Formation Mechanism of CaS-Al2O3 Inclusions in

Low Sulfur Al-Killed Steel After Calcium Treatment

JIANFEI XU, FUXIANG HUANG, and XINHUA WANG

The laboratory experiments of alumina inclusions modified by calcium treatment in Al-killed

steel were carried out at 1873 K (1600 °C), and the inclusions in steel samples were characterized
at 1, 5, and 10 minutes after calcium addition. The results show that the type of inclusions after
calcium treatment was determined by the sulfur and T.O contents of steel. CaS-Al2O3 inclusions
were obtained in steels with high sulfur and low T.O contents. The mass ratio between CaS and
Al2O3 was determined by T.Ca and T.O contents of steel. The influence of holding time after
calcium addition on the composition of inclusions was negligible. The thermodynamics for the
formation of CaS-Al2O3 inclusions after calcium treatment was discussed, and a simple
formation mechanism was proposed. Moreover, the CaO, Al2O3, and CaS contents in the
inclusions were predicted through the sulfur, total calcium (T.Ca), and T.O contents, and it was
found that the CaO content decreases with increasing S/T.O, while (pctCaS)/(pctAl2O3)1/3
increases with increasing T.Ca/T.O.

DOI: 10.1007/s11663-016-0599-8
Ó The Minerals, Metals & Materials Society and ASM International 2016

I. INTRODUCTION composed of the high melting point CaO-CaS system.

However, CaO-CaS inclusions were obtained in steel
CALCIUM is added to Al-killed steel for modifying with a high T.Ca content,[12,13] which increases the cost
solid alumina inclusions. Through proper calcium treat- of production. Meanwhile, it is found from the
ment, Al2O3 can be converted into low melting point CaO-Al2O3-CaS ternary phase diagram that CaS-Al2O3
calcium aluminates.[1–5] There are two methods for also has a high melting point and a poor deformability
predicting the composition of inclusions after calcium compared with CaO-CaS, and a new strategy for
treatment: One involves monitoring of the total calcium controlling the formation of CaS-Al2O3 inclusions in
(T.Ca) and acid-soluble aluminum ([Al]s) contents of the rolled steel plates was adopted. However, the formation
steel,[6,7] e.g., the alumina inclusions are converted into of CaS-Al2O3 inclusions after calcium treatment has
liquid calcium aluminates when T.Ca/[Al]s is larger than seldom been reported. Verma and Ren et al.[14–17] found
0.09. The other involves monitoring of the T.Ca and that CaS-Al2O3 inclusions were formed initially
total oxygen (T.O) contents of the steel,[8,9] and a T.Ca/ after calcium addition, and then, they converted to
T.O of 0.91 to 1.25 has been proposed for ensuring good calcium aluminates. Based on this, they proposed that
modification results. CaS-Al2O3 inclusions are the intermediate products in
However, calcium aluminate inclusions adversely the formation of calcium aluminates via calcium addi-
affect steel properties in the case of some steel grades, tion. However, the influence of T.O content on the
e.g., API X80 pipeline steel. This is because they can modification of Al2O3 inclusions by calcium treatment
aggregate to larger inclusions of 10 to 20 lm in was not evaluated in their experiments, and hence, the
continuous casting and can be deformed into stringer- formation of CaS-Al2O3 inclusions after calcium treat-
shaped inclusions in steel plates during rolling due to ment was not adequately clarified.
their good deformability,[10,11] which results in the In this study, the laboratory experiments of alumina
amount of B-type inclusions in rolled steel plates being inclusions modified by calcium treatment in Al-killed
higher than the standard value. Thus, inclusions with steel were performed and the characteristics of inclu-
the high melting point and poor deformability are sions in steel samples after calcium treatment were
favorable in these rolled steel plates, and some stud- investigated. Moreover, the formation mechanism of
ies[10,11] found that the severities of B-type inclusions can CaS-Al2O3 inclusions via calcium treatment was dis-
be lowered from ‘‘2.0’’ to ‘‘0’’ (ASTM E45 standard) cussed, and the relationship between steel compositions
when the inclusions found in the steel plates were all and inclusion compositions was analyzed.

