metals
Article
Formation and Removal Mechanism of Nonmetallic Inclusions
in 42CrMo4 Steel during the Steelmaking Process
Tong Qiao 1 , Guoguang Cheng 1, *, Yu Huang 1 , Yao Li 1 , Yanling Zhang 1 and Zhanchun Li 2
Citation: Qiao, T.; Cheng, G.; Huang,
Y.; Li, Y.; Zhang, Y.; Li, Z. Formation
and Removal Mechanism of
Nonmetallic Inclusions in 42CrMo4
Steel during the Steelmaking Process.
Metals 2022, 12, 1505. https://
doi.org/10.3390/met12091505
Academic Editors: Alexander
Ivanovich Zaitsev and
Lauri Holappa
Received: 13 August 2022
Accepted: 9 September 2022
Published: 11 September 2022
Publisher’s Note: MDPI stays neutral
with regard to jurisdictional claims in
published maps and institutional affil-
iations.
Copyright: © 2022 by the authors.
Licensee MDPI, Basel, Switzerland.
This article is an open access article
distributed under the terms and
conditions of the Creative Commons
Attribution (CC BY) license (https://
creativecommons.org/licenses/by/
4.0/).
Metals 2022, 12, 1505. https://doi.org/10.3390/met12091505
1 State Key Laboratory of Advanced Metallurgy, University of Science and Technology Beijing,
Beijing 100083, China
2 Jiangsu Yonggang Group Co., Ltd., Zhang Jiagang 215628, China
* Correspondence: chengguoguang@metall.ustb.edu.cn
Abstract: Nonmetallic inclusions are harmful to the quality of 42CrMo4 steel. Therefore, the formation
and removal mechanism of inclusions in 42CrMo4 steel during the steelmaking process is investigated
by industrial trials. The characteristics of inclusions in specimens were analyzed by scanning electron
microscopy and energy dispersive spectroscopy. The main type of inclusions in molten steel in the
early stage of ladle furnace (LF) refining is MgO-Al2 O3 inclusions of irregular shape. CaO begins
to appear in MgO-Al2 O3 inclusions in the middle and late stages of LF. In the vacuum degassing
(VD) refining stage, the inclusions in molten steel completely change into low-melting-point CaO-
MgO-Al2 O3 inclusions. The existence of [Mg] in molten steel is the fundamental reason for the
formation of a large number of MgO-Al2 O3 inclusions. Thermodynamic calculation shows that the
refractory mainly transfers [Mg] to the liquid steel in the LF refining stage, whereas the slag mainly
transfers [Mg] to the liquid steel in the VD refining stage. Kinetic calculation indicates that MgO-
Al2 O3 inclusions could be removed from molten steel faster than low-melting-point CaO-MgO-Al2 O3
inclusions. The fundamental reason for the different removal behavior of the two types of inclusions
is that the interfacial tension between the low-melting-point CaO-MgO-Al2 O3 inclusions and the
liquid steel is 50% lower than that of the MgO-Al2 O3 inclusions.
Keywords: nonmetallic inclusions; MgO-Al2 O3 ; CaO-MgO-Al2 O3 ; thermodynamic; kinetic
1. Introduction
The 42CrMo4 type of steel is a medium carbon alloy structural steel that has high
strength, hardenability, toughness, creep strength, and endurance strength at high tempera-
tures [1–5]. It is commonly used to prepare pitch and yaw bearings in wind power steel.
Because of the harsh working environment, complex load-bearing conditions, and the high
cost of lifting and replacing bearings, wind power equipment needs longer service life and
higher stability. Inclusions in steel are the key factor affecting fatigue life [6–8]. Therefore,
it is important to study the formation and removal mechanism of inclusions in 42CrMo4
steel during smelting to reasonably control inclusions in molten steel and improve the final
product quality of 42CrMo4 steel.
Aluminum deoxidizes 42CrMo4 steel, and the deoxidized product is Al2 O3 [9,10].
However, the main inclusions found in the smelting process are MgO-Al2 O3 inclusions,
which will change into CaO- MgO-Al2 O3 inclusions along with the smelting process.
The presence of [Mg] in liquid steel leads to the transformation of Al2 O3 inclusions into
MgO-Al2 O3 , but there is still controversy about the source of [Mg] in molten steel. Some
authors [11–14] have pointed out that slag containing MgO and magnesia refractory reacted
with liquid steel and transferred [Mg] to liquid steel together. Increasing [Mg] content
in liquid steel promotes inclusion change from simple Al2 O3 inclusions to MgO-Al2 O3
inclusions. However, some authors [15,16] have noted that the transfer of [Mg] from slag
to molten steel was dominant, and refractory materials had little effect on it. There are
https://www.mdpi.com/journal/metals
Metals 2022, 12, 1505 2 of 17

also authors [17–19] who have pointed out that the displacement reaction between [Al]
in molten steel and MgO in refractories was the main reason for the increase in [Mg] in
molten steel, which further determined the formation of MgO-Al2 O3 inclusions. As for
the transformation mechanism of MgO-Al2 O3 inclusions to CaO-MgO-Al2 O3 inclusions,
a large number of authors [20–24] have reported that [Ca] reduced MgO in MgO-Al2 O3
inclusions to CaO and eventually formed low-melting-point CaO-MgO-Al2 O3 inclusions.
However, compared with solid inclusions, low-melting-point liquid inclusions are more
difficult to remove from molten steel [25–31]. As a result, the main type of inclusions
observed in molten steel at the later stage of smelting is low-melting-point liquid inclusions,
which will be unfavorable to the final product quality. For 42CrMo4 steel, which uses the
Al deoxidation process and the smelting route is Electric furnace (EAF) → Ladle refining
furnace (LF) → Vacuum refining furnace (VD) → Continuous casting (CC), the source
of [Mg] in molten steel during smelting, the transformation process of inclusions from
MgO-Al2 O3 to liquid CaO-MgO-Al2 O3 , and the removal mechanism of inclusions have
never been specifically studied.
In the present work, the samples were taken at ladle furnace (LF) and vacuum de-
gassing (VD) in a plant trial to reveal the source of [Mg] in liquid steel and the formation
mechanism of CaO-MgO-Al2 O3 inclusions. The removal mechanism of low-melting-point
CaO-MgO-Al2 O3 inclusions was studied by kinetic calculation. The morphology, com-
position, and numbers of inclusions in samples were determined by scanning electron
microscopy and energy dispersive spectroscopy.

2. Methodology
2.1. Experimental Procedures
The industrial trials of 42CrMo4 steel were carried out in a steel plant. The steelmaking
route is Electric furnace (EAF) → Ladle refining furnace (LF) → Vacuum refining furnace
(VD) → Continuous casting (CC), as shown in Figure 1. Al blocks are used as deoxidizers for
strong deoxidation after steel tapping in the electric furnace, Si-Mn alloy and ferrochrome
alloy are used for alloying, and then lime is added for slagging. The slag used in the LF
stage has high basicity and strong reducing properties, and its binary basicity can reach
around 7. In vacuum treatment in the VD stage at vacuum degrees below 66.7 pa, argon gas
is passed through for stirring to promote rapid removal of hydrogen from the steel under
vacuum conditions. At the end of the vacuum, to make the steel composition qualified,
Al wire is added to increase the Al content in the molten steel to about 0.021%. The argon
blowing pressure and flow rate are adjusted for a soft blowing operation, which lasts a
total of 20 min. Finally, the liquid steel is transported to the continuous casting platform for
casting. The entire sampling process was carried out for a certain furnace: LF refining in
the early stage, LF refining in the middle stage, LF refining at the end, VD vacuum at the
end (sampling immediately after VD vacuum), soft blowing for 10 min, and soft blowing
at the end (for 20 min), respectively marked as samples 1~6. The slag was sampled at the
beginning of LF refining, the middle of LF refining, the end of LF refining and
Metals 2022, 12, x FOR PEER REVIEW 3 ofthe
19 end of
VD vacuum, respectively marked as samples 1~4. Since the composition of the slag after
VD vacuum was basically unchanged, the slag at the soft blowing stage was not sampled.
Steel samples
Deoxidizer: Al
Slag samples

EAF LF VD CC

1st 2nd 3rd 4th 5th 6th

Figure1.1.Schematic
Figure Schematic illustration
illustrationofofsmelting
smeltingprocess andand
process sampling locations.
sampling locations.

2.2. Composition Analysis and Inclusion Observation

The steel sample is processed into a round bar (Φ 5 × 50 mm), and the total oxygen
content (T.O) in the sample is detected by the inert gas fusion-infrared absorptiometry
method. The contents of Ca and Mg in steel were determined by the inductively coupled
plasma optical emission spectrometry method (ICP-OES, Thermo Fisher, Waltham, MA,
USA) and other alloy elements using the direct reading spectrometer ARL8860 (Thermo
Metals 2022, 12, 1505 3 of 17

2.2. Composition Analysis and Inclusion Observation

The steel sample is processed into a round bar (Φ 5 × 50 mm), and the total oxygen
content (T.O) in the sample is detected by the inert gas fusion-infrared absorptiometry
method. The contents of Ca and Mg in steel were determined by the inductively coupled
plasma optical emission spectrometry method (ICP-OES, Thermo Fisher, Waltham, MA,
USA) and other alloy elements using the direct reading spectrometer ARL8860 (Thermo
Fisher, Waltham, MA, USA) for testing; the main components of the sample were as shown
in Table 1. The chemical composition of the slag was detected by X-ray fluorescence
analysis (XRF, PANalytical B.V., Almelo, The Netherlands); the results were as shown in
Table 2, which shows the slag basicity was around 7. In order to facilitate observation
and statistics, the obtained steel samples were cut to obtain metallographic samples of
10 mm × 10 mm × 10 mm. After grinding and polishing, scanning electron microscopy
(FEI, Hillsboro, OR, USA) and energy dispersive spectrometry (FEI, Hillsboro, OR, USA)
were used for scanning statistics under fixed magnification, and the types and chemical
components of inclusions obtained were analyzed.

Table 1. Chemical compositions of 42CrMo4 steel (mass percent).

Sample No. C Si Mn S Cr Mo Al Ca Mg T.O

1 0.280 0.158 0.618 0.0040 0.879 0.132 0.036 0.0015 0.0006 0.0024
2 0.403 0.204 0.678 0.0022 0.985 0.156 0.024 0.0009 0.0006 0.0029
3 0.408 0.213 0.677 0.0019 0.982 0.155 0.019 0.0014 0.0007 0.0036
4 0.418 0.208 0.670 0.0018 0.986 0.154 0.008 0.0017 0.0006 0.0027
5 0.415 0.209 0.675 0.0017 0.986 0.156 0.021 0.0019 0.0005 0.0023
6 0.404 0.208 0.672 0.0017 0.978 0.155 0.021 0.0010 0.0005 0.0014

Table 2. Chemical compositions of slag (mass percent).

