distribution of NMIs as shown in Fig. 10, adapted from [65].
Non-uniform distribution of NMIs, such as the presence of
manganese sulfide clusters, play an important role in increas-
ing their influence on ductile fracture. Tanguy and co-workers
[66] discussed this issue in detail in relation to SA508 Cl.3
forgings used in the nuclear industry. They indicated that the
importance of MnS clusters decrease, as expected, at very
low sulfur content, such as the 40 ppm sulfur steel they stud-
ied.
The understanding of the effect of NMIs on ductile fracture
process and its anisotropy led to important developments:
one of these was inclusion modification treatments or sulfide
modification treatments. These treatments combine low sul-
fur steel and the formation of low plasticity sulfides mostly
via calcium treatment. As a result low anisotropy products
such as plates that fulfill tensile ductility requirements in the
short transverse (“through thickness” or “z”) direction could be
produced [67–69]. The concept of inclusion modification with
calcium further evolved to treatments leading to liquid inclu-
sions at casting temperatures, preventing nozzle clogging in
continuous casting [1,70,71]. For high strength steels for spe-
cial applications, inclusion engineering has evolved to include
the careful modeling of the fracture process [49] and studying
the specific behavior of different kinds of sulfides, for instance:
not only considering inclusion size and spacing [72] but also
how they contribute to fracture [73] and fracture toughness.
Finally, as mentioned in Section 4.1, understanding the influ-
ence of NMIs shape and formability on fracture had a dramatic
effect on the development of inclusion engineering for high
formability steels and tire cord wires [43,44].
Significant advances on the modeling of the fracture pro-
cess have been made, as discussed in [55]. At least three
2414 j m a t e r r e s t e c h n o l . 2 0 1 9;8(2):2408–2422

a 1000

400
900
Upper shelf energy in charpy-V test (J)

200
800

Fatigue limit (MPa)

0.1%C
100 700

60 600
0.2%C
SAE 4063
40
500 SAE 5150
SAE 4052

20 SAE 4140
400
SAE 4340
SAE 2340
10 300
0 0.2 0.4 0.6 0.8 1.0 20 30 40 50 60
Inclusion projected length per unit area (mm/mm2)
Hardness (HRC)
b
Short transverse or normal Fig. 11 – Relationship between hardness (HRC) and fatigue
limit (MPa).
1 Longitudinal
Adapted from Ref. [85].

2 Transverse
d1
Normal
plane
d2
d3
d3
d2
Transverse
plane
Fig. 10 – (a) Effect of inclusion projected length on upper
shelf energy in Charpy tests for steels with 0.1% and 0.2%C.
The symbols indicate individual steels used for the tests.
(b) Orientations and dimensions used to determine
projected lengths.
Adapted from Ref. [65].
important areas present challenges to further advances: (a)
the proper characterization of the properties of NMIs and their
interfaces with steel; (b) an adequate metric for character-
izing size, shape and distribution of the large multi-particle
population and (c) the difficulties associated with compu-
tational methods considering the multi-particle population.
When modeling fracture, detailed elastic-plastic calculations
of void nucleation can be performed if constitutive equations
are defined for the matrix, inclusions and interface (e.g. [74]).
Conversely, very interesting examples of how to deal with (a)
and (b) can be found, for instance, in Refs. [36,50]. A relevant
example of advances in computational methods is presented
in Ref. [23].
d1
4.3. Influence on fatigue
The importance of inclusions on the fatigue of steel has been
long recognized [75–77]. The effect of type, composition, shape
and size of NMIs on fatigue have been extensively discussed
in many publications without a firm conclusion, however. Cor-
relations between total oxygen content and rolling contact
Longitudinal
plane fatigue in rails [78] and with L10 in bearings (the number of
cycles at which 90% of fatigue samples fail) have been demon-
strated [79] but with limited quantitative success. A recent
review dedicated to bearing steels addresses in detail many of
these issues [80]. The work of Murakami and co-workers has
shed new light on understanding the effects of NMIs on fatigue
both before [81,82], and after [83] the publication of his book
[84]. His views can be summarized in the following statements:
The fatigue limit is correlated with the existence of non-
propagating cracks. It is not related to crack initiation. . .“the
fatigue limit is a threshold stress for crack propagation and not
the critical stress for crack initiation”. Fatigue limit correlates
with hardness, up to around 400HV (e.g. Fig. 11, [85]). In this
region, “the fatigue limit is determined by a material property
showing the average resistance to plastic deformation of the
material”.
As one passes the 400 HV threshold, the ideal fatigue limit,
associated with the material properties cannot be reached, in
general, due to the presence of defects (such as inclusions).
“Defects smaller than a critical size are non-damaging (not-
detrimental) to fatigue strength and the critical size is smaller
for materials having a higher static strength, so that a defect
of a given size is more detrimental to high strength steels than
to low strength metals.” Murakami demonstrated with several
experimental examples that since the fatigue limit is a stress
at which crack propagation does not occur, small defects can
and will have cracks starting from them that may or may not
 j m a t e r r e s t e c h n o l . 2 0 1 9;8(2):2408–2422 2415

3 a
(HV + 120) 4

10% error
1 3 CaO-AI2O3
σ

Globular

Harmful index
10% error
TiN
2
σ = 1.43 (HV + 120) /( area ) 1/6 AI2O3
0.3
10 100 1000 (Ca,Mn)S
1
area μm

Fig. 12 – Experimental data of a large number of fatigue 0

5 15 25 35 45
tests replotted by Murakami [84], showing the agreement
Average inclusion diameter, μm
with Eq. (1). For original data see Ref. [84]. b
1.5

lead to fatigue, depending on size and stresses. He argues that,

for this reason, when a crack originates at the inclusion–metal
1.0
interface or through inclusion cracking, the stresses within
the inclusion are relieved and the inclusion domain may be

σ / σw
regarded as mechanically equivalent to a stress-free defect or
pore. Thus, tessellated stresses, for instance [32] could be less Oxides, ≤10 μm
0.5 Oxides, 10 μm < <20 μm
important than previously thought. Using this approach, he Oxides, ≤20 μm
was able to find adequate relationships to predict the fatigue Tin, ≤6 μm
: Average diameter
limit of high strength steels, reconciling the endurance limit
0
relationship with hardness by including a term related to
105 106 107 108
the cross-section area transverse to the loading, occupied by
Cycles to failure, Nf
inclusion. Depending on the loading, position of the inclu-
sions can be important, and this is accounted for. Thus, for
Fig. 13 – (a) Harmfulness index of different inclusions as a
inclusions close to the surface in rotation-bending Murakami
function of composition and size, assessed by Monnot and
and co-workers proposed an empirical relation between the
co-workers [77]. (b) Monnot’s data reanalyzed by Murakami
endurance limit and hardness given by Eq. (1), where the area
[84] shows that inclusion “harmfulness” is very similar for
of the inclusion is considered. Murakami has developed simi-
the various types of inclusions and the fatigue behavior
lar equations for inclusions in other positions.
follows Eq. (1) (ordinate normalized by estimated fatigue
limit).
√ 1
⁄6
w ∼
= 1.43 (Hv + 120) /( areas ) (1)