JIANFEI XU, Ph.D. Student, FUXIANG HUANG, Doctor,

Lecturer, and XINHUA WANG, Doctor, Professor, are with the II. EXPERIMENTAL
School of Metallurgical and Ecological Engineering, University of
Science and Technology Beijing, 30 Xueyuan Road, Haidian District, A. Experimental Procedure
Beijing 100083, P.R. China. Contact e-mail: [email protected]
Manuscript submitted on November 19, 2015. The 16 experiments for the modification of alumina
Article published online February 8, 2016. inclusions by calcium treatment were performed in a

METALLURGICAL AND MATERIALS TRANSACTIONS B VOLUME 47B, APRIL 2016—1217

vacuum induction furnace, and the experimental setup is CaSi addition. After 5 minutes, the rod was withdrawn
shown in Figure 1(a). and the interface of reaction between the rod and molten
For each experiment, the Al lump was placed in the steel was observed by SEM/EDS.
feed opening. The CaSi alloy wrapped with steel was
fixed in a molybdenum rod. The chamber was first
B. Chemical Analysis and Characterization of Inclusions
evacuated to 20 Pa with a mechanical pump and then
backfilled with high-purity argon (99.999 pct) up to 1 Figure 2 shows the position of steel samples to
atm. The steel (1000 g) was melted in a magnesia analyze steel compositions (T.O, S, [Al]s and T.Ca)
crucible at 1873 K (1600 °C) under a 4 L/min flow of and inclusion compositions. The T.O and sulfur con-
high-purity argon. Then, a 0.3 g lump of aluminum was tents of steel were analyzed by infrared absorption. The
added into the molten steel to maintain a relatively T.Ca and [Al]s contents of steel were analyzed by
high-acid-soluble aluminum content. After 3 minutes, inductively coupled plasma-atomic emission spectrom-
pure CaSi alloy was added to the bottom of the molten etry. The inclusion compositions were analyzed by
steel by inserting the molybdenum rod. After a holding ASPEX PSEM EXPLORER with an automated feature
time of 1, 5, and 10 minutes, the melts were poured into analysis system.
an iron mold (bottom inner diameter, top inner diam-
eter, and height of 40, 55, and 140 mm, respectively).
The chemical composition of the steel used in these III. RESULTS
experiments is given in Table I. The original inclusions
of the steel were composed of Al2O3. The holding time A. Compositions of Steel Under Different Treatment
and the amounts of CaSi alloy for each experiment are Conditions
listed in Table II. The compositions of steel after calcium treatment
Moreover, using the same vacuum induction furnace, under different conditions are shown in Table III. The S,
immersion experiments were performed to analyze the T.Ca, and T.O contents fluctuate in the range of 8 to 28,
influence of T.O content on the extent of reaction 9 to 18, and 6 to 24 ppm, respectively, for different
between CaO in the inclusions and S in steel, and clarify holding time.
the formation mechanism of CaS-Al2O3 inclusions after
calcium treatment. The experimental setup is shown in
Figure 1(b). B. Characterization of Inclusions
A 10-mm-diameter MgO rod with some CaO impu- Figure 3 shows the composition of the inclusions at 1
rities was immersed for 5 minutes in molten steel minute after calcium addition. MgO was not taken into
prepared by the procedure described earlier, but without consideration as the MgO content of the inclusions was

Viewing Sampling box Feed Viewing Sampling box Feed

window opening window opening

Gas Gas
outlet outlet
Magnesia Magnesia
crucible crucible A 10-mm-diameter MgO rod with some CaO impurities

Argon gas Argon gas

Melt Melt

Magnesia lining Induction coil Magnesia lining Induction coil

(a) (b)

Fig. 1—Schematic diagram of the experimental setup.