Sample No. CaO SiO2 Al2 O3 MgO

1 58.6 8.57 24.2 5.33
2 56.4 7.92 26.5 5.87
3 55.4 8.26 27.2 5.93
4 54.2 8.7 28.2 6.31

3. Results
3.1. Characterization of Inclusions
By scanning the surface of each sample, it was found that the main inclusions in the
steel were MgO-Al2 O3 inclusions with irregular shapes in the early stage of LF smelting, as
shown in Figure 2a. The mass fraction of each component of the inclusions is marked in the
figure, and the mass fraction of MgO was lower than that of Al2 O3 . The elemental mapping
of typical MgO-Al2 O3 inclusions is shown in Figure 3a. Mg, Al, and O elements were
evenly distributed in the inclusions. In the middle stage of LF smelting, the main types
of inclusions remained unchanged and were still MgO-Al2 O3 inclusions, but CaO began
to appear in some inclusions, as shown in Figure 2(b2). The elemental mapping of typical
inclusions is shown in Figure 3b. The distribution of Mg and Al was relatively uniform,
whereas Ca was unevenly distributed in the outer layer of inclusion. In this smelting stage,
a small number of spherical inclusions with high CaO content also began to appear in the
molten steel, as shown in Figure 2(b3). At the end of LF smelting, CaO generally appeared
in inclusions, and the shape of CaO-MgO-Al2 O3 inclusions containing a small amount of
CaO was irregular, as shown in Figure 2(c1). With the further increase in CaO content in
inclusions, the morphology of inclusions began to change from irregular shape to spherical
inclusions with smooth external contour, as shown in Figure 2(c2). The elemental mapping
of typical inclusions is shown in Figure 3c. In the outer layer of the inclusions, the Ca
element further accumulated, whereas the Mg element disappeared, and the distribution
of the Al element decreased slightly. In the center of inclusion, the distribution of Mg and
Metals 2022, 12, 1505 4 of 17

Al was more concentrated, but the Ca element was not present. At the end of VD vacuum
refining, the main type of inclusions in molten steel was CaO-MgO-Al2 O3 inclusions with
a spherical shape, and the content of MgO in the inclusions decreased obviously, as shown
in Figure 2d. The elemental mapping of typical CaO-MgO-Al2 O3 inclusions is shown
in Figure 3d. It can be seen that the distribution of Ca and Al elements in the inclusion
is uniform. After 10 and 20 min of VD soft blowing, the main types of inclusions were
the same as those at the end of the VD vacuum. The main inclusions were CaO-MgO-
Al2 O3 inclusions containing a small amount of MgO, as shown in Figure 2. The elemental
mapping of typical CaO-MgO-Al2 O3 inclusions is shown in Figure 3e,f. It can be seen that
Metals 2022, 12, x FOR PEER REVIEWthe distribution of Ca and Al elements in the inclusions was uniform, but
5 of the
19 distribution

of Mg elements was uniform or a with small amount of aggregation.

Figure
Figure2.2.Morphology
Morphology andandcomposition
compositionof typical inclusions
of typical encountered
inclusions in samples:
encountered (a1) a MgO-
in samples: (a1) a MgO-Al2 O3
Al2O3 inclusion in sample 1; (a2) a MgO-Al2O3 inclusion in sample 1; (a3) a MgO-Al2O3 inclusion in
inclusion in sample 1; (a2) a MgO-Al2 O3 inclusion in sample 1; (a3) a MgO-Al2 O3 inclusion in sample
sample 1; (b1) a MgO-Al2O3 inclusion in sample 2; (b2) a CaO-MgO-Al2O3 inclusion in sample 2;
1; (b1)
(b3) a MgO-Al22O
a CaO-MgO-Al inclusioninin
O33 inclusion sample
sample 2; (b2)
2; (c1) a CaO-MgO-Al
a CaO-MgO-Al 2 O3 inclusion
2O3 inclusion in samplein3;sample
(c2) a 2; (b3) a CaO-
MgO-Al2 O32O
CaO-MgO-Al inclusion
3 inclusion inin
sample
sample2;3;(c1)
(c3)aaCaO-MgO-Al
CaO-MgO-Al2O 2O 3 inclusion
3 inclusion in in sample
sample 3; (c2)
3; (d1) a CaO-MgO-Al2 O3
a CaO-
MgO-Al 2O3 inclusion in sample 4; (d2) a CaO-MgO-Al2O3 inclusion in sample 4; (d3) a CaO-MgO-
inclusion in sample 3; (c3) a CaO-MgO-Al2 O3 inclusion in sample 3; (d1) a CaO-MgO-Al2 O3 inclusion in
Al2O3 inclusion in sample 4; (e1) a CaO-MgO-Al2O3 inclusion in sample 5; (e2) a CaO-MgO-Al2O3
sample 4;
inclusion in(d2)
samplea CaO-MgO-Al 2 O3 inclusion
5; (e3) a CaO-MgO-Al in sample 4; (d3) a CaO-MgO-Al2 O
2O3 inclusion in sample 5; (f1) a CaO-MgO-Al2O 3 inclusion in sample 4;
3 inclusion

(e1)
in a CaO-MgO-Al
sample O
6; (f2) a CaO-MgO-Al
2 3 inclusion
2O 3 in sample
inclusion in 5;
sample (e2)
6; a CaO-MgO-Al
(f3) a CaO-MgO-Al O2
2 3O 3 inclusion
inclusion inin sample 5; (e3) a CaO-
sample
6.
MgO-Al O inclusion in sample 5; (f1) a CaO-MgO-Al O inclusion in sample 6; (f2) a CaO-MgO-Al O
2 3 2 3 2 3
inclusion in sample 6; (f3) a CaO-MgO-Al2 O3 inclusion in sample 6.
Metals 2022,
Metals 2022, 12,
12, 1505
x FOR PEER REVIEW 65of 19
of 17

Figure 3. Elemental mapping of typical inclusions in samples: (a) a typical MgO-Al2O3 inclusion in
Figure 3. Elemental mapping of typical inclusions in samples: (a) a typical MgO-Al2 O3 inclusion in
sample 1; (b) a typical MgO-Al2O3 inclusion in sample 2; (c) a typical CaO-MgO-Al2O3 inclusion in
sample 1; (b) a typical MgO-Al2 O3 inclusion in sample 2; (c) a typical CaO-MgO-Al2 O3 inclusion
sample 3; (d) a typical CaO-MgO-Al2O3 inclusion in sample 4; (e) a typical CaO-MgO-Al2O3 inclu-
in sample
sion 3; (d)5; a(f)typical
in sample CaO-MgO-Al
a typical 2 O23Oinclusion
CaO-MgO-Al in sample 4; (e) a typical CaO-MgO-Al2 O3
3 inclusion in sample 6.
inclusion in sample 5; (f) a typical CaO-MgO-Al2 O3 inclusion in sample 6.
3.2. Composition
3.2. Composition Distribution
Distribution ofof Inclusions
Inclusions
The inclusion
The inclusion compositions
compositions observed
observed at at each
each stage
stage were
were marked
marked in in the
the CaO-MgO-
CaO-MgO-
Al 2O3 ternary phase diagram, as shown in Figure 4. The ternary phase diagram of CaO-
Al2 O3 ternary phase diagram, as shown in Figure 4. The ternary phase diagram of CaO-
MgO-Al2O
MgO-Al 3 at 1600 °C ◦ was calculated by FactSage8.0 thermodynamics software (version
2 O3 at 1600 C was calculated by FactSage8.0 thermodynamics software (version
8.0, CRCT, Montreal,
8.0, CRCT, Montreal, Canada),Canada), and and thethe red
red line
line area
area denotes
denotes the
the pure
pure liquid
liquid oxide
oxide area.
area.
Aluminum blocks were used for deoxidation in the smelting process, and Al2 O3 inclusions
Aluminum blocks were used for deoxidation in the smelting process, and Al 2O 3 inclusions
would be
would be rapidly
rapidly formed
formed in in the
the molten
molten steel.
steel. However,
However, at at the
the initial
initial stage
stage of
of LF
LF refining,
refining,
the main
the main type
type ofof inclusions
inclusionsin inliquid
liquidthe thesteel
steelwaswasMgO-Al
MgO-Al 2O3Oinclusions.
2 3 inclusions.With the
With devel-
the de-
opment of LF
velopment of refining,
LF refining,a new component,
a new component, CaO,CaO,began to appear
began in theinMgO-Al
to appear 2O3 inclu-
the MgO-Al 2 O3
sions. ManyMany
inclusions. studies have have
studies found that the
found thatimpure alloys
the impure [32] and
alloys [32] the
andreaction between
the reaction slag
between
and and
slag steelsteel
[33,34] transfer
[33,34] [Ca][Ca]
transfer to the steel,
to the resulting
steel, in the
resulting appearance
in the appearance of CaO
of CaOelements
elementsin
the inclusions of the non-calcium-treated LF stage. As can be seen from
in the inclusions of the non-calcium-treated LF stage. As can be seen from Figure 4, the Figure 4, the com-
position of the
composition inclusions
of the approached
inclusions approached thethe
liquid
liquidoxide
oxideregion
regionasasthe
thesmelting
smeltingproceeded,
proceeded,
and the morphology of the inclusions also changed changed from irregular
irregular toto a spherical
spherical shape.
shape.
the end
After the end ofof the
the VDVD vacuum
vacuum treatment,
treatment, MgO-Al
MgO-Al22O33 inclusions in liquid steel were
mostly transformed into sphericalspherical liquid
liquid low-melting-point
low-melting-point CaO-MgO-Al
CaO-MgO-Al22O33 inclusions,
which were principally in the pure liquid oxide region. region. After soft blowing, the main type
unchanged.
of inclusions remained unchanged.
Metals 2022, 12, x FOR PEER REVIEW
Metals 2022, 12, 1505 6 of 17

Figure 4. Composition distributions (mass fraction) of inclusions in CaO-MgO-Al2 O3 phase diagrams.