In this equation, w is thefatigue

 limit (MPa), Hv is the Vick-
ers hardness (HV) and areas ␮m2 is the area occupied by the
inclusion, calculated as described in [84].
Fig. 12 shows the agreement of this relation with experi-
mental results.
The effects of these insights on bearing steels develop-
ment (SAE 52100 or 100Cr6) have been very important. Thus,
for instance, the results of [77], where particular relevance
was ascribed to different inclusion compositions, could be
reappraised. The results indicate much less importance of
inclusion type when analyzed in accordance with Murakami’s
formalism, as shown in Fig. 13.
According to Murakami’s results, the largest inclusion
present in the stressed area will be responsible for fatigue
failure.
With the high cleanness of these steels, the classical meth-
ods of inclusion evaluation and quantification [2,86] became
quite ineffective in predicting fatigue behavior and extreme
value statistics has been presented as a solution [4,87]. In this
context, Murakami developed a method for extreme value
inclusion quantification [88]. Later, an ASTM standard was
developed [89], mostly with the bearing community in view.
With this method, Murakami and co-workers have been able
to predict fatigue properties based on extreme value statis-
tics for inclusions. Furthermore, they showed that, when the
inclusion population and inclusion size become exceedingly
small, as in extra-clean electron-beam (EB) melted steels,
microstructural heterogeneities (bainite areas) are larger than
the inclusions and act as fatigue nuclei [87].
The importance of inclusions in fatigue is still the subject of
frequent discussion, particularly in what is termed very high
cycle fatigue [90–92].
The developments led by Murakami and co-workers on the
understanding of the importance of NMIs in fatigue of high
strength steel also had a profound impact in the inclusion
engineering of spring and valve steels (see Refs. [84,93,94]).
Developments in the field of inclusion engineering in spring
steel in Europe were summarized in Ref. [95] in 2003.
Summarizing, when considering the literature on crack
origination and propagation in fatigue one must consider size
and volume fraction of NMIs. Crack origination may occur “in
the matrix” or related to second-phase particles, in special
NMIs. It seems that for lower strength steels, a critical crack
size larger than the larger NMIs is needed for fatigue to occur.
Thus, NMIs will play a less important role in low strength,
clean steels. Conversely, in high strength steel NMIs may be
2416 j m a t e r r e s t e c h n o l . 2 0 1 9;8(2):2408–2422
sufficiently large and play an important role. This relation-
ship between size and strength is consistent with Murakami’s
EB “clean steel” results reported above. The results of Pessard
and co-workers [96], who studied fatigue anisotropy in three
different steels, are also a good example: they did not observe
any anisotropy in the fatigue behavior of a low strength C35
steel ( U = 580 MPa, AISI 1035, where  U is the tensile strength)
with reasonably normal sulfur content (0.045%). On the other
hand, higher strength steels ( U = 1050–1150 MPa) with high
sulfur content (0.073–0.075%), presented fatigue anisotropy,
associated with NMI-related fatigue fracture.
The anisotropy of fatigue resistance is included in
√
Murakami’s formulation, since the term areas is calculated
using the projected area of the inclusion (or defect) onto a
plane perpendicular to the maximum tensile stress [84]. This
is confirmed by experimental results (e.g. [97,98]).
Some recent results in rolling contact fatigue (RCF) seem to
contradict – at least partially – Murakami’s concepts regarding
the effects of NMIs on fatigue. Both Sada and co-workers [99]
and Hashimoto and co-workers [100] observed that hot iso-
static pressing (HIP) of SAE 52100 bearing steel improved RCF
life. They attributed the effect to the elimination of the cavi-
ties between NMIs and the steel matrix. As the inclusion size is
not altered by HIP, these authors considered that the NMI–steel
interface condition also plays a role in fatigue life. On the other
hand, the NMI–matrix interface strength has been considered
an important factor in the microscopy phenomena involved
in fatigue cracking. Spriestersbach and co-workers [101], for
instance, noted that “classical” NMIs (oxides, complex oxides
and sulfides) debond easily due to the low NMI–steel inter-
facial strength. Furthermore, differences in CTEs [32] may
promote NMIs-matrix separation. Thus, classical NMIs may be
considered to behave as holes, as proposed by Murakami. Con-
versely, TiN, for instance, has a strong bond to the matrix and
the TiN–steel interface shows no separation. When TiN is sub-
jected to high stresses it cracks, and the cracks propagate into
the matrix. Thus, the correlation between TiN size and fatigue
behavior could be different from the one observed for “classi-
cal” NMIs [101]. Evidently, this is an important argument and
further research is needed to clarify these phenomena. This
is a very important issue for RCF. However, the microstruc-
tural changes observed during RCF [102] make investigating
the effects of the individual NMI very difficult, as experimental
results (e.g. [103]) demonstrate.
Furthermore, Murakami’s assumption [81] that the dimen-
sion of NMIs parallel to the stress is not important has been
questioned. Observations in steels with elongated NMIs larger
than those found in very clean steel (such as bearing steels)
suggest that these NMIs may have a larger than expected
influence, due to the long region of the material that will be
subjected to stress concentration [104].
4.4. Influence on the nucleation of ferrite
NMIs may play an important role in phase transformations.
Their influence on nucleation in continuous casting and ingot
casting was discussed in a previous publication [1]. NMIs also
play a critical role in the nucleation of fine acicular ferrite in
weld metal. This microstructure is important to achieve satis-
factory mechanical properties with low carbon compositions
without hot/controlled working. Thus, weld metal composi-
tion is tailored to cause the precipitation of adequate nuclei
for acicular ferrite. Some of the factors considered relevant
for a NMI to act as a nucleus [105,106] for acicular ferrite
are: crystal structure, differences in CTE and depletion in
austenite-stabilizing elements such as Mn around the inclu-
sions [107,108]. The latter is the most favored explanation,
followed by stresses generated by CTE differences between
inclusion and the austenite matrix [106]. This gains special
relevance with the prospect of thin plate casting, where the
extent of hot/controlled working that can be performed to
refine the austenite grain is very limited [109]. The impor-
tance of this phenomenon in welding is well discussed in Refs.
[106,110,111]. The control of NMIs for nucleation in thin foil
continuous casting is discussed by [109,112–114]. One of the
first and clearer in situ observations of the nucleation pro-
cess was done by Sugiyama and Shigesato [115] who discussed
in detail their observations on the importance of manganese
sulfide on ferrite nucleation. Li and co-workers have shown
experimentally and using first principle calculations that Zr
and Ti oxides promote Mn depleted zones in the NMI–matrix
interface, favoring ferrite nucleation [116,117]. Furthermore,
they have shown that MnS can nucleate on Zr oxide. As a
result, they have shown the beneficial effect of Zr-Ti deox-
idation in HSLA steels, promoting finer and more uniform
dispersions of MnS and acicular ferrite microstructures [118].
Grong and co-workers [119] reviewed the possibilities of pro-
ducing “dispersoids”, inclusions with a sufficiently fine size
and compositions to affect nucleation in solidification as well
as ferrite nucleation. In order to achieve this, these particles,
however, would be formed in a more complex way than just
resulting from classical NMI formation reactions, which are
the focus here.
4.5. Influence on machinability
Perhaps one of the properties most traditionally related to
