International Journal of Minerals, Metallurgy and Materials

Volume 17, Number 1, February 2010, Page 28

DOI: 10.1007/s12613-010-0105-2

Determination of inclusions in liquid steel after calcium treatment

Kabamba Tshilombo
Department of Material Science and Manufacturing, Council for Scientific and Industrial Research, Pretoria 0001, South Africa
(Received: 17 February 2009; revised: 20 March 2009; accepted: 25 March 2009)

Abstract: Nozzle blocking was eliminated by calcium-treated liquid steel through changing the chemical and phase composition of alumina
inclusions in aluminium-killed steel. Three different methods were applied to determine the composition of inclusions in liquid steel: total
oxygen content samples, sampling spoon samples, and “lollipop” steel samples. The results show that calcium modification of liquid steel in-
fluences the inclusion composition varying from 2wt% to 14wt% depending on the method used. The composition of inclusions contains
mainly Al2O3-CaO only, or is associated with SiO2 or MgO depending on the initial input. The methods used in this study are indicators of
the inclusion composition but can be improved to quantify the inclusion size.

Keywords: inclusion; modification; castability; deoxidation

1. Introduction Total oxygen content is used as a criterion of steel

cleanliness in the steel making process. The equilibrium
Calcium treatment is applied after steel refining to elimi- between aluminium and oxygen during deoxidation is rep-
nate nozzle blocking by changing the chemical and phase resented by the following equation:
composition of alumina inclusions in aluminium-killed steel
[1-3]. This treatment is based on calcium having a consid-
erably higher affinity for oxygen than aluminium and form-
( )
lg K = lg w[2Al] .w[3O] = −
62780
T
+ 20.54 (1)

ing a low-melting point mixture with aluminium oxide. where K is the equilibrium constant of the reaction, w[Al] the
Successful calcium treatment relies on the following: injec- aluminium content in steel, w[O] the oxygen content in steel,
tion of the calcium by wire feeding, use of a sufficient T the temperature.
amount of calcium, and avoiding the reaction of calcium
According to Eq. (1), if w[Al]=0.03wt%-0.06wt%, the free
with dissolved sulphur rather than with alumina inclusions
oxygen content is 3×10−6-5×10−6 at 1600°C. Because of the
[1]. If the calcium content in liquid steel is not rational,
small number of large inclusions in the steel and the small
however, the effect of calcium treatment is negative, the
sample size for total oxygen content measurement (normally
castability and fluidity of liquid steel rapidly deteriorates [3].
20 g), total oxygen content really represents the level of
From a detailed analysis of the link between inclusion com-
small oxide inclusions [4].
position and casting behaviour, it appears that clogging is
avoided if the inclusions contain more than 50wt% liquid The sampling spoon shown in Fig. 1 is a steel cup with a
phases in the liquid steel [1]. wooden lid which has a steel handle used to manipulate the
sampler. Before sampling, the wooden lid is fastened to the
The aim of the present work is to evaluate the sizes and cup using copper wire. The cup is then submerged in the
compositions of inclusions before and after calcium treat- steel, below the slag layer. After a few seconds of melting,
ment at the Mittal Saldanha Plant. Three different methods the copper wire releases the wooden lid, allowing liquid
were used: total oxygen content samples, sampling spoon steel to enter the cup. Once the sampling spoon is removed,
samples, and “lollipop” steel samples. it is cooled and the sample is subsequently taken out. The

Corresponding author: Kabamba Tshilombo E-mail: [email protected]

© University of Science and Technology Beijing and Springer-Verlag Berlin Heidelberg 2010
K. Tshilombo, Determination of inclusions in liquid steel after calcium treatment 29

sample from the spoon does not contain slag and the sam- method used, and the existing states of these inclusions in
pler does not filter out inclusions from the steel [5]. the liquid steel were ascertained according to the CaO-Al2O3
binary, CaO-MgO-Al2O3, and CaO-Al2O3-SiO2 ternary dia-
grams.
3. Results and discussion
From Figs. 2 and 3, the majority of the inclusions are
alumina with a low amount of sulphur. The calcium modi-
fication of inclusions (50wt% liquid) determined by
SEM-EDX analysis varies from 4wt% to 5wt% for total
oxygen content samples, 2wt% to 10wt% for sampling
spoon samples, and 2wt% to 14wt% for “lollipop” steel
samples. This indicates a partial and full liquefaction of the
oxide inclusions.
Fig. 1. Schematic drawing of the sampler used to collect liq-
uid steel samples (unit: mm).