Figure 4. Composition distributions (mass fraction) of inclusions in CaO-MgO-Al2O
(L: Liquid Oxide; CA2 : CaO-2Al2 O3 ; CM2 A8 : CaO-2MgO-8Al2 O3 ; CM2 A14 : CaO-2MgO-14Al2 O3 ;
grams. (L: 2Liquid
CA6 : CaO-6Al Oxide; 2CA
O3 ; Spinel:MgO-Al O3 ).2: CaO-2Al2O3; CM2A8: CaO-2MgO-8Al2O3; CM2A14: C
14Al2O3; CA6: CaO-6Al2O3; Spinel:MgO-Al2O3).
3.3. Quantity Distribution of Inclusions
It can be seen from Figure 4 that the inclusions are mainly divided into three types:
3.3. Quantity Distribution of Inclusions
MgO-Al2 O3 inclusions without CaO, CaO-MgO-Al2 O3 inclusions in the solid–liquid two-
phaseIt can be
region, andseen from Figure
CaO-MgO-Al 4 thatinthe
2 O3 inclusions inclusions
the pure are region.
liquid phase mainly Thedivided
content into t
of CaO in the CaO-MgO-Al O inclusions in the
MgO-Al2O3 inclusions without CaO, CaO-MgO-Al2O3 inclusions in
2 3 solid–liquid two-phase region ranged
the solid–l
from 0 to 30%. The number of inclusions of different types in the samples was further
phase
analyzed, region, andinCaO-MgO-Al
as shown Figure 5. In this 2O3 inclusions in the pure liquid phase region. T
paper, all inclusions larger than 3 µm on the
of CaOofin
surface thesample
each CaO-MgO-Al
were counted, 2O3so inclusions in the
the total number solid–liquid
of inclusions in each two-phase
sample regi
was different.
from 0 to 30%.In samples
The number1~3, the main type of inclusions
of inclusions was MgO-Al
of different types 2 Oin
3 inclusions,
the samples w
but four CaO-containing inclusions were found in sample 1, and the CaO content was
analyzed, as shown in Figure 5. In this paper, all inclusions larger than 3 μm on
less than 30%. The number of inclusions in sample 2 increased to 10 and included two
of each
liquid sample were
CaO-MgO-Al counted, so the total number of inclusions in each sample
2 O3 inclusions. In sample 3, the number of liquid CaO-MgO-Al2 O3
ent. In samples
inclusions increased1~3,
to 7, the main
whereas thetype
numberof inclusions
of MgO-Al2 Owas MgO-Al
3 inclusions 2O3 inclusions,
decreased slightly. but
In samples 4–6, the main inclusions changed
containing inclusions were found in sample 1, and the CaOinto liquid CaO-MgO-Al O inclusions. In
2 3 content was less
sample 4, the number of MgO-Al2 O3 inclusions decreased to 1. The number of CaO-MgO-
The number of inclusions in sample 2 increased to 10 and included two liquid C
Al2 O3 inclusions containing less than 30% CaO also decreased significantly. The number
Al 2O3 inclusions.
of liquid CaO-MgO-Al In2 Osample 3, the number of liquid CaO-MgO-Al2O3 inclusions
3 inclusions increased significantly to 17. As for samples 5 and
to 7, whereas
6, the number ofthe number
liquid CaO-MgO-Al of MgO-Al 2O3 inclusions
2 O3 inclusions decreased
was basically unchanged,slightly.
at 17 andIn samp
15, respectively, whereas the number of other types of inclusions
main inclusions changed into liquid CaO-MgO-Al2O3 inclusions. In sample 4, t was very small, and no
MgO-Al2 O3 inclusions were found in sample 6.
of MgO-Al2O3 inclusions decreased to 1. The number of CaO-MgO-Al2O3 inclu
taining less than 30% CaO also decreased significantly. The number of liquid C
Al2O3 inclusions increased significantly to 17. As for samples 5 and 6, the numb
CaO-MgO-Al2O3 inclusions was basically unchanged, at 17 and 15, respectivel
the number of other types of inclusions was very small, and no MgO-Al2O3
were found in sample 6.
Metals 2022, 12, x FOR PEER REVIEW 8 of 19
Metals 2022, 12, 1505 7 of 17

30
Liquid CaO-MgO-Al2 O3
CaO-MgO-Al2O 3(CaO<30%)
25 MgO-Al2O3

2 7

Number of inclusions
20

10
15 9
4

17
10 17 15

12 12
5 10

1 2 3
0 1 1
1 2 3 4 5 6
Sample number

Number of
Figure5.5.Number
Figure of inclusions
inclusionsofofdifferent
differenttypes in all
types in samples.
all samples.
4. Discussion
4.4.1.
Discussion
Thermodynamic Analysis of [Mg] Sources in Molten Steel
4.1. Thermodynamic Analysis
After the reaction of [Al]of [Mg]
with Sourcesand
refractory in Molten
slag in Steel
steel, the refractory and slag will
transfer
After[Mg]
the to the steel,
reaction ofwhich is therefractory
[Al] with fundamental
andcause
slag of
in the formation
steel, of MgO-Al
the refractory and2 Oslag
3 will
inclusions in this experiment. During the smelting process, the reactions between liquid
transfer [Mg] to the steel, which is the fundamental cause of the formation of MgO-Al2O3
steel and refractory, liquid steel, and slag are as follows [10]:
inclusions in this experiment. During the smelting process, the reactions between liquid
3MgO
steel and refractory, (s)in
liquid steel, and+slag
refractory 2[Alare
] = as
(Alfollows
2 O3 )in slag + 3[Mg]
[10]: (1)

3MgO
3(MgO )in slag
s in+ 2[Al
refracto
 2[Al]  (Al 2 O 3+
ry ] = (Al2 O3 )in slag
) in3sla[Mg
g  ]3[Mg ] (2) (1)
The reaction between steel and slag and refractory erosion in 42CrMo4 steel was theo-
retically calculated by using3(MgO) in slag 
FactSage8.0 2[Al]  ( Al2 O 3software.
thermodynamic )in slag  3[Mg ]
The actual composition (2)
of liquid steel detected in Table 1 was adopted. In order to simplify the calculation, the
The reaction
following between steel
model assumptions wereandused:slag (1)
and refractory
the erosion molten
reaction between in 42CrMo4 steel
steel and was the-
slag
oretically calculated by
reached equilibrium; (2)using FactSage8.0
the reaction between thermodynamic
liquid steel andsoftware.
refractoryThe wasactual composi-
balanced;
(3) the temperature was constant at 1600 was◦ C; (4) the erosion of refractory by slag was not
tion of liquid steel detected in Table1 adopted. In order to simplify the calculation,
considered.
the followingThe modelmassassumptions
fraction ratio of weremolten
used:steel
(1)tothe
slagreaction
and molten steel tomolten
between refractory
steel and
was 100:1 during calculation. The calculation results are shown in Figure 6. To increase
slag reached equilibrium; (2) the reaction between liquid steel and refractory was bal-
the number of samples and thus improve the accuracy of the model, we conducted five
anced; (3) the temperature was constant at 1600 °C; (4) the erosion of refractory by slag
experiments under the same conditions, and full-process sampling was performed for each
was not considered.
experiment. A~E are the Thefivemass fraction conducted,
experiments ratio of moltenand 1~4 steel
are to
theslag and molten
sampling numberssteel to
refractory was 100:1
for LF refining in theduring calculation.
early stage, LF refining Theincalculation
the middle results
stage, LF are shownatin
refining theFigure
end, 6. To
increase
and VD the number
vacuum at theofend
samples
for eachand thus improve
experiment. theseen
It can be accuracy
that both of slag
the and
model,
refrac-we con-
ducted five experiments
tory transferred under
[Mg] to liquid the In
steel. same conditions,
the LF and full-process
refining process, the transfer sampling
of [Mg] from was per-
refractory
formed fortoeach
molten steel was more
experiment. A~Ethan areslag,
the which played a majorconducted,
five experiments role. With the andprogress
1~4 are the
sampling numbers for LF refining in the early stage, LF refining in the middlewas
of smelting, the content of [Mg] transferred by refractory gradually decreased, which stage, LF
related to the decrease in [Al] content in molten steel. The decrease in [Al] content reduced
refining at the end, and VD vacuum at the end for each experiment. It can be seen that
the corrosion ability of liquid steel to refractory. The transfer of [Mg] content from slag to
both slag and refractory transferred [Mg] to liquid steel. In the LF refining process, the
molten steel gradually increased to a stable level. In the VD refining process, molten steel
transfer
and slagofwere
[Mg]fully
from refractory
stirred. to molten [Mg]
Slag-transferred steelcontent
was more than slag,
to molten steel which played
was higher thana major
role. With the progress of smelting, the content
that transferred by refractory, which played a major role. of [Mg] transferred by refractory gradually
decreased, which was related to the decrease in [Al] content in molten steel. The decrease
in [Al] content reduced the corrosion ability of liquid steel to refractory. The transfer of
[Mg] content from slag to molten steel gradually increased to a stable level. In the VD
refining process, molten steel and slag were fully stirred. Slag-transferred [Mg] content to
molten steel was higher than that transferred by refractory, which played a major role.
Metals 2022, 12, x FOR PEER REVIEW 9 of 19
Metals
Metals2022,
2022,12,
12,x1505
FOR PEER REVIEW 9 of
8 of19
17

0.0005
0.0005 from slag
from refractory
from slag
from refractory
0.0004
0.0004

of Mg/wt.%
0.0003

of Mg/wt.%
0.0003
Content
0.0002
Content

0.0002

0.0001
0.0001

0.0000
1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4
0.0000
1 2 A3 4 1 2 B3 4 1 2 C3 4 1 2 D3 4 1 2 E3 4
A B C D E
Five experiments
Five experiments
Figure 6. The content of magnesium provided by slag and refractory to molten steel at 1600 °C.
Thecontent
Figure6.6.The contentofofmagnesium
magnesiumprovided
providedby
byslag
slagand
andrefractory
refractoryto
tomolten
moltensteel
steelat
at1600 ◦ C.
1600°C.
Figure
In order to verify the accuracy of the thermodynamic calculation, error analysis was
Inorder
In
conducted order to
ontothe verify
verify theaccuracy
the
calculationaccuracy of
resultsof the
ofthe thermodynamic
thethermodynamic calculation,
calculation,
five furnace experiments, error
error
as analysis
analysis
shown was
was
in Figure
conducted
conducted
7. Mg is a trace on the
on the calculation
calculation
element, results
and results
Kang [35]of the
of the five furnace
five furnace
pointed experiments,
out thatexperiments, as shown in Figure
as shown inerror
there is 0.0001–0.0002% Figurein7.
Mg
7.
theMg isisa atrace
determination element,
trace element,
of Mg and
andKang
Kang
content [35]
by pointed
[35]
the pointed outout
ICP method. that there
that
The thereis 0.0001–0.0002%
resultisof0.0001–0.0002%
thermodynamic error in the
error in
calcu-
determination
the determination
lation of
is comparedofwithMg content
Mg content by the
byof
the result ICP method.
theexperimental The result
ICP method.detection, of thermodynamic
The resultand of thermodynamic calculation
the fluctuation calcu-
of the
is compared
lation
value with than
compared
is not more the result
with the of experimental
result
0.00015%. detection,
of experimental
Therefore, and theand
detection,
the thermodynamic fluctuation of results
the value
the fluctuation
calculation theis
of can
not
value more than 0.00015%. Therefore, the thermodynamic calculation results can reflect the
reflectisthenot more of
content than
[Mg]0.00015%.
in moltenTherefore,
steel well.the thermodynamic calculation results can
content of [Mg] in molten steel well.
reflect the content of [Mg] in molten steel well.
0.0010
1
0.0010 2
13
24
0.0008 3
4
0.0008
of Mg/wt.%