NMIs is machinability [120]. The effects of sulfides [121] are
well known and the design of these inclusions for machin-
ability has been quite successful [122]. A detailed review on
the role of NMIs on machinability, covering many types of
inclusions has been published recently [33]. Computational
thermodynamics has been used to design steels with good
machinability by tailoring sulfides to substitute Pb “metallic
inclusions” [123]. Lead added steel presents important health
hazards during steelmaking and significant environmental
impact. Presently the automotive industry is defining a min-
imum sulfur content for non-resulfurized steels to improve
their machinability. This has posed an interesting challenge
to bar manufacturers who need to adjust their processes to
prevent nozzle clogging by the use of calcium in presence of
sulfur in the range of 0.020% [124]. Stringent process control
is required, in this case.
4.6. Influence on surface finish
Albeit surface finish can be considered a machinability issue,
the case of tool steels for plastic molds, for instance, presents
extraordinary requirements. Studies have shown that both
microstructure and cleanness play an important role. Simple
 j m a t e r r e s t e c h n o l . 2 0 1 9;8(2):2408–2422
forms of cleanness quantification, however, are not able, in
general, to correlate with polishing quality [125]. Inclusion
type play a definite role in the process: in the case of ESR of
P20 steel, for instance, it was demonstrated that the typical
desulfurization of ESR is deleterious for surface finish and
inert atmosphere remelting must be used to prevent desul-
furization and guarantee that sulfides will cover the oxide
inclusions allowing a good surface finish [126]. The compared
results of VAR and the so-called PESR (ESR under inert gas) in
[125] can be explained in the light of this observation.
4.7. Influence on corrosion
Two examples of the influence of NMIs on the corrosion perfor-
mance of steels are the importance of inclusions on hydrogen
related failures such as hydrogen-induced cracking (HIC) and
on the formation of pits, as discussed in the next sections.
4.7.1. NMIs and hydrogen related failures
The importance of inclusions as traps and nuclei for HIC has
been recognized at least since the 1970s [127]. The importance
of the synergistic effect of segregation and NMIs, particularly
manganese sulfide was soon also recognized. Nakai and co-
workers [128] observed that shape control of sulfides had a
great influence on HIC. They showed that higher oxygen steel
with type I sulfides (that do not have high plasticity) had better
resistance to HIC than Al-killed steel with type II sulfides that
elongated during rolling. However, they preferred either Ca or
Ce sulfide modification in order to guarantee good properties.
It was also clear that simply reducing sulfur and controlling
sulfide shape was not sufficient to guarantee excellent HIC
resistance [128] since crack propagation was controlled by seg-
regation. Thus, lower C and lower Mn steels were developed,
as well as accelerated cooling strategies [129] to promote less
segregation, particularly banding, and more uniform hard-
ness in the microstructure. The interaction of segregation
and calcium modification was demonstrated by Matsumiya
and others [130,131] who showed that in large segregates
usual calcium treatment could be ineffective to prevent the
formation of MnS. Thus, very low sulfur and avoidance of
MnS has become the rule to guarantee good HIC resistance
[132,133]. Nonetheless, inclusions continue to play an impor-
tant role on HIC crack nucleation. In very clean steels, it has
been shown that not only MnS promotes hydrogen cracking.
When studying the resistance of API X120 HSLA steel, Huang
and co-workers [134] related steel cleanness to reduced effect
of hydrogen, regardless of the inclusion type. They did not
give information on sulfur content of their steel, however. Jin
and co-workers [133] reported that in a Ca treated API X100
steel with 50 ppm S, oxides were detrimental to hydrogen
resistance. Domizzi and co-workers [135] could not correlate
sulfide length or sulfur content to HIC resistance in steels
with sulfur in the 50–150 ppm range. They propose that sul-
fur content and size influence resistance to hydrogen: “. . . a
small number of very elongated inclusions may reduce the HIC
resistance in the same way as a greater number of shorter par-
ticles.” They also emphasize the relevance of banding, which
in HSLA steels is normally associated with higher Mn con-
tents. The effects of banding were confirmed by [136] in SA
516 steels and by Dunne and co-workers [137,138] in API X70
2417
steels. Banding was also shown to be critical to the hydro-
gen induced failure of AISI 4140 bolts in sub-sea applications
subjected to cathodic protection [139]. Du and co-workers
[140] have shown the beneficial effects of generating a fine
dispersion of oxide and sulfide NMIs through Zr + Ti deox-
idation on HIC resistance, when compared to conventional
Al deoxidation. In a recent review, Ohaeri and co-workers
[141] confirmed that NMIs in general can be deleterious to
resistance to hydrogen degradation but confirmed that elon-
gated inclusions apparently have a more negative effect. The
importance of elongated sulfides on the extent of hydro-
gen blistering has also been demonstrated [135]. Thus, NMIs
shape, amount and type play an important role in hydrogen
cracking. The eventual clustering of NMIs, particularly regions
of microstructural banding, has a synergistic effect in pro-
moting problems associated with hydrogen and should be
carefully avoided.
Additionally, it should be noted that Murakami has demon-
strated that hydrogen trapping at NMIs has a significant effect
on super long life fatigue phenomena [142].
It is evident from the above discussion that inclusion engi-
neering will play an increasingly important role in the design
of hydrogen resistant steels.
4.7.2. Pitting
Wranglén [143] reported the importance of MnS as pitting
initiation sites both in carbon and stainless steels. He pro-
posed that in carbon steels, the attack starts in the matrix
close to the sulfide inclusion, which is nobler than the
matrix while in stainless steels, the attack would start at
the sulfide inclusion proper. With the development of char-
acterization techniques, Ryan and co-workers [144] measured
the presence of a chromium depleted region surrounding sul-
fide inclusions in stainless steels and proposed that these
would be the initiation sites. They did not, however, pro-
pose a mechanism for the formation of these regions nor
describe the thermal history of their sample. Their results
were contested by measurements performed by Meng and
co-workers in various steels (including Ryan’s original sam-
ple) [145] and by [146]. The observations of Ryan resemble
the composition profiles around chromium carbides in sen-
sitized stainless steel. While a clear explanation exists for
the formation of chromium depleted regions in the matrix
around carbides in sensitization that is not the case for
the matrix surrounding sulfide inclusions. More recently,
Williams and co-workers [147] observed a layer of Fe rich
sulfide surrounding the sulfide inclusions in stainless steels.
This layer would preferentially dissolve and start the pit-
ting process. In their conclusions, they suggest that inclusion
engineering could be used to control the composition of
the Mn-Fe-Cr-S sulfides and prevent this from happening.
Park and Kang [148] recently reviewed the issue of NMIs
in stainless steels. They discussed the process of solid-
ification of the sulfides that could lead to the situation
observed by Williams. It seems clear that subtle chemical
composition differences around NMIs can be of paramount
importance for the pitting of stainless steel. Results presently
available indicate that the composition variations caused
during sulfide formation could play a very important role.
Liu and co-workers recently [149] demonstrated the effect
2418 j m a t e r r e s t e c h n o l . 2 0 1 9;8(2):2408–2422