Table I. Chemical Composition of the Steel Used in the Experiments (Weight Percent)

C Si Mn P S [Al]s T.Ca T.O

0.05 0.02 0.25 0.008 0.0008 to 0.0028 0.03 <0.0005 0.0015

1218—VOLUME 47B, APRIL 2016 METALLURGICAL AND MATERIALS TRANSACTIONS B

small (<5 pct). It was found that the Al2O3 inclusions from CaO-Al2O3-CaS to CaS-Al2O3 with increasing the
were converted into CaO-Al2O3-CaS inclusions when sulfur content of steel. However, in contrast to the
the sulfur content of steel was less than 12 ppm. results of 1 minute after calcium addition, CaO-Al2O3-
However, CaS-Al2O3 inclusions were the main product CaS inclusions were still obtained for steel with a sulfur
when the sulfur content was higher than 14 ppm (here content of 14 to 20 ppm.
the inclusions were defined as CaS-Al2O3 inclusions if The composition of the inclusions at 10 minute
the CaS + Al2O3 content in the inclusions was above after calcium addition is shown in Figure 5. The
90 pct), which indicates that the sulfur content of steel modification results were similar to those at 1 and
has an impact on the modification of alumina inclusions 5 minute after calcium addition, but the critical sulfur
by calcium treatment. content, at which the product changes from CaO-
Figure 4 shows the composition of the inclusions at Al2O3-CaS to CaS-Al2O3, still differed, which implies
5 minute after calcium addition. The product changed that the modification of alumina inclusions by cal-
cium treatment is probably affected by other factors
as well.
Figure 6 shows the influence of T.Ca and T.O
contents of steel on the type of inclusion. The relation-
ship between the type of inclusion and T.Ca content was
not obvious. However, the T.O content was significantly
lower in steel samples with CaS-Al2O3 inclusions com-
pared to those with CaO-Al2O3-CaS inclusions, which
means that the formation of CaS-Al2O3 and CaO-
Al2O3-CaS inclusions after calcium treatment also
depends on the T.O content of steel; moreover, this
explains why CaS-Al2O3 inclusions cannot be obtained
for steel with a sulfur content of 14 to 20 ppm at
Fig. 2—Analytical position of steel samples. 5 minutes after calcium addition.

Table II. The Holding Time and the Amounts of CaSi Alloy for Each Experiment

No. Holding Time Amounts of CaSi Alloy (g) No. Holding Time Amounts of CaSi Alloy (g)
1 1 min 0.79 9 5 min 1.5
2 1 min 0.7 10 5 min 2
3 1 min 0.85 11 5 min 2
4 1 min 0.91 12 5 min 2.5
5 1 min 0.84 13 10 min 2.8
6 1 min 0.7 14 10 min 3
7 5 min 1.8 15 10 min 3
8 5 min 2 16 10 min 3.2

Table III. Compositions of Steel Under Different Treatment Conditions, ppm

No. Holding Time [Al]s S T.Ca T.O

1 1 min 350 8 13 15.2
2 1 min 340 12 9 14.5
3 1 min 210 14 11 8.8
4 1 min 350 15 16 10.8
5 1 min 370 20 15 8.1
6 1 min 190 22 9 11
7 5 min 180 9 13 16.2
8 5 min 150 14 17 24.1
9 5 min 300 19 11 18.6
10 5 min 260 20 10 12.9
11 5 min 520 21 15 6.3
12 5 min 280 28 18 10.9
13 10 min 300 9 12 13.8
14 10 min 300 12 14 12.3
15 10 min 370 18 15 8.7
16 10 min 360 21 18 8.2

METALLURGICAL AND MATERIALS TRANSACTIONS B VOLUME 47B, APRIL 2016—1219

Fig. 3—Composition of the inclusions at 1 min after calcium addition: (a) no. 1 and [S] = 8 ppm, (b) no. 2 and [S] = 12 ppm, (c) no. 3
and [S] = 14 ppm, (d) no. 4 and [S] = 15 ppm, (e) no. 5 and [S] = 20 ppm, and (f) no. 6 and [S] = 22 ppm.

1220—VOLUME 47B, APRIL 2016 METALLURGICAL AND MATERIALS TRANSACTIONS B

Fig. 4—Composition of the inclusions at 5 min after calcium addition: (a) no. 7 and [S] = 9 ppm, (b) no. 8 and [S] = 14 ppm, (c) no. 9
and [S] = 19 ppm, (d) no. 10 and [S] = 20 ppm, (e) no. 11 and [S] = 21 ppm, and (f) no. 12 and [S] = 28 ppm.