0.0006
of Mg/wt.%

0.0006
Content

0.0004
Content

0.0004

0.0002
0.0002

0.0000
1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4
0.0000
1 2 A3 4 1 2 B3 4 1 2 C3 4 1 2 D3 4 1 2 E3 4
A B C
Five experiments D E

Five experiments
Erroranalysis
Figure7.7.Error
Figure analysisof
ofcalculation
calculationresults
resultsof
offive
fivefurnaces.
furnaces.
Figure 7. Error analysis of calculation results of five furnaces.
4.2. Formation Mechanism of CaO-MgO-Al2 O3 Inclusions
4.2. Formation Mechanism of CaO-MgO-Al2O 3 Inclusions
After slag
4.2. Formation and refractories
Mechanism of transferred
CaO-MgO-Al [Mg] into liquid steel, MgO-Al O inclusions
2O3 Inclusions
After slag and refractories transferred [Mg] into liquid steel, MgO-Al22O33inclusions
were rapidly generated in liquid steel, so a large number of such inclusions were observed
wereAfter slag
rapidly and refractories
generated in liquidtransferred [Mg]number
steel, so a large into liquid steel,
of such MgO-Alwere
inclusions 2O3 inclusions
observed
in the early stage of LF refining. The Mg-Al-O equilibrium phase diagram of 42CrMo4
were
in therapidly generated
early stage of LFin liquid
refining.steel,
The so a large
Mg-Al-O number of
equilibrium such inclusions
phase diagram were
of observed
42CrMo4
steel at 1600 ◦ C was calculated by FactSage8.0 software, as shown in Figure 8. The dotted
in theatearly
steel 1600 stage
°C was ofcalculated
LF refining.by The Mg-Al-O
FactSage8.0 equilibrium
software, phaseindiagram
as shown Figure 8.ofThe
42CrMo4
dotted
line shows the dissolved oxygen contents in molten steel. Figure 9a contains four areas:
steel at 1600the
line shows °C dissolved
was calculated by contents
oxygen FactSage8.0 software,
in molten as shown
steel. Figure in9aFigure 8. The
contains fourdotted
areas:
the upper part is the MgO inclusion generation area, the middle part is the MgO-Al2 O3
line showspart
the upper the dissolved
is the MgO oxygen contents
inclusion in molten
generation area,steel. Figure 9a
the middle contains
part four areas:
is the MgO-Al 2O3
inclusion generation area, the lower part is the Al2 O3 inclusion generation area, and the
the upper part is the MgO inclusion generation area, the middle part is the MgO-Al2O3
 Metals 2022, 12, x FOR PEER REVIEW 10 of 19

Metals 2022, 12, 1505 inclusion generation area, the lower part is the Al2O3 inclusion generation area, and 9 ofthe
17
lower-left corner is the liquid steel area. It can be seen that the formation area of MgO-
Al2O3 inclusions is the largest. A small amount of [Mg] contents in liquid steel could form
MgO-Al2Ocorner
lower-left 3 inclusions. Figuresteel
is the liquid 8b,c area.
showsItthe caneffect
be seen of adding
that the1formation
ppm and 3area ppm of [Ca]
MgO- to
liquid steel on inclusions, respectively. There are four areas:
Al2 O3 inclusions is the largest. A small amount of [Mg] contents in liquid steel could form the upper part is the MgO
inclusion
MgO-Al 2 Ogeneration
3 inclusions. area, the middle
Figure 8b,c shows part is the theeffect
MgO-Al 2O3 inclusion
of adding 1 ppm generation
and 3 ppmarea, [Ca] the
to
lower steel
liquid part is onthe CaO-MgO-Al
inclusions, 2O3 liquid
respectively. oxideare
There generation
four areas: area,
theandupper the part
lower-left
is the corner
MgO
is the liquid
inclusion steel area.
generation area, the middle part is the MgO-Al2 O3 inclusion generation area, the
lowerThepartcomposition
is the CaO-MgO-Al of liquid 2 Osteel
3 liquidin LF andgeneration
oxide VD is marked area,inandFigure 8. Without
the lower-left [Ca], is
corner as
shown in Figure
the liquid steel area. 8a, the composition points are all located in the formation area of MgO-
Al2OThe
3 inclusions,
composition indicating
of liquidthat a large
steel in LFnumber
and VDofisMgO-Al marked2O inclusions
in3 Figure were formed
8. Without [Ca], as in
the molten
shown in Figure steel8a, under this smeltingpoints
the composition condition.
are allWhen located 1 ppm
in the[Ca] existedarea
formation in the liquid
of MgO-
steel,
Al 2 O3 as shown inindicating
inclusions, Figure 8b,that theacomposition
large number points
of MgO-Alwere still2 O3 all located were
inclusions in theformed
area of
in the molten
MgO-Al steel underHowever,
2O3 inclusions. this smelting the Al condition.
2O3 inclusion When 1 ppm [Ca]
formation region existed in the liquid
disappeared. The
steel, as shown2Oin
CaO-MgO-Al Figureoxide
3 liquid 8b, the composition
formation region points were still
appeared. The all located
MgO-Al 2Oin the area of
3 inclusion re-
MgO-Al
gion became O inclusions. However,
2 3 smaller. It can be found that the Al O inclusion formation region
2 1 3ppm [Ca] was not enough to convert inclusions disappeared. The
CaO-MgO-Al
in 42CrMo4 steel 2 O3 liquid oxide formation
to CaO-MgO-Al region
2O3 liquid appeared.
oxide. WhenThe MgO-Al
3 ppm O3 inclusion
[Ca]2existed in theregion
liquid
became
steel, assmaller.
shown in It Figure
can be 8c,
found that 1 ppm
the MgO-Al 2O3 [Ca] was not
inclusion enougharea
formation to convert inclusions
in the phase diagram in
42CrMo4 steel to CaO-MgO-Al
almost disappeared 2 O3 liquid
and only occurred whenoxide. the When
dissolved 3 ppm [Ca]in
oxygen existed in the
the liquid liquid
steel was
steel,
high.asTheshown in Figure
formation area8c,ofthe MgO-Al2 O32O
CaO-MgO-Al inclusion formation
3 liquid oxide in thearea
phasein the phase diagram
diagram was fur-
almost
ther expanded. The composition points of liquid steel are all located in this region, was
disappeared and only occurred when the dissolved oxygen in the liquid steel indi-
high.
catingThe thatformation
CaO-MgO-Al area of2O CaO-MgO-Al
3 liquid oxide 2 O3was liquid oxide
stably in the phase
generated diagram was
in 42CrMo4 steelfurther
when
expanded.
there was The 3 ppm composition
[Ca] in liquidpoints of liquid
steel. It cansteel
be foundare allthat
located in this region,
increasing different indicating
calcium that con-
CaO-MgO-Al
tent in molten O
2 steel
3 liquid
had no effect on the MgO inclusion formation region, butwas
oxide was stably generated in 42CrMo4 steel when there 3 ppm
increasing
[Ca]
onlyin1 liquid steel. It can
ppm calcium be found that
significantly increasing
affected the MgO-Aldifferent calcium content in molten steel
2O3 inclusion and Al2O3 inclusion
had
formation area. Thus, a slight increase in calcium could change the only
no effect on the MgO inclusion formation region, but increasing type 1ofppm calcium
inclusions in
significantly
molten steel. affected the MgO-Al 2 O 3 inclusion and Al 2 O 3 inclusion formation area. Thus,
a slight increase in calcium could change the type of inclusions in molten steel.
0.01 0.01
(a) Number: [O ppm] 1 2 3 4 5 6 (b) Number: [O ppm] 1 2 3 4 5 6

MgO MgO
Content of Mg / wt.%
Content of Mg / wt.%

0.001 0.001

MgO-Al 2O3
MgO-Al2O3
70 50 40 30 20 15 10 5

0.0001
1E-4 0.0001
1E-4
200 150 100 70 50 40 30 20 15 10 5

Liquid Oxide
Liquid Steel Liquid Steel
100
200 150 Al2O3
0.00001
1E-5 0.00001
1E-5
1E-5
0.00001 0.0001
1E-4 0.001 0.01 0.1 1 0.00001
1E-5 0.0001
1E-4 0.001 0.01 0.1 1
Content of Al / wt.% Content of Al / wt.%

0.01
(c) Number: [O ppm] 1 2 3 4 5 6

MgO

MgO-Al2O3
Content of Mg / wt.%

0.001

200 150 100 70 50 40 30 20 15 10 7 5

0.0001
1E-4

Liquid Oxide
Liquid
Steel

0.00001
1E-5
0.00001
1E-5 0.0001
1E-4 0.001 0.01 0.1 1
Content of Al / wt.%

Figure8.8.Effect
Figure Effectofofdifferent
differentcalcium
calciumcontent
contenton
onMgO-Al2O
MgO-Al 3 3inclusions:
2O inclusions: (a)
(a) no
nocalcium;
calcium;(b)
(b)11ppm
ppm
[Ca];
[Ca];(c)
(c)33ppm
ppm[Ca]
[Ca](C:
(C:0.4%-Cr:
0.4%-Cr:0.98%-Mo:
0.98%-Mo:0.16%).
0.16%).
 the content of CaO-2Al2O3 inclusions began to decrease and liquid oxides began to appear.
When the calcium content reached 16ppm, the MgO-Al2O3 inclusions disappeared com-
pletely and only liquid low-melting-point oxides remained in the molten steel. The tran-
sition route of inclusions was: MgO-Al2O3 → CaO-2MgO-8Al2O3 → CaO-2MgO-8Al2O3 +
CaO-2Al2O3 → Liquid CaO-MgO-Al2O3. The content of CaO in inclusions increased grad-
Metals 2022, 12, 1505 ually. The inclusions in Figure 9b–d had the same transition route as Figure 9a, except10that
of 17
the quality of inclusions generated under different aluminum contents was different.