of alumina clusters on pitting of carbon steels. Ma and be the most adequate way to properly characterize NMIs in
co-workers [150] have shown the anisotropic behavior of pit- clean steel.
ting associated to manganese sulfide NMIs. Thus, it is clear
that NMIs, in particular sulfides, have a crucial role in pitting.
Park and Kang [148] remark that the presence of oxide inclu-
6. Conclusions
sions can also play an important role in pitting of stainless
Non-metallic inclusions influence many properties of steels
steels.
relevant to their performance in mechanical and struc-
tural applications. The understanding of how this happens
has evolved in the last decades. Significant progress in the
5. The importance of characterization and quantification of this understanding has been made. Some
quantification of NMIs important limitations persist, however: while the modeling
of ductile fracture has significantly evolved, there has been
little progress in measuring both NMI and NMI-steel interfa-
Based on the above discussion, it is evident that proper
cial properties, a knowledge that is essential to the progress
characterization and quantification of the amount, shape,
of modeling efforts. Furthermore, considering multi-particle
distribution and properties of NMIs is critical for steel charac-
populations and non-uniform distributions in models remains
terization and for the further development of improved steels.
a significant challenge. This is also a challenge for charac-
The importance of the new methods of inclusion characteriza-
terization and quantification of NMIs: while some clustering
tion in the 1990s that made possible proper fatigue predictions
of NMIs will be detected by enhanced non-destructive test-
was also briefly reviewed and discussed in a previous publica-
ing, current methods of quantification focus on the size
tion [1]. This gives an example of how important and dynamic
of particles of the average background NMI population and
the field is. An important challenge in characterization and
do not, in general, measure particle spacing or account for
quantification of NMIs in current clean steel [4] is being able
non-uniform distribution of particles. Combination of charac-
to detect and quantify the “cloud” (or background) of small
terization methods should be used if proper understanding
inclusions and the few large ones. If they occur in large enough
of NMI population is desired. Great progress has been made
sizes, they may be detected by non-destructive testing (NDT),
in the understanding of the importance of NMIs in high
as discussed in item 4.1 above. Characterization techniques
cycle fatigue. This has been accompanied by progress in
that sample large volumes of steel using NDT, such as MIDAS
specific NMI quantification methods and inclusion engineer-
[151,152] are limited by the detection limit of the technique
ing, resulting in significant improvements in materials for
used. Furthermore, in some applications, inclusions consid-
these applications. The use of NMIs for improved machin-
ered “large” are below the limits of reasonable NDT. Scanning
ability has greatly advanced with the understanding of NMI
a sample or the whole volume of steel using NDT becomes
genesis. Inclusion engineering has led to new compositions
useless in this case. Furthermore, clustering and concen-
with improved machinability and surface finish. Significant
tration of secondary NMIs associated with microsegregation
developments also happened in understanding the effect of
[1,29,131] may lead to NMIs behaving, effectively, as if they
NMIs on ferrite nucleation. These developments have reached
were larger [96,98]. Thus, intermediate size clusters or regions
industrial maturity. The understanding of the importance of
where NMIs are concentrated and will influence steel behavior
NMIs in pitting corrosion and hydrogen related fracture has
may not be detected with the current characterization tech-
made enormous progress. However, the interaction of NMIs
niques and may be too small for the current non-destructive
and segregation, inherent to casting processes, still poses a
examination technique. When these clusters are associated
significant challenge to the complete understanding of NMI
with production events (mainly in the casting process (e.g.
influence on the complex hydrogen related fracture phe-
[153]), stringent process control may limit their occurrence
nomenon.
or identify when they occur. However, some route to charac-
Progress in the understanding and control of the effects of
terize the non-uniform distribution of secondary inclusions,
NMIs on the properties of steels in mechanical and structural
caused by microsegregation, may require additional develop-
applications require the close collaboration between physical
ment efforts. With the presently available techniques, more
metallurgy, process metallurgy and steelmaking. Those who
efforts could be placed on evaluating neighbor spacing dis-
realized this have made dramatic progress in steel develop-
tribution by different techniques [4,154,155] which can rely
ment in recent decades and will probably continue to do so.
on usual quantitative metallography tools and are frequently
applied to composites. Present efforts at development new
techniques of three dimensional (3D) characterization such as
Conflicts of interest
X-ray Micro-CT [156] or other 3D reconstruction techniques
[157,158] may also become viable solutions in the future.
The author declares no conflicts of interest.
Presently, techniques such as automated SEM-EDS analysis
(frequently mentioned using the commercial name ASPEX® )
offer the possibility of sampling large areas. Combined with Acknowledgements
3D reconstruction (e.g. [153]) this may also give valuable infor-
mation on particle distribution and clustering. While different The author thanks all his former and present students for
methods will give different information about NMI population inspiration and important contributions to the work on non-
[3], most probably a combination of methods will continue to metallic inclusions in steels. The collaboration of colleagues
 j m a t e r r e s t e c h n o l . 2 0 1 9;8(2):2408–2422
in many steelmakers and research organizations is gratefully
acknowledged.
The support of CNPq, CAPES and FAPERJ during different
stages of the author’s work is also acknowledged.
references
[1] Costa e Silva A. Non-metallic inclusions in steels – origin
and control (review). J Mater Res Technol 2018;7(3):283–99.