2. Materials and experimental procedure

Steel samples were obtained from an integrated steel-
works which produced aluminium-killed low-carbon
(0.05wt% C) steel. The plant used two ladle furnaces as
holding vessels before continuous casting, with the holding
time ranging from 45 to 401 min. The total content of FeO
and MnO in the ladle slag was below 1wt% in all cases after
LF refining. The ladle refractory was MgO-carbon bricks at
the slagline and in the intermediate zone and alumina-spinel
carbon bricks in the barrel and bottom. Calcium treatment
was applied just before sending the ladle to the continuous
caster, and calcium wire was used for treatment.
The samples were obtained from different heats with the
steel mass ranging from 170 to 180 t. Steel temperatures af-
ter calcium treatment varied from 1591°C to 1604°C with
tundish temperatures from 1561°C to 1569°C. The calcium
content of the steel was from 16×10−6 to 34×10−6 in the ladle
after calcium treatment, and from 16×10−6 to 19×10−6 in the
tundish. The aluminium content was in the range of
0.01wt%-0.03wt%, and the sulphur content was in the range
of 30×10−6-50×10−6. “Lollipop” steel samples were taken
before and after calcium treatment in the ladle and in the
tundish [1].
All the samples were cut and polished for scanning elec-
tron microscopy (SEM) using backscattered electron imag-
ing to identify the inclusions. Inclusion compositions were
analysed by energy-dispersive X-ray microanalysis (EDX),
generally using 100 s of live time per inclusion. The
amounts of calcium, aluminium, manganese, and sulphur
Fig. 2. EDX of the total oxygen samples (a), “Lollipop” steel
were quantified. At least 20 inclusions were analysed per samples (b), and sampling spoon samples (c) in the tundish,
sample. The inclusions were classified according to the respectively.
30 Int. J. Miner. Metall. Mater., Vol.17, No.1, Feb 2010

Fig. 3. Backscattered electron images of inclusions after calcium treatment: (a) total oxygen samples; (b) “Lollipop” steel samples;
(c) sampling spoon samples.

The inclusions in the calcium treated steel can be classi- with MgO in the top slag [3].
fied according to their composition as: (1) Al2O3-CaO bi- In Fig. 4, different symbols are for the samples from dif-
nary inclusions formed by the combination of alumina with ferent heats, where A, C, and M indicate Al2O3, CaO, and
calcium oxide after calcium addition; (2) Al2O3-SiO2-CaO MgO, respectively. Calcium treatment changes the inclusion
ternary inclusions formed by the combination of calcium compositions considerably, which is away from MgO-Al2O3
aluminate with silicate; (3) Al2O3-CaO-MgO ternary inclu- spinels, and shifts to higher CaO contents and lower MgO
sions formed by the combination of deoxidation product contents [1].

Fig. 4. Compositions independent of the method used.

Fuhr and Cicutti studied the relationship between the in- increment of CaO could be caused by the strong interactions
clusion composition and nozzle blocking, and found that the with the slag because of vigorous argon bubbling performed
K. Tshilombo, Determination of inclusions in liquid steel after calcium treatment 31

to promote steel desulphurization [6]. remain almost unchanged during the process, being mostly
liquid at the steelmaking temperature. This ensures good
Fig. 5 shows that the CaS content of the inclusions in-
castability and absence of deposits in the nozzle wall. The
creases at higher Ca/O mass ratios once the inclusions have
castability problems are mainly attributed to an insufficient
been modified to the liquid form. Geldenhuis and Pistorius
modification of alumina inclusions.
studied the minimisation of calcium additions in steel, and
found that the CaS contents were generally higher than the Acknowledgements
equilibrium prediction, presumably reflecting a kinetic trend
[7]. The author is grateful to ArcelorMittal Saldanha, the
However, calcium treatment is not efficient for high sul- University of Pretoria for support of this work. In particular,
phur steels as calcium reacts with sulphur to form solid cal- the author thanks Roelof Hatting of ArcelorMittal Saldanha
cium sulphide instead of liquefying the alumina [5]. for providing samples, Prof. Chris Pistorius of the Univer-
sity of Pretoria for his helpful discussions on inclusions in
liquid steel.

References
[1] P. Chris, Magnesium: origin and role in calcium-treated inclu-
sions, [in] 2006 TMS Fall Extraction and Processing Division:
Sohn International Symposium, California, 2006, p.373.
[2] Z.J. Han, L. Liu, M. Lind, and L. Holappa, Mechanism and
kinetics of transformation of alumina inclusions by calcium
treatment, Acta Metall. Sin. Engl. Lett., 19(2006), p.1.
[3] F. Yuan, X. Wang, and X. Yang, Influence of calcium content
Fig. 5. Relationship between the CaS content and the Ca/O
on solid ratio of inclusions in Ca-treated liquid steel, J. Univ.
mass ratio of inclusions.
Sci. Technol. Beijing, 13(2006), p.486.
[4] L.F. Zhang, Indirect methods of detecting and evaluating in-
4. Conclusion clusions in steel-a review, J. Iron Steel Res. Int., 13(2006), p.1.
[5] K.G. Rackers and B.G. Thomas, Clogging in continuous
Within the investigated temperature range varying from
casting nozzles, [in] 78th Steelmaking conference proceedings,
1560°C to 1600°C, the microinclusions observed in the ladle Warrendale, 1995, p.723.
slag FeO are mainly composed by alumina before calcium [6] E. Fuhr, C. Cicutti, G. Walter, and G. Torga, Relationship
addition. After the treatment, the microinclusions are con- between nozzle deposits and inclusion composition in the con-
verted into calcium aluminates with small quantities of MgO tinuous casting of steels, Iron Steelmaker, 30(2003), p.53.
or SiO2. Inclusions from calcium modification of liquid steel [7] J.M.A. Geldenhuis and P.C. Pistorius, Minimisation of cal-
vary from 2wt% to 14wt% depending on the method used. cium additions to low carbon steel grades, Ironmaking Steel-
Under soft bubbling, the compositions of microinclusions making, 27(2000), p.442.