0.008 0.008
(a) (b)

0.007 0.007

Mass fraction of inclusions / wt.%

0.006 0.006

Mass fraction of inclusions / wt.%

0.005 Liquid Oxide 0.005
Liquid Oxide
Spinel
0.004
Spinel 0.004

0.003 0.003

0.002 CA2 0.002

CA2
CM2A8 CM2A8
0.001 0.001

0.000 0.000
0.0000 0.0005 0.0010 0.0015 0.0020 0.0000 0.0005 0.0010 0.0015 0.0020
Content of Ca / wt.% Content of Ca / wt.%
0.008 0.008
(c) (d)

0.007 0.007

0.006 0.006

Mass fraction of inclusions / wt.%

Mass fraction of inclusions / wt.%

0.005 0.005
Liquid Oxide
Spinel Liquid Oxide
0.004 0.004 Spinel

0.003 0.003

0.002 CA2 0.002

CM2 A8
CM2 A8
0.001 0.001 CA2

0.000 0.000
0.0000 0.0005 0.0010 0.0015 0.0020 0.0000 0.0005 0.0010 0.0015 0.0020
Content of Ca / wt.% Content of Ca / wt.%

Effect of
of calcium
calcium content ◦
Figure9.9.Effect
Figure content on
on inclusions
inclusionsininsmelting
smeltingprocess
processatat1600
1600C:
°C:(a)(a)
Al:Al:
0.036%; (b)(b)
0.036%; Al:
0.024%;
Al: (c) Al:
0.024%; 0.02%;
(c) Al: 0.02%;(d) Al:
(d) 0.008%; C: 0.4%;
Al: 0.008%; Cr: 0.98%;
C: 0.4%; Mo: 0.16%;
Cr: 0.98%; Mg: 0.0006%;
Mo: 0.16%; O: 0.003%;
Mg: 0.0006%; Spinel:
O: 0.003%;
Spinel:
MgO-Al MgO-Al 2O23A
2 O3 ; CM ; CM 2A8: CaO-2MgO-8Al
8 : CaO-2MgO-8Al 2O32; :CA
2 O3 ; CA 2: CaO-2Al
CaO-2Al O23S:
2 O3 ;2C ; CCaO-2SiO
2S: CaO-2SiO
2 . 2).

According
4.3. Removal to the observed
Mechanism inclusions distribution, MgO-Al2 O3 inclusions gradually
of Inclusions
changed to CaO-MgO-Al O
2 3
According to the observationinclusions
and from LF to
analysis ofVD. The CaO
inclusions incontent
42CrMo4 in the inclusions
steel, it was
increased gradually until the end of VD soft blowing, and most of the inclusions
shown that low-melting-point CaO-MgO-Al2O3 inclusions became the main inclusions in were in
the liquid oxide region. The calcium and aluminum in the liquid steel fluctuated greatly
liquid steel after LF, VD, and soft blowing. Soft blowing treatment could not effectively
according to the composition of the liquid steel detected. Therefore, in order to explore the
promote the removal of such inclusions. The solid MgO-Al2O3 inclusions disappeared af-
effect of calcium content on inclusion transformation in liquid steel, FactSage8.0 software
ter refining, indicating that the removal effects of different types of inclusions in molten
was used to calculate the effect of calcium content on inclusions in liquid steel under differ-
steel were different. The removal process of inclusions was divided into three steps:
ent aluminum content conditions, as shown in Figure 9. The calculation range of calcium
content was 0~0.004%; Mg content was 0.0006%; the total oxygen content was 0.003%. It can
be seen from Figure 9a that with the increase in calcium content, the content of MgO-Al2 O3
inclusions in molten steel gradually decreased, and CaO-2MgO-8Al2 O3 and CaO-2Al2 O3
inclusions began to appear. The content of CaO-2MgO-8Al2 O3 inclusions began to decrease
when the calcium content increased to 2 ppm, whereas the content of CaO-2Al2 O3 inclu-
sions continued to increase. When the calcium content increased to 5 ppm, the content
of CaO-2Al2 O3 inclusions began to decrease and liquid oxides began to appear. When
the calcium content reached 16ppm, the MgO-Al2 O3 inclusions disappeared completely
and only liquid low-melting-point oxides remained in the molten steel. The transition
route of inclusions was: MgO-Al2 O3 → CaO-2MgO-8Al2 O3 → CaO-2MgO-8Al2 O3 + CaO-
2Al2 O3 → Liquid CaO-MgO-Al2 O3 . The content of CaO in inclusions increased gradually.
The inclusions in Figure 9b–d had the same transition route as Figure 9a, except that the
quality of inclusions generated under different aluminum contents was different.

4.3. Removal Mechanism of Inclusions

According to the observation and analysis of inclusions in 42CrMo4 steel, it was shown
that low-melting-point CaO-MgO-Al2 O3 inclusions became the main inclusions in liquid
steel after LF, VD, and soft blowing. Soft blowing treatment could not effectively promote
the removal of such inclusions. The solid MgO-Al2 O3 inclusions disappeared after refining,
indicating that the removal effects of different types of inclusions in molten steel were
different. The removal process of inclusions was divided into three steps: polymerization
Metals 2022, 12, x FOR PEER REVIEW 12 of 19

Metals 2022, 12, 1505 11 of 17

polymerization growth, floating, and separation at the interface between steel and slag.
The third step determines whether the inclusion could be removed from the liquid steel
[26]. This section discusses the displacement, velocity, and force variation of different in-
growth, floating, and separation at the interface between steel and slag. The third step
clusions
determinesat whether
the interface between
the inclusion steel
could beand slag. from the liquid steel [26]. This section
removed
Based on the Strandh model [28,29], the
discusses the displacement, velocity, and force variation force analysis diagram
of different of inclusions
inclusions at the at the
interface betweensteel
interface between steel
andand slag is shown in Figure 10. Inclusion in the interface between
slag.
steel Based
and slag sported
on the Strandhby model
the four forcesthe
[28,29], worked together:
force analysis the buoyancy
diagram forceatFthe
of inclusions b, the drag
interface between steel and slag is shown in Figure 10. Inclusion in the interface
force Fd, the added mass force Ff, and the rebound force Fr. Inclusions of movement fol- between
steel and
lowed slag sported
Newton’s second the four forces worked together: the buoyancy force Fb , the
by law:
drag force Fd , the added mass force Ff , and the rebound force Fr . Inclusions of movement
followed Newton’s second law: 4 3 d 2Z
Fa  ma=Fb  Fr  Fd  Ff   RI I 2 (3)
4 33 d2 Z dt
Fa = ma = Fb + Fr + Fd + Ff = πR I ρ I (3)
3 dt2

10. The
Figure 10.
Figure Theforce
forceanalysis
analysisdiagram of inclusions.
diagram of inclusions.
The size of the buoyancy force, the drag force, the added mass force, and the rebound
forceThe size
could be of the buoyancy
calculated force,
according the drag force,
to Equations (4)–(7):the added mass force, and the rebound
force could be calculated according to  Equations (4)–(7):
Z σ − σIS
Fr = 2πRσMS − 1 − IM (4)
R I  Z σMS   
IM IS
Fr  2  R MS   1     (4)
µ M Z2  RI µM  MSZ 
µ dZ
2
Fd = 4πR I µS A · −1 2
−2 −1 + M (5)
µS RI µS RI µS dt
2 3
 
1 2ρM   ZM 3 ZρM  MZ ρZM  d2 ZM  dZ
 3  2  2
Ff = πRF ρS44RI ρS A−1  3  −1 −21 R 2+ 1 ρ  dt2  (6) (5)
3 dI  RI 4 R ρ2S  
 
S

 3 S 
 I 
 2 S
I
 RSI S dt
4 1 ρM Z 3 ρM Z ρ
Fb = πR3I g −1 3
− −1 2
+ M ρS − ρI (7)
3 4 ρS RI 4 ρ3 S RI ρS 2
2 3  1  M  Z 3  M  Z M  d Z
2
 RI S σMS is the 1interfacial
where R is the radiusFfofinclusion,  3  tension
  1between
 2  molten  steel
2 and (6)
 4  S inclusions
slag, σIM is the interfacial 3tension between  RI and S steel,
4  liquid  RI σISisS the
 dtinterfacial
tension between inclusions and slag, µM is the viscosity of liquid steel, µS is the viscosity of
slag, ρM is the density of liquid steel, ρS is the slag
4 3
 1  M
density, ρ is the density of inclusions,
 Z 3 3  M I  Z 2  M   Z
is the displacement Fb of
 inclusions
 RI g   at the interface
 1 3 between
  steel  1and2 slag,
 and  St 
is 
time. In
I  (7)
this model, it was assumed3 that the  S position
 4initial  RI of the S S  the
 RIwas below
4  inclusions 
interface.