[2] Zhang L, Thomas BG. State of the art in evaluation and
control of steel cleanliness. ISIJ Int 2003;43(3):271–91.
[3] Kaushik P, Lehmann J, Nadif M. State of the art in control of
inclusions their characterization, and future requirements.
Metall Mater Trans B 2012;43(4):710–25.
[4] Atkinson HV, Shi G. Characterization of inclusions in clean
steels: a review including the statistics of extremes
methods. Prog Mater Sci 2003;48(5):457–520.
[5] Alexis J, Andersson M, Björkvall J, Sichen D, Sandberg A.
Optimization of secondary metallurgy with respect to
non-metallic inclusions. Report No.: Jernkontorets
Forskning D836 JK 23045. Stockholm, Sweden: Jernkontoret;
2011.
[6] IISI. IISI study on clean steel. Brussels, Belgium:
International Iron and Steel Institute; 2004.
[7] Kay DAR. Inclusions engineering. CIM Bull 1982;75(842):130.
[8] You D, Michelic SK, Presoly P, Liu J, Bernhard C. Modeling
inclusion formation during solidification of steel: a review.
Metals 2017;7(11):460.
[9] Gaye H. Inclusion formation in steels. In: Cramb A, editor.
The making shaping and treating of steel, casting volume.
11th ed. Pittsburgh, PA: AISE; 2003 [chapter 3].
[10] Kiessling R. Non-metallic inclusions in steels. London: The
Iron and Steel Institute; 1968.
[11] Kiessling R. The influence of non-metallic inclusions on the
properties of steel. J Met 1969;21:47–54.
[12] Gove KB, Charles JA. The high-temperature hardness of
various phases in steel. Met Technol 1974;1(1):279–83.
[13] Baker TJ, Gave KB, Charles JA. Inclusion deformation and
toughness anisotropy in hot-rolled steels. Met Technol
1976;3(1):183–93.
[14] Luo C, Doctoral thesis Modeling the behavior of inclusions
in plastic deformation of steels. Stockholm, Sweden: KTH
Royal Institute of Technology; 2001.
[15] Vodopivec F, Gabrovšek M. Relative plasticity of manganese
sulphide inclusions during rolling of some industrial steels.
Met Technol 1980;7(1):186–91.
[16] Yu H, Bi H, Liu X, Chen L, Dong N. Behavior of inclusions
with weak adhesion to strip matrix during rolling using
FEM. J Mater Process Technol 2009;209(9):4274–80.
[17] Ervasti E, Ståhlberg U. Void initiation close to a
macro-inclusion during single pass reductions in the hot
rolling of steel slabs: a numerical study. J Mater Process
Technol 2005;170(1):142–50.
[18] Luo CH, Stahlberg U. Deformation of inclusions during hot
rolling of steels. J Mater Process Technol 2001;114(1):
87–97.
[19] Malkiewicz T, Rudnik S. Deformation of non-metallic
inclusions during rolling of steel. J Iron Steel Inst
1963;201:33–8.
[20] Robinson SW, Martin IW, Pickering FB. Formation of
alumina in steel and its dissemination during mechanical
working. Met Technol 1979;6(1):157–69.
[21] Cicutti C, Valdez M, Pérez T, Ares R, Panelli R, Petroni J.
Optimisation of calcium treatment to improve castability.
In: 18th Process Technology Conference ISS. 2001. p.
871–82.
2419
[22] Zhang L, Guo C, Yang W, Ren Y, Ling H. Deformability of
oxide inclusions in tire cord steels. Metall Mater Trans B
2018;49(2):803–11.
[23] Gupta A, Cecen A, Goyal S, Singh AK, Kalidindi SR.
Structure–property linkages using a data science approach:
application to a non-metallic inclusion/steel composite
system. Acta Mater 2015;91:239–54.
[24] Sauveur A. Metallography and heat treatment of iron and
steel. 4th ed. New York, NY: McGraw-Hill; 1935.
[25] Bernard G, Riboud PV, Urbain G. Etude de la plasticite
d’inclusions d’oxides. Rev Metall – CIT 1981:421–34.
[26] Matsuoka N, Terano M, Ishiguro T, Abe E, Yukawa N,
Ishikawa T, et al. Computer simulation of deformation
behavior of non-metallic inclusion in hot-rolling. Procedia
Eng 2014;81:120–5.
[27] Misaka Y, Yoshimoto T. Formulation of mean resistance of
deformation of plain carbon steel at elevated temperature. J
Jpn Soc Technol Plast 1967;8(79):414–22.
[28] Yamamoto K, Yamamura H, Suwa Y. Behavior of
non-metallic inclusions in steel during hot deformation
and the effects of deformed inclusions on local ductility.
ISIJ Int 2011;51(12):1987–94.
[29] Colpaert H. Metallography of steels: interpretation of
structure and the effects of processing, updated and
translated by Costa e Silva A. Materials Park, OH: ASM
International; 2018.
[30] Rocabois P, Pontoire JN, Lehmann J, Gaye H. Crystallization
kinetics of Al2 O3 -CaO-SiO2 based oxide inclusions. J
Non-Cryst Solids 2001;282:98–109.
[31] Faraji M, Wilcox DP, Thackray RP, Todd I, Howe AA,
Tsakiropoulos P. Effect of early stages of thermomechanical
processing on inclusions in high carbon steel. Mater Sci
Technol 2014;30(10):1177–88.
[32] Brooksbank D, Andrews KW. Stress fields around inclusions
and their relation to mechanical properties. J Iron Steel Inst
1972:246–55.
[33] Ånmark N, Karasev A, Jönsson P. The effect of different
non-metallic inclusions on the machinability of steels.
Materials 2015;8(2):751–83.
[34] ASTM. ASTM E1245-03 standard practice for determining
the inclusion or second-phase constituent content of
metals by automatic image analysis. West Conshohocken,
PA: ASTM – American Society for Testing and Materials;
2003.
[35] Pardoen T, Scheyvaerts F, Simar A, Tekoğlu C, Onck PR.
Multiscale modeling of ductile failure in metallic alloys. C R
Phys 2010;11(3):326–45.
[36] Pineau A, Benzerga AA, Pardoen T. Failure of metals I: brittle
and ductile fracture. Acta Mater 2016;107:424–83.
[37] Thomason PF. Ductile fracture of metals. 1st ed. New York:
Pergamon Press; 1990. p. 219.
[38] O’Keeffe SC, Tang S, Kopacz AM, Smith J, Rowenhorst DJ,
Spanos G, et al. Multiscale ductile fracture integrating
tomographic characterization and 3-D simulation. Acta
Mater 2015;82:503–10.
[39] Pickering FB. Toward improved toughness und ductility. The
Climax Molybdenum Ltd; 1971.
[40] Le Roy G, Embury JD, Edwards G, Ashby MF. A model of
ductile fracture based on the nucleation and growth of
voids. Acta Metall 1981;29(8):1509–22.
[41] Poruks P, Yakubtsov I, Boyd JD. Martensite–ferrite interface
strength in a low-carbon bainitic steel. Scr Mater
2006;54(1):41–5.
[42] Gladman T, Fourlaris G, Talafi-Noghani M. Grain refinement
of steel by oxidic second phase particles. Mater Sci Tecnol
1999;15:1414–24.
[43] Suito H, Inoue R. Thermodynamics on control of inclusions
composition in ultra-clean steels. ISIJ Int 1996;36(5):528–36.
2420 j m a t e r r e s t e c h n o l . 2 0 1 9;8(2):2408–2422
[44] Onoe T, Ito S, Ogawa K, Mimura T, Matsumoto H, Maeda S.
Shape control of inclusions for steel tire cord (development
in ladle arc refining). Trans ISIJ 1987;27:B249.