IV. DISCUSSION
½Ca þ 1=3ðAl2 O3 Þ ! ðCaOÞ þ 2=3½Al
A. Formation Thermodynamic of CaS-Al2O3 Inclusions ½1
DG01 ¼ 356297:7 þ 71:8 T
The Gibbs free energies of the following reactions
were considered to clarify the formation of CaS-Al2O3
inclusions after calcium treatment.[12,13]

METALLURGICAL AND MATERIALS TRANSACTIONS B VOLUME 47B, APRIL 2016—1221

Fig. 5—Composition of the inclusions at 10 min after calcium addition: (a) no. 13 and [S] = 9 ppm, (b) no. 14 and [S] = 12 ppm, (c) no. 15
and [S] = 18 ppm, and (d) no. 16 and [S] = 21 ppm.

Fig. 6—Relationship between the type of inclusion and the T.Ca and T.O contents of steel: (a) T.Ca and (b) T.O.

1222—VOLUME 47B, APRIL 2016 METALLURGICAL AND MATERIALS TRANSACTIONS B

 Table IV. Activities of CaO, Al2O3, and CaS for Steel Samples with CaO-Al2O3-CaS Inclusions

Average Composition Activities

No. Al2O3 CaS CaO aCaO aAl2O3 aCaS

1 0.563 0.151 0.286 0.11 0.253 1
2 0.584 0.248 0.168 0.087 0.312 1
7 0.45 0.193 0.357 0.301 0.08 1
8 0.515 0.104 0.381 0.263 0.097 1
9 0.626 0.124 0.25 0.087 0.312 1
10 0.556 0.253 0.191 0.087 0.312 1
13 0.374 0.162 0.464 0.651 0.02 1
14 0.529 0.152 0.319 0.169 0.165 1

Table V. First- and Second-Order Interaction Coefficients at 1873 K (1600 °C)[12,13]

eji C Si Mn P S Al Ca O
Al 0.091 0.056 0.0065 0.033 0.03 0.045 0.05 6.6
Ca 0.34 0.097 0.0156 0.1 138 0.07 0 9000
S 0.11 0.063 0.026 0.029 0.03 0.035 110 0.27
O 0.42 0.066 0.021 0.07 0.13 1.17 3600 0.17
rCa 5
O ¼ 5:7  10 ; rO;Ca
O ¼ 2:9  106 ; rO 6
Ca ¼ 3:6  10 ; rCa;O
Ca ¼ 2:9  106 .

Fig. 7—Gibbs free energies of Reactions [1] through [4] at different [O] and [Ca] contents: (a) [Ca] = 5ppm, [S] = 12 ppm, and [Al]s = 340
ppm and (b) [Ca] = 10 ppm, [S] = 12 ppm, and [Al]s = 340 ppm.

½Ca þ ½O ! ðCaOÞ For the steel samples with CaO-Al2O3-CaS inclusions,
½2 i.e., no. 1, 2, 7, 8, 9, 10, 13, and 14 steel samples, the
DG02 ¼ 645421 þ 146 T
activities of CaO, Al2O3, and CaS were calculated by
FactSage 6.4 according to the average compositions of
the inclusions and the pure solid state was regarded as
½Ca þ ½S ! ðCaSÞ the reference state. FToxide (for oxides) and FactPS (for
½3 CaS) databases were employed in FactSage 6.4 Equilib
DG03 ¼ 530900 þ 116:2 T
Module, and the Modified Quasichemical Model was
used. The results are presented in Table IV.
The activities of [Ca], [Al]s, [S], and [O] dissolved in
ðCaOÞ þ ½S ! ðCaSÞ þ ½O steel were calculated by Eq. [5] and 1 mass pct steel
½4 solution was considered as the reference state. The
DG04 ¼ 114521  29:8 T