When the displacement value reached twice the radius of the inclusions, it was considered
where
that theRinclusions
is the radius
brokeofaway
inclusion,
from the σMS is the interfacial
interface tension
and completely between
entered the slag.molten steel and
slag, According to Newton’s
σIM is the interfacial secondbetween
tension law, a second-order
inclusions anddifferential equation
liquid steel, could
σIS is be
the interfacial
obtained by simultaneous operation of Equations (3)–(6), as shown in Equation
tension between inclusions and slag, μM is the viscosity of liquid steel, μS is the viscosity (8). The
change
of slag, in
ρMinclusion displacement
is the density of liquid with time
steel, ρScould
is thebe obtained
slag by solving
density, ρI is thethe equation.
density of inclusions,
Z is the displacement of inclusions at the interface between steel and slag, and t is time. In
this model, it was assumed that the initial position of the inclusions was below the inter-
face. When the displacement value reached twice the radius of the inclusions, it was con-
sidered that the inclusions broke away from the interface and completely entered the slag.
According to Newton’s second law, a second-order differential equation could be
obtained by simultaneous operation of Equations (3)–(6), as shown in Equation (8). The
change in inclusion displacement with time could be obtained by solving the equation.
  1  Z3 3   Z2   
2  S   M  1 3   M  1 2  M    I  g
 
d 2Z  4  S  RI 4  S  RI S 
2
  
dt   1  M Z 3
3  Z 2
  
Metals 2022, 12, 1505  S    1 3   M  1 2  M   2  I  12 of 17
 
  4  S  RI 4  S  RI S  
Z    IS 
3 MS   1  IM 
    3   RI  2   
ρM MS
2 ρS 4 ρ −1
1 ρM Z 3 ρM Z
− 4 ρ −1 +ρ − ρI g
d2 Z S R 3 S 3 RI 2 S (8)
dt2
=  2   ρ  1  M3I  Z  32   M   Z 2 M  
R 1  M
ρS I 4 sρ  −1 Z 31 ρ M
− 4 ρ −3 1  Z  ρ
+ρ M +2ρI 2 
1    2  I 
 S 4  IS R S  RI S 
R 3 S R4I2
I  S 
σ −σ
3σMS RZ −1− IM σMS
IS
−     I
 3 1  
  Z2 3 Z 2 M 
3    
2 ρ 1 ρ M2 1 Z − 3 M
s S4 ρRI −
ρ
 S 3 4 ρM − 11 Z 2 + ρρM +2ρMI  1  (8)
I 6A
R
R
   2 
r    I 4  S   RI I 4  S   RI
S S R 3 S 2
S  
l


2 ρ 1 ρM Z 3 3 ρM Z 2 ρM
−6Aµ S R S 2
   I  Z 4 ρS  − 1 − − 1
R I Z 4 ρS  dZ R I
3 2 + ρS + 2ρ l
   M 1 2 2  2  M   1  M 
µM 

M
×  − S 1  R
µS
Z
2 I− 2 
RI
µ
S− 1 ZRI+ MS dZdt
µS
µ
RI µS dt

The low-melting-point CaO-MgO-Al2O3 inclusions detected in the smelting process

The low-melting-point CaO-MgO-Al2 O3 inclusions detected in the smelting process
were substituted into Equation (7) for solution, as shown in Figure 11. It can be seen that
were substituted into Equation (7) for solution, as shown in Figure 11. It can be seen
the displacement of the low-melting-point CaO-MgO-Al2O3 inclusions generated in the
that the displacement of the low-melting-point CaO-MgO-Al2 O3 inclusions generated in
smelting process at the interface between steel and slag was less than 2R. This kind of
the smelting process at the interface between steel and slag was less than 2R. This kind
inclusion could not be removed from the molten steel, but oscillated at the interface be-
of inclusion could not be removed from the molten steel, but oscillated at the interface
tween steel and slag, and finally stayed there. In actual production, the inclusions were
between steel and slag, and finally stayed there. In actual production, the inclusions were
likely to return to the molten steel due to the fluctuation of the interface.
likely to return to the molten steel due to the fluctuation of the interface.

2.0
Ratio of inclusion displacement to radius (Z/R)

1.5

1.0

0.5

0.0
0.0 5.0×10 -6 1.0×10 -5 1.5×10-5 2.0×10-5 2.5×10 -5 3.0×10 -5
Time / s

Figure 11. Displacement variation of

of low-melting-point
low-melting-point CaO-MgO-Al
CaO-MgO-Al22O
O33 inclusions at the interface
between steel and slag.

Since
Since aalarge
largenumber
numberofofMgO-Al
MgO-Al 2O2O3 3inclusions
inclusionswerewereobserved
observed in in
thethe
LF LF
refining process,
refining pro-
the
cess,removal process
the removal of suchofinclusions
process at the interface
such inclusions was calculated,
at the interface as shownas
was calculated, in shown
Figure 12.
in
The
Figuredisplacement changes of MgO-Al
12. The displacement changes2ofO3MgO-Al
inclusion2Oand low-melting-point
3 inclusion CaO-MgO-Al
and low-melting-point 2 O3
CaO-
inclusion
MgO-Al2O at3 inclusion
the interface were
at the compared
interface wereunder the same
compared radius
under (15 µm),
the same molten
radius steelmol-
(15 μm), and
slag conditions. The values of interfacial tension, density, and viscosity data
ten steel and slag conditions. The values of interfacial tension, density, and viscosity data required for the
calculations were calculated according to the study by Xuan et al. [27], as
required for the calculations were calculated according to the study by Xuan et al. [27], asshown in Table 3.
The displacement of the MgO-Al O inclusion reached 2R at 5.04 × 10 −6 s, whereupon it
shown in Table 3. The displacement 2 3 of the MgO-Al2O3 inclusion reached 2R at 5.04 × 10−6
could completely separate from
s, whereupon it could completely separate the moltenfrom steelthe
and entered
molten theand
steel slag. The displacement
entered the slag. The
of the low-melting-point CaO-MgO-Al2 O3 inclusion was obviously smaller than that of
the MgO-Al2 O3 inclusion at the same time, and reached the maximum displacement at
1.2 × 10−5 s. However, at this time, the inclusion could not completely break away from
the molten steel. Then, the displacement decreased, and the inclusion oscillated at the
interface until it stayed at the interface. Finally, the displacement of the inclusion was 1.57 R.
The MgO-Al2 O3 inclusion separated from the liquid steel and entered the slag phase faster
than the low-melting-point CaO-MgO-Al2 O3 inclusions. Therefore, the low-melting-point
 slag phase faster than the low-melting-point CaO-MgO-Al2O3 inclusions. Therefore, the
low-melting-point CaO-MgO-Al2O3 inclusion was not easily removed from liquid steel,
which was also the reason why a large number of such inclusions were still seen after VD.

Table 3. Calculated values of inclusion radius, interfacial tension, density, and viscosity.
Metals 2022, 12, 1505 13 of 17
RI σMS σIM σIS μM μS ρM ρS ρI
Inclusion Type
(μm) (N/m) (N/m) (N/m) (Pa•s) (Pa•s) (kg/m3) (kg/m3) (kg/m3)
Liquid CaO-MgO-Al2O3 15 1.36 O inclusion
CaO-MgO-Al 1.34 0.57not easily
was 0.0049removed
0.089from7001.65 2809.42
liquid steel, 3055.52
which was also the
2 3
Solid MgO-Al2O3 15 1.36 3.13 0.82 0.0049 0.089 7001.65 2809.42
reason why a large number of such inclusions were still seen after VD. 3550 [36]

5.0410-6s 1.210-5s
2.0
MgO-Al2O 3
Low Melting Point CaO-MgO-Al2O3

Ratio of inclusion displacement to radius (Z/R) 1.57

1.5

1.0

0.5

0.0
0 1×10 -5 2×10-5 3×10-5 4×10-5 5×10-5
Time / s

Figure12.
Figure Thedisplacement
12.The displacementof
ofdifferent
different inclusions
inclusions at
at the
the interface
interface between
between steel
steel and
and slag.
slag.

Table 3. Calculated values of inclusion radius, interfacial tension, density, and viscosity.
The velocity and resultant force changes of the low-melting-point CaO-MgO-Al2O3
RI inclusion
σ MS in Figureσ IM12 were σfurther calculated
µM andµanalyzed, ρas shown inρ Figure 13.ρIt can
IS S M S I
Inclusion Type
(µm) be seen
(N/m) that the(N/m)
inclusion velocity
(N/m) decreased
(Pa•s) to 0(Pa
at •1.2
s) × 10−5 s, when
(kg/m 3) the inclusion
(kg/m 3) had 3the
(kg/m )
Liquid
maximum displacement. As the resultant force on the inclusion was the resistance to in-
15 1.36removal,1.34
clusion the value 0.57
is −4×10−5 0.0049
N. Under the 0.089
action of7001.65 2809.42
the resultant 3055.52
force, the inclu-
CaO-MgO-Al2 O3
Solid MgO-Al2 O3 15 sion began
1.36 to change
3.13 from moving
0.82 out of
0.0049liquid steel
0.089 to moving
7001.65into liquid
2809.42steel.
3550 [36]it
Then,
oscillated until the velocity and resultant force changed to zero and stayed at the interface.

The velocity and resultant force changes of the low-melting-point CaO-MgO-Al2 O3

inclusion in Figure 12 were further calculated and analyzed, as shown in Figure 13. It can
be seen that the inclusion velocity decreased to 0 at 1.2 × 10−5 s, when the inclusion had
the maximum displacement. As the resultant force on the inclusion was the resistance to
inclusion removal, the value is −4 × 10−5 N. Under the action of the resultant force, the
inclusion began to change from moving out of liquid steel to moving into liquid steel. Then,
it oscillated until the velocity and resultant force changed to zero and stayed at the interface.
Under the same conditions of steel and slag, the low-melting-point CaO-MgO-Al2 O3
inclusions and MgO-Al2 O3 inclusions exhibited completely different movement behaviors
at the interface of steel and slag, which was related to their interfacial tension [26]. Therefore,
the interfacial tension of the two inclusions was calculated, as shown in Figure 14. The low-
melting-point CaO-MgO-Al2 O3 inclusions in the liquid oxide region had similar physical
properties, so the mean value was used to represent them. It can be seen that the interfacial
tension between low-melting-point CaO-MgO-Al2 O3 inclusions and slag was slightly
smaller than that between MgO-Al2 O3 inclusions and slag under the same conditions.
However, the interfacial tension between low-melting-point CaO-MgO-Al2 O3 inclusions
and molten steel was much less than that between MgO-Al2 O3 inclusions and molten steel,
which was less than 50% of the former. The large difference in the interfacial tension of the
two inclusions resulted in different movement behaviors at the interface between steel and
slag, which was the reason why the low-melting-point inclusions could not be removed
from molten steel and became the main type of inclusions in molten steel.
 0 0.0

R
-5.0×10 -5
-2
1.210 -5 s
Metals
Metals2022,
2022,12,
12,x1505
FOR PEER REVIEW -1.0×10 -4 15 of1419
of 17
0 1×10 -5 2×10 -5 3×10-5 4×10-5 5×10 -5 6×10 -5
Time / s

Figure
6 13. The velocity and resultant force changes of the low-melting-point
2.0×10 -4 CaO-MgO-Al2O3 inclu-
sions. Velocity
Resultant Force
-4
1.5×10
Under the same conditions of steel and slag, the low-melting-point
4
CaO-MgO-Al2O3
inclusions and MgO-Al2O3 inclusions exhibited completely different movement behaviors
1.0×10 -4interfacial tension [26]. There-