[45] Batista RP, Martins AA, Costa e Silva A. The effects of
deoxidation practice on the quality of thin foil low-carbon
steel. J Min Metall Sect B Metall 2017;53(3B):357–63.
[46] Kobayashi H, Kurokawa T, Shimomura T, Matsudo K,
Miyahara S. Effect of non-metallic inclusion of flange
cracking of drawn and ironed can from tinplate. Trans Iron
Steel Inst Jpn 1983;23(5):410–6.
[47] Kaushik P, Lowry M, Yin H, Pielet H. Inclusion
characterisation for clean steelmaking and quality control.
Ironmak Steelmak 2012;39(4):284–300.
[48] Jacobi H. Metallurgy and materials technology of
high-purity steels. Stahl Eisen 1994;114(11):45–56.
[49] Bandstra JP, Koss DA, Geltmacher A, Matic P, Everett RK.
Modeling void coalescence during ductile fracture of a steel.
Mater Sci Eng – Struct Mater Prop Microstruct Process
2004;366(2):269–81.
[50] Srivastava A, Ponson L, Osovski S, Bouchaud E, Tvergaard V,
Needleman A. Effect of inclusion density on ductile fracture
toughness and roughness. J Mech Phys Solids 2014;63:
62–79.
[51] Edelson BI, Baldwin WM Jr. The effect of second phases on
the mechanical properties of alloys. Trans ASM
1962;55:230–50.
[52] Speich GR, Spitzig WA. Effect of volume fraction and shape
of sulfide inclusions on through-thickness ductility and
impact energy of high-strength 4340 plate steels. Metall
Trans A 1982;13(12):2239–58.
[53] Abyazi A, Ebrahimi AR. Mechanical anisotropy in Ca-treated
and ultra-low sulphur HSLA-100 steel. Mater Sci Technol
2016;32(10):976–84.
[54] Knott J. Fundamentals of fracture mechanics. London:
Butherworths; 1973.
[55] Benzerga AA, Leblond J-B. Ductile fracture by void growth to
coalescence. Adv Appl Mech 2010;44:169–305.
[56] Savic V, Hector LG. Tensile deformation and fracture of
press hardened boron steel using digital image correlation.
SAE paper 2007-01-0790. In: Steel Innovations, Fatigue
Research, Sheet/Hydro/Gas Forming Technology &
Advanced High Strength Steel Development (SP-2103) SAE
World Congress & Exhibition, Detroit, Michigan. 2007.
[57] Bardelcik A, Vowles CJ, Worswick MJ. A mechanical,
microstructural, and damage study of various tailor hot
stamped material conditions consisting of martensite,
bainite, ferrite, and pearlite. Metall Mater Trans A
2018;49(4):1102–20.
[58] Kadkhodapour J, Butz A, Ziaei Rad S. Mechanisms of void
formation during tensile testing in a commercial,
dual-phase steel. Acta Mater 2011;59(7):2575–88.
[59] Avramovic-Cingara G, Saleh CAR, Jain MK, Wilkinson DS.
Void nucleation and growth in dual-phase steel 600 during
uniaxial tensile testing. Metall Mater Trans A
2009;40(13):3117–27.
[60] Samei J, Green DE, Cheng J, de Carvalho Lima MS. Influence
of strain path on nucleation and growth of voids in dual
phase steel sheets. Mater Des 2016;92:1028–37.
[61] Jamwal RS, Gokhale AM, Bhat SP. Quantitative fractographic
analysis of variability in the tensile ductility of a high
strength dual-phase steel. Metallogr Microstruct Anal
2013;2(1):30–4.
[62] Chawla KK. Composite materials science and engineering.
New York: Springer-Verlag; 1987.
[63] Ashby MF, Blunt FJ, Bannister M. Flow characteristics of
highly constrained metal wires. Acta Metall Mater
1989;37(7):1847.
[64] Turkalo AM, Low JR Jr. The effect of carbide dispersion on
the strength of tempered martensite. Trans AIME
1958;212(December):750–8.
[65] Spitzig WA. Effect of sulfides and sulfide morphology on
anisotropy of tensile ductility and toughness of hot-rolled
C-Mn steels. Metall Trans A 1983;14A:471–84.
[66] Tanguy B, Besson J, Piques R, Pineau A. Ductile to brittle
transition of an A508 steel characterized by Charpy impact
test. Part I. Experimental results. Eng Fracture Mech
2005;72(1):49–72.
[67] Luyckx L, Bell JR, McLean A, Korchynsky M. Sulfide shape
control in high strength low alloy steels. Metall Trans
1970;1(12):3341–50.
[68] Tivelius B, Sohlgren T, Wretlind C. The processing of z-steel
in an ASEA-SKF ladle furnace. In: Secondary steelmaking :
proceedings of a conference/organized by Activity Group
Committee V (Ironmaking and Steelmaking) of The Metals
Society. London, England: The Metals Society; 1977.
[69] Banks TM, Gladman T. Sulphide shape control. Met Technol
1979;6(1):81–94.
[70] Bannenberg N. Inclusion modification to prevent nozzle
clogging. In: 78th Steelmaking Conference Proceedings.
Nashville: ISS-AIME, Warrendale PA. 1995. p. 457–63.
[71] Cramb AW, Rastogi R, Maddalena RL. Nozzle clogging. In:
Cramb AW, editor. The making, shaping and treating of
steel, casting volume. 11th ed. Pittsburgh, PA: AISE; 2003.
[72] Maloney JL, Garrison WM. The effect of sulfide type on the
fracture behavior of HY180 steel. Acta Mater
2005;53(2):533–51.
[73] Iorio LE, Garrison WM. Effects of gettering sulfur as CrS or
MnS on void generation behavior in ultra-high strength
steel. Scr Mater 2002;46(12):863–8.
[74] Xu X-P, Needleman A. Void nucleation by inclusion
debonding in a crystal matrix. Model Simul Mater Sci Eng
1993;1(2):111.
[75] Thornton PA. The influence of nonmetallic inclusions on
the mechanical properties of steel: a review. J Mater Sci
1971;6(4):347–56.
[76] Cogne JY, Heritier B, Monnot J. Cleanness and fatigue life of
bearing steels. In: Clean Steel 3; Balatonfured; Hungary; 2–4
June 1987. 1986. p. 26–31.
[77] Monnot J, Heritier B, Cogne JY. Relationship of melting
practice, inclusion type, and size with fatigue resistance of
bearing steels. In: Effect of Steel Manufacturing Processes
on the Quality of Bearing Steels (ASTM STP 987); Phoenix,
Arizona; USA; 4–6 Nov 1986. 1988. p. 149.
[78] Shur EA, Bychkova NY, Trushevsky SM. Physical metallurgy
aspects of rolling contact fatigue of rail steels. Wear
2005;258(7–8):1165–71.
[79] Unigame Y, Hiraoka K, Takasu I, Kato Y. Evaluation
procedures nonmetallic inclusions in steel for highly
reliable bearings. J ASTM Int 2006;3(5):1–8.
[80] Bhadeshia HKDH. Steels for bearings. Prog Mater Sci
2012;57(2):268–435.
[81] Murakami Y, Kodama S, Konuma S. Quantitative-evaluation
of effects of non-metallic inclusions on fatigue-strength of
high-strength steels. 1. Basic fatigue mechanism and
evaluation of correlation between the fatigue
fracture-stress and the size and location of non-metallic
inclusions. Int J Fatigue 1989;11(5):291–8.
[82] Murakami Y, Endo M. Effects of defects, inclusions and
inhomogeneities on fatigue strength. Int J Fatigue
1994;16(3):163–82.
[83] Murakami Y, Matsunaga H, Abyazi A, Fukushima Y. Defect
size dependence on threshold stress intensity for
high-strength steel with internal hydrogen. Fatigue Fract