METALLURGICAL AND MATERIALS TRANSACTIONS B VOLUME 47B, APRIL 2016—1223

first- and second-order interaction coefficients are listed Reactions [1] through [3] are complete, CaO reacts with
in Table V. [S] to generate CaS, and subsequently, the CaO content
X j X j 2
of the inclusions decreases.
lg a½i ¼ lg½ipct þ ei ½pctj þ ri ½pctj For the steel samples with CaS-Al2O3 inclusions, the
j j Gibbs free energies of Reactions [1], [2], and [3] are also
XX ½5
ðj;kÞ lower than that of Reaction [4] as the activities of CaO
þ ri ½pctj½pctk
j
and Al2O3 in the CaS-Al2O3 system are, respectively,
k
lower and higher than those in the CaO-Al2O3-CaS
Taking no.2 as an example, the changes in the Gibbs system. Thus, the dissolved calcium first reacts with
free energies for Reactions [1] through [4] at different [O] Al2O3, [S], and [O]. However, the extent of reaction
and [Ca] contents are shown in Figure 7. The dissolved between CaO in the inclusions and S in steel is thought
calcium first reacts with Al2O3, [S], and [O] to generate to be different in these two systems, because their T.O is
CaO and CaS because the Gibbs free energies of different, and CaO can react with S only at a low-dis-
Reactions [1] through [3] are lower than that of Reaction solved oxygen content.
[4], regardless of the [O] and [Ca] contents. Conse- To clarify the extent of reaction between CaO and S,
quently, the CaO and CaS contents of the inclusions the influence of T.O content on the dissolved oxygen
increase, while the Al2O3 content decreases. After content at equilibrium between steel and inclusions was
first analyzed by FactSage 6.4 using the Modified
Quasichemical Model. The results are plotted in
Figure 8.
In the above calculation, FToxide (for oxides) and
FactPS (for CaS) databases were employed in FactSage
6.4 Equilib Module. The [Ca], [Al]s, [S], [O], [Fe], and
(Al2O3) were used as the input parameters. The [Ca] and
[S] contents ranged between 10 to 20 ppm and 10 to 30
ppm, respectively. The [Al]s content was valued at 300
ppm. The dissolved oxygen ([O]) content before calcium
addition was 3 ppm based on Al-O equilibrium ther-
modynamics. The (Al2O3) content was calculated by
Eq. [6]. It can be seen from Figure 8 that the dissolved
oxygen content of steel at equilibrium between steel and
inclusions decreases with decreasing T.O content at
given contents of [Al]s, [Ca], and [S]. This is because of
the existence of interaction between steel and inclusions.
Thus, it can be concluded that a low T.O content is in
favor of the reaction between CaO and S.
Fig. 8—Influence of T.O content on the equilibrium dissolved oxy-
gen content. ðAl2 O3 Þ ¼ 102=48  ðT:O  ½OÞ ½6

Fig. 9—Images of the scanned area of typical phases of the MgO rod.

1224—VOLUME 47B, APRIL 2016 METALLURGICAL AND MATERIALS TRANSACTIONS B

Fig. 10—Images of the scanned area at the reaction interface ([S] = 11 ppm, T.O = 6.1 ppm).

Fig. 11—Images of the scanned area at the reaction interface ([S] = 11 ppm, T.O = 10 ppm).

Next, the immersion experiments of the MgO rod Therefore, the reaction of CaO and S was
with some CaO impurities were conducted to further restrained for steel with a high T.O content. In this
validate the influence of T.O on the reaction extent case, only a fraction of CaO in the inclusions could
between CaO and S. The typical phases of the MgO rod react with [S] to generate CaS, and many CaO-
before it was immersed into molten steel were MgO and Al2O3-CaS inclusions were retained in steel. How-
CaO-SiO2 phases, as shown in Figure 9. ever, the extent of reaction between CaO and S was
Figure 10 shows the images of the scanned area of the large for steel with a low T.O content, and thus
reaction interface after the MgO rod was immersed for 5 CaS-Al2O3 inclusions were formed. Moreover, the
minute into molten steel with a sulfur content of 11 ppm Gibbs free energy for the reaction between CaO and
and a T.O content of 6.1 ppm. Sulfur was detected at the S decreased with increasing sulfur content of steel,
reaction interface, indicating that the CaO-SiO2 reacted which indicates that CaS-Al2O3 inclusions could be
with S in steel. also obtained at a high sulfur content without
However, as shown in Figure 11, sulfur was not changing the T.O content.
detected at the reaction interface for the steel sample Figure 12 shows the relationship between S/T.O and
with a sulfur content of 11 ppm and a T.O content of 10 the proportion of CaS-Al2O3 inclusions in the total
ppm, and thus, a low T.O content facilitated the inclusions according to the experimental results of 16
reaction between CaO and S. steel samples; it was found that CaS-Al2O3 inclusions