Resultant Force / N
at the interface of steel and slag, which was related to their
Velocity / ms-1
fore,2the interfacial tension of the two inclusions was calculated, as shown in Figure 14.
The low-melting-point CaO-MgO-Al2O3 inclusions in the liquid
5.0×10 -5
oxide region had similar
physical properties, so the mean value was used to represent them. It can be seen that the
interfacial
0 tension between low-melting-point CaO-MgO-Al 0.0 2O3 inclusions and slag was

slightly smaller than that between MgO-Al2O3 inclusions and slag under the same condi-
tions. However, the interfacial tension between low-melting-point-5.0×10 -5
CaO-MgO-Al2O3 inclu-
sions-2 and molten steel was much less than that between MgO-Al2O3 inclusions and molten
steel, which was1.210
less-5than
s 50% of the former. The large difference in the interfacial tension
-1.0×10 -4
of the0 two inclusions
1×10 -5
2×10 resulted
-5
3×10in
-5 different
4×10 -5 movement
5×10 -5
6×10behaviors
-5 at the interface between
steel and slag, which was the Timereason
/s why the low-melting-point inclusions could not be
removed from molten steel and became the main type of inclusions in molten steel.
Figure13.
Figure Thevelocity
13.The velocityand
andresultant
resultant force
force changes
changes of
ofthe
thelow-melting-point
low-melting-pointCaO-MgO-Al2 O2O
CaO-MgO-Al 3 inclusions.
3 inclu-

sions.
3.5
Low Melting Point CaO-MgO-Al2O3
Under theMgO-Al
same2Oconditions
3
of steel and slag, the 3.13
low-melting-point CaO-MgO-Al2O3
3.0
inclusions and MgO-Al2O3 inclusions exhibited completely different movement behaviors
at the interface of steel and slag, which was related to their interfacial tension [26]. There-
Interfacial Tension / N·m-1

fore,2.5
the interfacial tension of the two inclusions was calculated, as shown in Figure 14.
The low-melting-point CaO-MgO-Al2O3 inclusions in the liquid oxide region had similar
2.0 properties, so the mean value was used to represent them. It can be seen that the
physical
interfacial tension between low-melting-point CaO-MgO-Al2O3 inclusions and slag was
slightly
1.5 smaller than that between MgO-Al2O3 inclusions and slag under the same condi-
1.34
tions. However, the interfacial tension between low-melting-point CaO-MgO-Al2O3 inclu-
sions1.0and molten steel was much less than that between MgO-Al2O3 inclusions and molten
0.82
steel, which was less than 50% of the former. The large difference in the interfacial tension
0.57
of the
0.5two inclusions resulted in different movement behaviors at the interface between
steel and slag, which was the reason why the low-melting-point inclusions could not be
removed
0.0
from molten steel and became the main type of inclusions in molten steel.
Slag Liquid Steel
3.5
Figure14.
Figure Comparison
14.Comparison ofof
Low Melting interfacial
interfacial
Point tension
tension
CaO-MgO-Al between
between low-melting-point
low-melting-point CaO-MgO-Al
CaO-MgO-Al2O23O 3 inclusions
inclusions
2O3
and 3.13
andMgO-Al
MgO-Al 2OO3 inclusions.
3 inclusions.
2 MgO-Al 2O 3
3.0

According to the above calculation and analysis, the inclusion generation and removal
Interfacial Tension / N·m-1

mechanism
2.5 in 42CrMo4 steel could be summarized as the following process, as shown in
Figure 15. In the EAF, aluminum blocks were used for deoxidation, and Al2 O3 inclusions
were2.0formed in the molten steel. In the LF, [Mg] was transferred to the molten steel by [Al]
reducing refractory and slag, and MgO-Al2 O3 inclusions began to appear in the molten
steel.
1.5 With the transfer of [Ca] from slag to molten steel, the MgO-Al2 O3 inclusions were
1.34
transformed into CaO-MgO-Al2 O3 inclusions. In the VD, the content of [Ca] in molten steel
continued
1.0 to increase, and then the content of CaO in CaO-MgO-Al2 O3 inclusions increased,
which was transformed0.82 into low-melting-point CaO-MgO-Al2 O3 inclusions. A fraction
of the MgO-Al 0.57 O inclusions was removed from the molten steel during smelting. The
0.5 2 3
unremoved inclusions were converted into CaO-MgO-Al2 O3 inclusions. Because of the
low interfacial tension between the low-melting-point CaO-MgO-Al2 O3 inclusions and
0.0
the liquid steel, such Slag inclusions were notLiquid removed
Steel but remained at the interface. The
interface between steel and slag fluctuated constantly during smelting. Therefore, the
Figure 14. Comparison CaO-MgO-Al
low-melting-point of interfacial tension between low-melting-point CaO-MgO-Al2O3 inclusions
2 O3 inclusions had the risk of being involved in liquid
and MgO-Al 2O3 inclusions.
steel, which would pollute the liquid steel.

Metals 2022, 12, 1505
cause of the low interfacial tension between the low-melting-point CaO-MgO-Al2O3 inclu-
sions and the liquid steel, such inclusions were not removed but remained at the interface.
The interface between steel and slag fluctuated constantly during smelting. Therefore, the
low-melting-point CaO-MgO-Al2O3 inclusions had the risk of being involved in liquid
steel, which would pollute the liquid steel. 15 of 17

Slag

CaO of slag
Liquid [Ca]
[Al] + [O]
Steel [Ca] [Mg]
CMA

MgO-Al2O3 MgO-Al2O3

Refractory Al2O3
[Mg] Low Melting
MgO Low Mel ting
Al2O3 Point CMA
Point CMA

Steel and Slag

EAF LF VD Interface(After the VD
vacuum treatment)

Figure
Figure15.
15.Schematic
Schematic of inclusionformation
of inclusion formation and
and removal
removal mechanism.
mechanism.

5. 5.Conclusions
Conclusions
In order to investigate the formation and removal mechanism of oxide inclusion in
In order to investigate the formation and removal mechanism of oxide inclusion in
42CrMo4 steel, molten steel specimens were taken at different stages in industrial trials.
42CrMo4
According steel, molten
to the steel
analysis specimensinwere
of inclusions taken
samples, at different
combined with stages in industrial
thermodynamic and trials.
According to the analysis of inclusions in samples, combined
kinetic calculation, the following conclusions were obtained: with thermodynamic and
kinetic calculation, the following conclusions were obtained:
1. The MgO-Al2 O3 inclusions were the main inclusions generated in aluminum-deoxidized
1. The MgO-Al
liquid 2O
steel in 3 inclusions
the LF. With thewere the main
progress inclusions
of smelting, generated began
CaO components in aluminum-deoxi-
to appear
dized liquid2 O
in MgO-Al steel in the LF.After
3 inclusions. With thethe
VD, progress
main type of smelting,
of inclusionsCaO was components
CaO- MgO-began
toAl 2 O3 inclusions,
appear in MgO-Al and2Othe proportionAfter
3 inclusions. of CaOVD, components
the main type in inclusions increased
of inclusions was CaO-
significantly. Combined with the phase diagram, it was found
MgO-Al2O3 inclusions, and the proportion of CaO components in inclusions in- that all the inclusions
were located
creased in the liquid
significantly. phase area
Combined andthe
with belonged
phasetodiagram,
low-melting-point
it was found CaO-MgO-
that all the
Al2 O3 inclusions.
inclusions were located in the liquid phase area and belonged to low-melting-point
2. The existence of [Mg] in molten steel was the fundamental reason for the formation of
CaO-MgO-Al 2O3 inclusions.
MgO-Al2 O3 inclusions. The thermodynamic calculation of the [Mg] transfer capacity
2. The existence
of slag of [Mg]toin
and refractory molten
liquid steelsteel was that
indicated the [Mg]
fundamental reason
in liquid steel at LFfor
wasthe formation
mainly
ofprovided
MgO-Al2by O3refractory,
inclusions.whereas
The thermodynamic
[Mg] in liquid steel at VD was mainly providedcapac-
calculation of the [Mg] transfer
itybyofslag.
slag and refractory to liquid steel indicated that [Mg] in liquid steel at LF was
3. mainly
Thermodynamic
provided calculation
by refractory, indicated
whereasthat[Mg]
the mass fraction
in liquid of [Ca]
steel at VDin molten steel pro-
was mainly
increased to 3 × 10 −4 %, and the dominant area of liquid oxide in the Mg-Al-O phase
vided by slag.
diagram would becalculation
3. Thermodynamic significantlyindicated
expanded,thatwhich
themade
massa fraction
large number of CaO-MgO-
of [Ca] in molten steel
Al 2 O 3 inclusions form in liquid steel. During this period, the
increased to 3 × 10 %, and the dominant area of liquid oxide in the Mg-Al-O phase
−4 theoretical transition
route of MgO-Al O3 inclusions was as follows: MgO-Al2 O3 → CaO-2MgO-8Al2 O3
diagram would be2 significantly expanded, which made a large number of CaO-MgO-
→ CaO-2MgO-8Al2 O3 + CaO-2Al2 O3 → Liquid CaO-MgO-Al2 O3 , and the content of
AlCaO
2O3 inclusions form in liquid steel. During this period, the theoretical transition
in the inclusions increased gradually.
4. route of MgO-Al
Kinetic calculation2O3 inclusions was as follows: MgO-Al2O3 → CaO-2MgO-8Al2O3 →
indicates that MgO-Al2 O3 inclusions with a radius of 15 µm can
be removed at 5.04 × 10−6 s under the same steel and slag conditions, whereas the
low-melting-point CaO-MgO-Al2 O3 inclusions with a radius of 15 µm oscillate at the
interface of steel and slag, and finally stay at the interface, which is in accordance
with the characteristics of inclusions in industrial trials. The low interfacial tension
between the low-melting-point CaO-MgO-Al2 O3 inclusions and the liquid steel is
only 1.34 N·m−1 , but the interfacial tension of MgO-Al2 O3 inclusions is 3.13 N·m−1 ,
which causes the low-melting-point CaO-MgO-Al2 O3 inclusions not to be removed.

Author Contributions: Conceptualization, T.Q.; methodology, T.Q. and Y.H.; software, T.Q. and Y.L.;
validation, T.Q., Y.H. and Z.L.; formal analysis, T.Q. and G.C.; investigation, T.Q.; supervision, T.Q.;
project administration, G.C. and Y.Z. All authors have read and agreed to the published version of
the manuscript.
Funding: This research was funded by the National Natural Science Foundation of China, grant
numbers U196021 and 51874034.
Metals 2022, 12, 1505 16 of 17

Institutional Review Board Statement: Not applicable.