Eng Mater Struct 2013;36(9):836–50.
 j m a t e r r e s t e c h n o l . 2 0 1 9;8(2):2408–2422
[84] Murakami Y. Metal fatigue: effects of small defects and
nonmetallic inclusions. Kidlington, England: Elsevier
Science; 2002.
[85] Garwood MF, Gensamer M, Zurburg HH, Burwell JT, Erickson
MA, La Que FL. Interpretation of tests and correlation with
service. Materials Park, OH: ASM International; 1951.
[86] Bartosiaki BG, Pereira JAM, Bielefeldt WV, Vilela ACF.
Assessment of inclusion analysis via manual and
automated SEM and total oxygen content of steel. J Mater
Res Technol 2015;4(3):235–40.
[87] Murakami Y, Toriyama T, Tsubota K, Furumura K. What
happens to the fatigue limit of bearing steel without
nonmetallic inclusions? Fatigue strength of electron beam
remelted super clean bearing steel. In: Bearing steels: into
the 21st century (STP 1327). ASTM International; 1998. p.
87–105.
[88] Murakami Y, Toriyama T, Coudert EM. Instructions for a
new method of inclusion rating and correlations with the
fatigue limit. J Test Eval 1994;22(4):318–26.
[89] ASTM. ASTM E2283 “Standard practice for extreme value
analysis of nonmetallic inclusions in steel and other
microstructural features”. W Conshohocken, PA: Am Soc
Test Mater; 2004.
[90] Sakai T, Harada H, Oguma N. Crack initiation mechanism of
bearing steel in very high cycle fatigue. In: Proceeding
ECF-16. 2006.
[91] Sakai T. Review and prospects for current studies on very
high cycle fatigue of metallic materials for machine
structural use. J Solid Mech Mater Eng 2009;3(3):425–39.
[92] Li SX. Effects of inclusions on very high cycle fatigue
properties of high strength steels. Int Mater Rev
2012;57(2):92–114.
[93] Shiwaku K, Yamada Y, Koarai J, Kawaguchi Y. Improvement
of fatigue life of valve spring wire by reducing non-metallic
inclusions, SAE Technical Paper 850364. SAE Trans; 1985. p.
3.32–43.
[94] Oshiro T, Ikeda T, Matsuyama H, Okushima S, Oki Y, Ibaraki
N. Verbesserung der Dauerhaltbarkeit von Ventilfederdraht.
Stahl Eisen 1989;10(21):1011–5.
[95] Bertrand C, Molinero J, Landa S, Elvira R, Wild M, Barthold G,
et al. Metallurgy of plastic inclusions to improve fatigue life
of engineering steels. Ironmak Steelmak 2003;30(2):165–9.
[96] Pessard E, Morel F, Bellett D, Morel A. A new approach to
model the fatigue anisotropy due to non-metallic
inclusions in forged steels. Int J Fatigue 2012;41:168–78.
[97] Roiko A, Hänninen H, Vuorikari H. Anisotropic distribution
of non-metallic inclusions in a forged steel roll and its
influence on fatigue limit. Int J Fatigue 2012;41:158–67.
[98] Pessard E, Morel F, Verdu C, Flacelière L, Baudry G.
Microstructural heterogeneities and fatigue anisotropy of
forged steels. Mater Sci Eng A 2011;529:289–99.
[99] Sada T, Nonaka Y, Mikami T, Kizawa K. Improvement of
rolling bearing life by bonding of non-metallic inclusions
with matrix (part 1). J Jpn Soc Tribol 2016;61(4):264–74.
[100] Hashimoto K, Fujimatsu T, Tsunekage N, Hiraoka K, Kida K,
Santos EC. Effect of inclusion/matrix interface cavities on
internal-fracture-type rolling contact fatigue life. Mater Des
2011;32(10):4980–5.
[101] Spriestersbach D, Grad P, Kerscher E. Influence of different
non-metallic inclusion types on the crack initiation in
high-strength steels in the VHCF regime. Int J Fatigue
2014;64:114–20.
[102] Sadeghi F, Jalalahmadi B, Slack TS, Raje N, Arakere NK. A
review of rolling contact fatigue. J Tribol 2009;131(4):041403.
[103] Fu H, Rydel JJ, Gola AM, Yu F, Geng K, Lau C, et al. The
relationship between 100Cr6 steelmaking, inclusion
microstructure and rolling contact fatigue performance. Int
2421
J Fatigue 2018, http://dx.doi.org/10.1016/j.ijfatigue.2018.11.
011.
[104] Vaara J, Mäntylä A, Frondelius T. Brief review on high-cycle
fatigue with focus on non-metallic inclusions and forming.
Raken Mek 2017;50(3):146.
[105] Sarma DS, Karasev AV, Jönsson PG. On the role of
non-metallic inclusions in the nucleation of acicular ferrite
in steels. ISIJ Int 2009;49(7):1063–74.
[106] Babu SS. The mechanism of acicular ferrite in weld
deposits. Curr Opin Solid State Mater Sci 2004;8(3–4):267–78.
[107] Byun JS, Shim JH, Cho YW, Lee DN. Non-metallic inclusion
and intragranular nucleation of ferrite in Ti-killed C-Mn
steel. Acta Mater 2003;51(6):1593–606.
[108] Gregg JM, Bhadeshia H. Titanium-rich mineral phases and
the nucleation of bainite. Metall Mater Trans - Phys Metall
Mater Sci 1994;25(8):1603–11.
[109] Mizoguchi S, Ph.D. thesis A study on segregation and oxide
inclusions for the control of steel properties. University of
Tokyo; 1996.
[110] Koseki T, Thewlis G. Overview Inclusion assisted
microstructure control in C–Mn and low alloy steel welds.
Mater Sci Technol 2005;21(8):867–79.
[111] Babu SS, David SA. Inclusion formation and microstructure
evolution in low alloy steel welds. ISIJ Int
2002;42(12):1344–53.
[112] Killmore CR, Creely H, Phillips A, Kaul H, Campbell P,
Schueren M, et al. Development of ultra-thin cast strip
products by the Castrip® process. In: Mater Forum; 2008. p.
13–28.
[113] Liu Z, Kobayashi Y, Yin F, Kuwabara M, Nagai K. Nucleation
of acicular ferrite on sulfide inclusion during rapid
solidification of low carbon steel. ISIJ Int 2007;47(12):
1781–8.
[114] Loder D, Michelic SK, Bernhard C. Acicular ferrite formation
and its influencing factors – a review. J Mater Sci Res
2016;6(1):24.
[115] Sugiyama M, Shigesato G. Development of in-situ
microstructure observation technique in steels. Nippon
Steel Tech Rep 2005;91:13–7.
[116] Li Y, Wan XL, Cheng L, Wu KM. First-principles calculation
of the interaction of Mn with ZrO2 and its effect on the
formation of ferrite in high-strength low-alloy steels. Scr
Mater 2014;75:78–81.
[117] Li Y, Wan XL, Cheng L, Wu KM. Effect of oxides on
nucleation of ferrite: first principle modelling and
experimental approach. Mater Sci Technol 2016;32(1):88–93.
[118] Li Y, Wan XL, Lu WY, Shirzadi AA, Isayev O, Hress O, et al.
Effect of Zr-Ti combined deoxidation on the microstructure
and mechanical properties of high-strength low-alloy
steels. Mater Sci Eng A 2016;659:179–87.
[119] Grong Ø, Kolbeinsen L, van Der Eijk C, Tranell G.
Microstructure control of steels through dispersoid
metallurgy using novel grain refining alloys. ISIJ Int
2006;46(6):824–31.
[120] Merchant ME, Zlatin N. Basic reasons for good machinability
of free machining steels. Trans Am Soc Met 1949;41:647–77.
[121] Ramalingam S, Watson JD. Inclusion chemistry control for
machinability enhancement in steels. Mater Sci Eng
1980;43(2):101–8.