METALLURGICAL AND MATERIALS TRANSACTIONS B VOLUME 47B, APRIL 2016—1225

coexisted with CaO-Al2O3-CaS inclusions in all the steel oxygen and the sulfur contents of steel, and CaO-
samples. The proportion of CaS-Al2O3 inclusions in the Al2O3-CaS inclusions are retained in the steel samples
total inclusions increased with increasing S/T.O, indi- with high T.O and low sulfur contents. In contrast, high
cating that a high transformation ratio of CaO-Al2O3- amounts of CaO-Al2O3-CaS inclusions are converted to
CaS inclusions to CaS-Al2O3 inclusions were obtained CaS-Al2O3 inclusions in the steel samples with low T.O
at a low T.O content and a high sulfur content, which is and high sulfur contents.
consistent with the above analysis.
Based on the above results, a simple mechanism for
B. Relationship Between Steel Compositions and
the formation of CaS-Al2O3 inclusions after calcium
Inclusion Compositions
treatment is proposed, as shown in Figure 13.
The formation of CaS-Al2O3 inclusions after calcium Figure 14 shows the relationship between S/T.O in
treatment is divided into two stages. In the first stage, steel and the CaO content in the inclusions. The CaO
the dissolved calcium reacts with Al2O3, [S], and [O], content decreases with increasing S/T.O, and subse-
resulting in the formation of CaO-Al2O3-CaS inclusions quently becomes zero when S/T.O exceeds 2.5, indicat-
in steel. In the second stage, CaO reacts with [S], the ing that CaS-Al2O3 inclusions can be obtained after
reaction extent of which is determined by the dissolved calcium treatment at a high S/T.O value.
Holappa et al.[18] found that the window of liquid
calcium aluminates becomes narrow at a low oxygen
content and a high sulfur content. They proposed that the
oxygen content in calcium-treated steel should be con-
trolled to 10 to 20 ppm or even higher to obtain liquid
calcium aluminates. Janke et al.[19] showed that the high
activity ratio of S and O in steel facilitated the formation
of sulfides, while high oxygen content facilitated oxide
formation, which are consistent with the present results.
Figure 15 shows the relationship between T.Ca/T.O
in steel and (pctCaS)/(pctAl2O3)1/3 in the inclusions.
(pctCaS)/(pctAl2O3)1/3 was derived from the stoichio-
metric coefficients of Reaction [7], which is obtained by
combining Reactions [1] and [4].
½Ca þ 1=3ðAl2 O3 Þ þ ½S ! ðCaSÞ þ 2=3½Al þ ½O ½7

Fig. 12—Relationship between S/T.O and the proportion of CaS- Figure shows that (pctCaS)/(pctAl2O3)1/3 in the inclu-
Al2O3 inclusions in the total inclusions. sions increases with increasing T.Ca/T.O, regardless of

Fig. 13—Formation mechanism of CaS-Al2O3 inclusions.

1226—VOLUME 47B, APRIL 2016 METALLURGICAL AND MATERIALS TRANSACTIONS B

 (1) The type of inclusions after calcium treatment is
determined by the sulfur and T.O contents of steel.
CaS-Al2O3 inclusions are obtained in steels with
high sulfur and low T.O contents, whereas CaO-
Al2O3-CaS inclusions are obtained under the
opposite condition.
(2) The thermodynamics for the formation of CaS-
Al2O3 inclusions after calcium treatment is discussed,
and a simple formation mechanism is proposed.
(3) The relationship between steel compositions and
inclusion compositions is discussed, and it is found
that the CaO content in the inclusions decreases
with increasing S/T.O, while (pctCaS)/(pctAl2O3)1/3
in the inclusions increases with increasing T.Ca/T.O,
regardless of the holding time after calcium addi-
tion.