Informed Consent Statement: Not applicable.
Data Availability Statement: Not applicable.
Acknowledgments: The authors would like to express appreciation to the National Natural Science
Foundation of China for the financial support, the State Key Laboratory of Advanced Metallurgy
at the University of Science and Technology Beijing, and to Jiangsu Yonggang Group Co., Ltd., for
academic and experimental assistance.
Conflicts of Interest: The authors declare no conflict of interest.

Nomenclature
Fa The combined force on the inclusions (N)
Fb The buoyancy force (N)
Fd The drag force (N)
Ff The added mass force (N)
Fr The rebound force (N)
RI The radius of inclusions (µm)
σMS The interfacial tension between molten steel and slag (N/m)
σIM The interfacial tension between inclusions and liquid steel (N/m)
σIS The interfacial tension between inclusions and slag (N/m)
µM The viscosity of liquid steel (Pa•s)
µS The viscosity of slag (Pa•s)
ρM The density of liquid steel (kg/m3 )
ρS The slag density (kg/m3 )
ρI The density of inclusions (kg/m3 )
Z The displacement of inclusions at the interface between steel and slag (µm)
t The time (s)
References
1. He, P.; Hong, R.; Wang, H.; Lu, C. Fatigue life analysis of slewing bearings in wind turbines. Int. J. Fatigue 2018, 111, 233–242.
[CrossRef]
2. Seleznev, M.; Wong, K.Y.; Stoyan, D.; Weidner, A.; Biermann, H. Cluster detection of non-metallic inclusions in 42CrMo4 steel.
Steel Res. Int. 2018, 89, 1800216. [CrossRef]
3. Zhang, H.; Zhu, M.; Ji, S.; Zhang, J.; Fan, H. Numerical simulation and experimental study on laser hardening process of the
42CrMo4 Steel. Adv. Mech. Eng. 2021, 13, 16878140211044648. [CrossRef]
4. Escalero, M.; Blasón, S.; Zabala, H.; Torca, I.; Urresti, I.; Muniz-Calvente, M.; Fernández-Canteli, A. Study of alternatives and
experimental validation for predictions of hole-edge fatigue crack growth in 42CrMo4 steel. Eng. Struct. 2018, 176, 621–631.
[CrossRef]
5. Cubillas, D.; Olave, M.; Llavori, I.; Ulacia, I.; Larrañaga, J.; Zurutuza, A.; Lopez, A. Numerical analysis of the wind turbine pitch
bearing raceway tribo-contact due to cyclic loading under constant pitch angle. In Proceedings of the 8th International Conference on
Fracture, Fatigue and Wear; Abdel Wahab, M., Ed.; Springer: Singapore, 2021; pp. 757–769.
6. Yang, Z.G.; Li, S.X.; Zhang, J.M.; Zhang, J.F.; Li, G.Y.; Li, Z.B.; Hui, W.J.; Weng, Y.Q. The fatigue behaviors of zero-inclusion and
commercial 42CrMo steels in the super-long fatigue life regime. Acta Mater. 2004, 52, 5235–5241. [CrossRef]
7. Zhang, J.; Li, S.; Yang, Z.; Li, G.; Hui, W.; Weng, Y. Influence of inclusion size on fatigue behavior of high strength steels in the
gigacycle fatigue regime. Int. J. Fatigue 2007, 29, 765–771. [CrossRef]
8. Tsunekage, N.; Hashimoto, K.; Fujimatsu, T.; Hiraoka, K.; Beswick, J.; Dean, S.W. Initiation behavior of crack originated from
non-metallic inclusion in rolling contact fatigue. J. ASTM Int. 2010, 7, 97–110. [CrossRef]
9. Liu, C.; Gao, X.; Ueda, S.; Guo, M.; Kitamura, S. Composition changes of inclusions by reaction with slag and refractory: A review.
ISIJ Int. 2020, 60, 1835–1848. [CrossRef]
10. Park, J.H.; Todoroki, H. Control of MgO·Al2 O3 spinel inclusions in stainless steels. ISIJ Int. 2010, 50, 1333–1346. [CrossRef]
11. Kumar, D.; Pistorius, P.C. A study on calcium transfer from slag to steel and its effect on modification of alumina and spinel
inclusions. In Advances in Molten Slags, Fluxes, and Salts: Proceedings of the 10th International Conference on Molten Slags, Fluxes and
Salts 2016; Reddy, R.G., Chaubal, P., Pistorius, P.C., Pal, U., Eds.; Springer International Publishing: Cham, Switzerland, 2016;
pp. 145–153.
12. Yang, S.; Wang, Q.; Zhang, L.; Li, J.; Peaslee, K. Formation and modification of MgO·Al2 O3 -based inclusions in alloy steels. Metall.
Mater. Trans. B 2012, 43, 731–750. [CrossRef]
13. Yin, X.; Sun, Y.H.; Yang, Y.D.; Bai, X.F.; Deng, X.X.; Barati, M.; McLean, A. Inclusion Evolution during refining and continuous
casting of 316L stainless steel. Ironmak. Steelmak. 2016, 43, 533–540. [CrossRef]
Metals 2022, 12, 1505 17 of 17

14. Park, J.H.; Kang, Y. Inclusions in stainless steels—A review. Steel Res. Int. 2017, 88, 1700130. [CrossRef]
15. Verma, N.; Pistorius, P.C.; Fruehan, R.J.; Potter, M.S.; Oltmann, H.G.; Pretorius, E.B. Calcium modification of spinel inclusions in
aluminum-killed steel: Reaction steps. Metall. Mater. Trans. B 2012, 43, 830–840. [CrossRef]
16. Okuyama, G.; Yamaguchi, K.; Takeuchi, S.; Sorimachi, K. Effect of slag composition on the kinetics of formation of Al2 O3 -MgO
inclusions in aluminum killed ferritic stainless steel. ISIJ Int. 2000, 40, 121–128. [CrossRef]
17. Zhang, L. Several important scientific research points of non-metallic inclusions in steel. Steelmaking 2016, 32, 1–16.
18. Alhussein, A.; Yang, W.; Zhang, L. Effect of interactions between Fe–Al alloy and MgO-based refractory on the generation of
MgO·Al2 O3 spinel. Ironmak. Steelmak. 2020, 47, 424–431. [CrossRef]
19. Huang, F.; Zhang, L.; Zhang, Y.; Ren, Y. Kinetic modeling for the dissolution of MgO lining refractory in Al-killed steels. Metall.
Mater. Trans. B 2017, 48, 2195–2206. [CrossRef]
20. Wang, X.; Li, X.; Li, Q.; Huang, F.; Li, H.; Yang, J. Control of stringer shaped non-metallic inclusions of CaO-Al2 O3 System in API
X80 linepipe steel plates. Steel Res. Int. 2014, 85, 155–163. [CrossRef]
21. Deng, Z.; Zhu, M. Evolution mechanism of non-metallic inclusions in Al-killed alloyed steel during secondary refining process.
ISIJ Int. 2013, 53, 450–458. [CrossRef]
22. Jing, G.; Shu-Sen, C.; Zi-Jian, C. Mechanism of non-metallic inclusion formation and modification and their deformation during
compact strip production (CSP) process for aluminum-killed steel. ISIJ Int. 2013, 53, 2142–2151. [CrossRef]
23. Jiang, M.; Wang, X.; Chen, B.; Wang, W. Laboratory study on evolution mechanisms of non-metallic inclusions in high strength
alloyed steel refined by high basicity slag. ISIJ Int. 2010, 50, 95–104. [CrossRef]
24. Liu, C.; Gao, X.; Ueda, S.; Kitamura, S. Change in composition of inclusions through the reaction between Al-killed steel and the
slag of CaO and MgO saturation. ISIJ Int. 2019, 59, 268–276. [CrossRef]
25. Strandh, J.; Nakajima, K.; Eriksson, R.; Jönsson, P. Solid inclusion transfer at a steel-slag interface with focus on tundish conditions.
ISIJ Int. 2005, 45, 1597–1606. [CrossRef]
26. Strandh, J.; Nakajima, K.; Eriksson, R.; Jönsson, P. A mathematical model to study liquid inclusion behavior at the steel-slag
interface. ISIJ Int. 2005, 45, 1838–1847. [CrossRef]
27. Shannon, G.N.; Sridhar, S. Modeling Al2 O3 inclusion separation across steel-slag interfaces. Scand. J. Metall. 2005, 34, 353–362.
[CrossRef]
28. Xuan, C.; Persson, E.S.; Sevastopolev, R.; Nzotta, M. Motion and detachment behaviors of liquid inclusion at molten steel–slag
interfaces. Metall. Mater. Trans. B 2019, 50, 1957–1973. [CrossRef]
29. Liu, C.; Yang, S.; Li, J.; Zhu, L.; Li, X. Motion behavior of nonmetallic inclusions at the interface of steel and slag. Part I: Model
development, validation, and preliminary analysis. Metall. Mater. Trans. B 2016, 47, 1882–1892. [CrossRef]
30. Liu, W.; Yang, S.; Li, J.; Wang, F.; Yang, H. Numerical model of inclusion separation from liquid metal with consideration of
dissolution in slag. J. Iron Steel Res. Int. 2019, 26, 1147–1153. [CrossRef]
31. Lee, S.H.; Tse, C.; Yi, K.W.; Misra, P.; Chevrier, V.; Orrling, C.; Sridhar, S.; Cramb, A.W. Separation and dissolution of Al2 O3
inclusions at slag/metal interfaces. J. Non-Cryst. Solids 2001, 282, 41–48. [CrossRef]
32. Liu, C.; Jia, Y.; Hao, L.; Han, S.; Huang, F.; Yu, H.; Gao, X.; Ueda, S.; Kitamura, S. Effects of slag composition and impurities of
alloys on the inclusion transformation during industrial ladle furnace refining. Metals 2021, 11, 763. [CrossRef]
33. Liu, L.; Fan, J.; Wang, P.; Wang, L. Generation mechanism of large inclusions during bearing steels refining process by tracer
method. Iron Steel 2017, 52, 34–41.
34. Xu, J.; Wang, K.; Wang, Y.; Qu, Z.; Tu, X.; Meng, X. Influence mechanism of silicon content in al-killed steel on compositions of
inclusions during LF refining. Ironmak. Steelmak. 2021, 48, 127–132. [CrossRef]
35. Kang, D.; Yu, Y. Determination of Content of Trace Magnesium and Calcium in Iron and Steel by Inductively Coupled Plasma
Mass Spectrometry. Angang Technol. 2013, 5, 37–40.
36. Howie, R.A.; Zussman, J.; Deer, W. An Introduction to the Rock-Forming Minerals; Longman: London, UK, 1992.