[122] Oikawa K, Ishida K, Nishizawa T. Effect of titanium addition
on the formation and distribution of MnS inclusions in
steel during solidification. ISIJ Int 1997;37(4):332–8.
[123] Oikawa K, Mitsui H, Ebata T, Takigushi T, Shimitzu T,
Ishikawa K, et al. A new Pb-free machinable ferritic
stainless steel. ISIJ Int 2002;42(7):806–7.
[124] Mercier M, M.Sc. thesis Adição Dinamica de CaSi. Belo
Horizonte, Brazil: UFMG; 2014.
2422 j m a t e r r e s t e c h n o l . 2 0 1 9;8(2):2408–2422
[125] Klocke F, Dambon O, Behrens B. Analysis of defect
mechanisms in polishing of tool steels. Prod Eng
2011;5(5):475–83.
[126] Mesquita RA, Schneider R. Tool steel quality and surface
finishing of plastic molds. Exacta 2010;8(3):307–18.
[127] Miyoshi E, Tanaka T, Terasaki F, Ikeda A. Hydrogen-induced
cracking of steels under wet hydrogen sulfide environment.
J Eng Ind 1976;98(4):1221–30.
[128] Nakai Y, Emi T, Kurahashi H, Haida O. Development of steel
resistant to hydrogen induced cracking in wet hydrogen
sulfide environment. Kawasaki Steel Tech Rep
1980;(1):47–59.
[129] Tamehiro OH, Takeda T, Matsuda S, Yamamoto K, Okumura
N. Effect of accelerated cooling after controlled rolling on
the hydrogen induced cracking resistance of line pipe steel.
Trans Iron Steel Inst Jpn 1985;25(9):982–8.
[130] Matsumyia T, Kajioka H, Mizoguchi S, Ueshima Y, Esaka H.
Mathematical analysis of segregations in continuously-cast
slabs. Trans Iron Steel Inst Jpn 1984;24(11):873–82.
[131] Matsumiya T. Mathematical analyses of segregations and
chemical compositional changes of nonmetallic inclusions
during solidification of steels. Mater Trans JIM
1992;33(9):783–94.
[132] Kim WK, Koh SU, Yang BY, Kim KY. Effect of environmental
and metallurgical factors on hydrogen induced cracking of
HSLA steels. Corros Sci 2008;50(12):3336–42.
[133] Jin TY, Liu ZY, Cheng YF. Effect of non-metallic inclusions
on hydrogen-induced cracking of API5L X100 steel. Int J
Hydrog Energy 2010;35(15):8014–21.
[134] Huang F, Liu J, Deng ZJ, Cheng JH, Lu ZH, Li XG. Effect of
microstructure and inclusions on hydrogen induced
cracking susceptibility and hydrogen trapping efficiency of
X120 pipeline steel. Mater Sci Eng A 2010;527(26):6997–7001.
[135] Domizzi G, Anteri G, Ovejero-Garcia J. Influence of sulphur
content and inclusion distribution on the hydrogen
induced blister cracking in pressure vessel and pipeline
steels. Corros Sci 2001;43:325–39.
[136] Kulesza BS, Shick JE, Fletcher FB, McDowell EE. Hydrogen
induced CRACKING of ASTM A516 pressure vessel plate
steels and the apparent effect of microscopic banding. In:
Corrosion 2016 paper 7498. Houston, TX: NACE; 2016.
[137] Dunne DP, Hejazi D, Saleh AA, Haq AJ, Calka A, Pereloma EV.
Investigation of the effect of electrolytic hydrogen charging
of X70 steel: I. The effect of microstructure on
hydrogen-induced cold cracking and blistering. Int J Hydrog
Energy 2016;41(28):12411–23.
[138] Saleh AA, Hejazi D, Gazder AA, Dunne DP, Pereloma EV.
Investigation of the effect of electrolytic hydrogen charging
of X70 steel: II. Microstructural and crystallographic
analyses of the formation of hydrogen induced cracks and
blisters. Int J Hydrog Energy 2016;41(28):12424–35.
[139] Paes MTP, Piana LA, Marinho RR, Costa e Silva A. Subsea
hydrogen induced stress cracking of low alloy steel
fasteners (IBP1860 16). In: Anals of Rio Oil & Gas Expo and
Conference 2016. Rio de Janeiro, Brasil: IBP; 2016.
[140] Du Y, Wu KM, Cheng L, Li Y, Isayev O, Hress O. Effect of Zr–Ti
deoxidisation on the hydrogen-induced cracking of X65
pipeline steels. Mater Sci Technol 2016;32(7):728–35.
[141] Ohaeri E, Eduok U, Szpunar J. Hydrogen related degradation
in pipeline steel: a review. Int J Hydrog Energy
2018;43(31):14584–617.
[142] Murakami Y, Nomoto T, Ueda T. On the mechanism of
fatigue failure in the superlong life regime (N > 107 cycles).
Part 1: influence of hydrogen trapped by inclusions. Fatigue
Fract Eng Mater Struct 2000;23(11):893–902.
[143] Wranglen G. Pitting and sulphide inclusions in steel. Corros
Sci 1974;14(5):331–49.
[144] Ryan MP, Williams DE, Chater RJ, Hutton BM, McPhail DS.
Why stainless steel corrodes. Nature 2002;415(6873):
770–4.
[145] Meng Q, Frankel GS, Colijn HO, Goss SH. Stainless-steel
corrosion and MnS inclusions: metallurgy (communication
arising). Nature 2003;424(6947):389–90.
[146] Schmuki P, Hildebrand H, Friedrich A, Virtanen S. The
composition of the boundary region of MnS inclusions in
stainless steel and its relevance in triggering pitting
corrosion. Corros Sci 2005;47(5):1239–50.
[147] Williams DE, Kilburn MR, Cliff J, Waterhouse GIN.
Composition changes around sulphide inclusions in
stainless steels, and implications for the initiation of
pitting corrosion. Corros Sci 2010;52(11):3702–16.
[148] Park JH, Kang Y. inclusions in stainless steels – a review.
Steel Res Int 2017;88(12):1700–2130.
[149] Liu Q, Yang S, Zhao M, Zhu L, Li J. Pitting corrosion of steel
induced by Al2 O3 inclusions. Metals 2017;7(9):347.
[150] Ma J, Zhang B, Wang J, Wang G, Han E-H, Ke W. Anisotropic
3D growth of corrosion pits initiated at MnS inclusions for
A537 steel during corrosion fatigue. Corros Sci
2010;52(9):2867–77.
[151] Jacobi H, Wünnenberg K. Methods for determination of
oxide cleanness in steel. In: IISI study on clean steel.
Brussels, Belgium: International Iron and Steel Institute;
2004.
[152] Jacobi H, Wünnenberg K. Improving oxide cleanness on
basis of MIDAS method. Ironmak Steelmak
2003;30(2):130–41.
[153] Deng X, Ji C, Cui Y, Tian Z, Yin X, Shao X, et al. Formation
and evolution of macro inclusions in IF steels during
continuous casting. Ironmak Steelmak 2017;44(10):
739–49.
[154] Wang D, Shanthraj P, Springer H, Raabe D. Particle-induced
damage in Fe–TiB2 high stiffness metal matrix composite
steels. Mater Des 2018;160:557–71.
[155] Shehata MT, Boyd JD. Measurement of spatial distribution
of inclusions. In: R Rungta, org, editor. Inclusions and their
influence on material behavior. Materials Park OH: ISM and
ASM International; 1988. p. 123–31.
[156] Li T, Shimasaki S, Taniguchi S, Narita S, Uesugi K.
Three-dimensional Observation of nonmetallic inclusion
clusters in solid metal by X-ray micro-CT. ISIJ Int
2016;56(11):1989–95.
[157] Bi Y, Doctoral thesis Three dimensional determinations of
inclusions in ferroalloys and steel samples. Stockholm,
Sweden: KTH Royal Institute of Technology;; 2014.
[158] Bi Y, Karasev AV, Jönsson PG. Three dimensional evaluations
of REM clusters in stainless steel. ISIJ Int 2014;54(6):1266–73.