Fig. 14—Relationship between S/T.O in steel and the CaO content

in the inclusions.

REFERENCES
1. G. Ye, P. Jönsson, and T. Lund: ISIJ Int., 1996, vol. 36, pp. S105–
S108.
2. Y.I. Ito, M. Suda, Y. Kato, H. Nakato, and K.I. Sorimachi: ISIJ
Int., 1996, vol. 36, pp. S148–50.
3. Y. Higuchi, M. Numata, S. Fukagawa, and K. Shinme: ISIJ Int.,
1996, vol. 36, pp. S151–54.
4. W. Yang, L. Zhang, X. Wang, Y. Ren, X. Liu, and Q. Shan: ISIJ
Int., 2013, vol. 53, pp. 1401–410.
5. M. Lind and L. Holappa: Metall. Mater. Trans. B, 2010, vol. 41B,
pp. 359–66.
6. S.P. He, C.R. Wang, Z.Y. Lai, Q. Wang, B.C. Huang, and J.C.
Peng: J. Univ. Sci. Technol. Beijing, 2007, vol. 29, pp. S18–S21.
7. L. Zhang, J. Zhi, F. Mei, L. Zhu, X. Jiang, J. Shen, J. Cui
K. Cai, and B.G. Thomas: Ironmak. Steelmak., 2006, vol. 33,
pp. 129–39.
8. Z. Deng and M. Zhu: Steel Research Int., 2013, vol. 6,
pp. 519–25.
9. J.M.A. Geldenhuis and P.C. Pistorius: Ironmak. Steelmak., 2000,
vol. 27, pp. 442–49.
Fig. 15—Relationship between T.Ca/T.O in steel and (pctCaS)/ 10. X. Wang, X. Li, Q. Li, F. Huang, H. Li, and J. Yang: Steel Res.
(pctAl2O3)1/3 in the inclusions. Int., 2014, vol. 85, pp. 155–63.
11. X. Wang, X. Li, Q. Li, F. Huang, H. Li, and J. Yang: Acta Metall.
Sinica, 2013, vol. 49, pp. 553–61.
the holding time after calcium addition. Therefore, it is 12. G. Yang, X. Wang, F. Huang, D. Yang, P. Wei, and X. Hao:
reasonable to infer that the type of inclusions after Metall. Mater. Trans. B, 2015, vol. 46B, pp. 145–54.
13. G. Yang and X. Wang: ISIJ Int., 2015, vol. 55, pp. 126–33.
calcium treatment is determined by the sulfur and T.O 14. N. Verma, P.C. Pistorius, R.J. Fruehan, M. Potter, M. Lind, and
contents of steel, and the mass ratio between Al2O3 and S. Story: Metall. Mater. Trans. B, 2011, vol. 42B, pp. 711–19.
CaS is determined by the T.Ca and T.O contents of steel. 15. N. Verma, P.C. Pistorius, R.J. Fruehan, M. Potter, M. Lind, and
S. Story: Metall. Mater. Trans. B, 2011, vol. 42B, pp. 720–29.
16. N. Verma, P.C. Pistorius, R.J. Fruehan, M.S. Potter, H.G.
Oltmann, and E.B. Pretorius: Metall. Mater. Trans. B, 2012,
V. CONCLUSIONS vol. 43B, pp. 830–840.
17. Y. Ren, L. Zhang, and S. Li: ISIJ Int., 2014, vol. 54, pp. 2772–
The formation mechanism of CaS-Al2O3 inclusions 2779.
after calcium treatment was investigated based on the 18. L. Holappa, M. Hämäläinen, M. Liukkonen, and M. Lind: Iron-
mak. Steelmak., 2003, vol. 30, pp. 111–115.
laboratory experiments and thermodynamic analysis. 19. D. Janke, Z. Ma, P. Valentin, and A. Heinen: ISIJ Int., 2000,
The following conclusions were obtained. vol. 40, pp. 31–39.

METALLURGICAL AND MATERIALS TRANSACTIONS B VOLUME 47B, APRIL 2016—1227