Mass transfer behaviour of Mg in low carbon aluminium killed steel during LF refining

Abstract

Inclusions containing Mg existed in low carbon aluminium killed steel even though Mg is not added during LF treatment. To investigate the mass transfer mechanism of Mg in low carbon aluminium killed steel, both industrial practice and kinetic calculations were carried out in the present work. The results from industrial practice showed that Mg concentration in molten steel and inclusions increased with refining time during ladle furnace treatment. The inclusion size tended to become smaller with the increase of Mg concentration in the inclusions. The erosion rate of refractory with different composition was tallied. A refractory-slag-metal-inclusion multiphase reaction model was developed to investigate mass transfer mechanism underlying the variation of Mg among the steel, the slag, inclusions and the refractory. The calculated results exhibited a good predictability of the content of Mg in the molten steel, slag and inclusions. The results showed that Mg dissolved into molten steel in two ways: the first is in the way of slag/steel reaction, the second is in the way of refractory erosion which is the main way.

Introduction

Ladle furnace is commonly used in high-quality steelmaking to obtain molten steel with the suitable composition and control the composition of inclusions in a low Melton zone. There are serials of complex reactions occurring in the ladle refining among the steel, slag, inclusions and refractory. Many reports showed that Mg existed in the molten and inclusions during LF refining under no Mg addition condition.^1–3 The mechanisms underlying the variation of Mg among the steel, the slag, inclusions and the refractory need to be further investigated.

Many scholars investigated the behaviour of Mg in the reaction between molten steel and slag. Harada et al. ^4,5 studied the dissolution behaviour of Mg form MgO into molten steel in laboratory. Kasimagwa et al. ⁶ studied the reactions between MgO–C refractory during second refining in laboratory. Jansson et al. ⁷ studied the effect of relative velocity, temperature and steel composition on the dissolution rate of MgO–C refractory in Al-deoxidized molten steel by rotating cylinder method in the laboratory. However, the studies on the mass transfer behaviour of Mg based on the industrial production were rarely reported. For real industrial conditions, there was much difference between the laboratory and the industrial production. It is important to investigate the behaviour of Mg during LF refining under no Mg addition condition in industrial production.

In this study, the transfer behaviour of Mg in low carbon aluminium killed steel during LF refining was investigated in industrial production. The effect of Mg in the inclusions on the size distribution of inclusions was also studied. Finally, a kinetic model was developed to investigate the mechanisms underlying the variation of Mg concentration among the molten steel, the slag and inclusions.

Industrial practice

In order to investigate the transfer behaviour of Mg in low carbon aluminium killed steel during LF refining, sampling was performed in low carbon Al-killed steel with grade MR produced by Tangsteel plant. The production process was described as following: 100 ton BOF→LF→RH→1015 × 200 mm² CC. The operation conditions and the sampling scheme are shown in Fig. 1. In this operation, aluminium was added as deoxidiser first at 10 s, and aluminium was added second at 600 s. Ferro-manganese was added at 1200 s and Ca–Al line was added at 1500 s. The Ar gas flow rate for stirring varied from 0.05 to 0.8 Nm³ min⁻¹. The molten steel and the slag were sampled at 0 s (at the beginning of LF), 200 s (after Al addition first), 350 s (after Al addition second), 1550 s (after Ca addition), 1740 s (soft blowing for 4 min) during LF treatment. The composition of molten steel was analysed by optical emission spectroscopy. Mg and Ca contents in the metal were analysed by inductively coupled plasma atomic emission spectroscopy. The slag composition was analysed by fluorescent X-ray analysis. The composition of inclusions was analysed by scanning electron microscope (SEM, ZEISS ULTRA55) and energy dispersive spectroscopy (EDS, OXFORD INSTRUMENTS INCA X-MAX50).

Operational condition in tangsteel plant

Results

To investigate the transfer behaviour of Mg in low carbon aluminium killed steel during LF refining, the analyses of the compositions of the molten steel, the slag and inclusions were necessary. Figure 2 showed that the Mg concentration in the molten steel and in the slag. The Mg concentration in the molten steel increased with time, reaching 3.0 ppm at the end of LF refining in industrial production. MgO concentration in the slag almost unchanged. EDS results of inclusions were converted into mass percentage of CaO, MgO and Al₂O₃, and were shown in the MgO–CaO–Al₂O₃ ternary system phase diagram with Factsage software, as given in Fig. 3. The region with w_CaO between 40% and 60%, w_Al2O3 between 35% and 65% and w_MgO between 0% and 20% was the most favourable region for avoiding clogging nozzles where inclusions were liquid at 1873 K as shown in Fig. 3. In this work, 1500–3500 inclusions with the size of >1 µm were detected by scanning electron microscope (SEM, ZEISS ULTRA55) for each of the samples. Figure 3 showed the inclusions distribution and morphology during LF treatment. Initially, most of the inclusions were alumina as shown in Fig. 3 a. The MgO concentration increased with time and most of the inclusion converted into MgO–Al₂O₃–CaO ternary under the condition of no Mg addition as shown in Fig. 3 b and c. Figure 3 d showed that the inclusions compositions became much more scattered after Ca addition. Figure 3 e showed that average MgO content in inclusions reached 8.9% at the end of LF refining. A little part of inclusions moved into the low Melton point region while most of them were not in low Melton point region as shown in Fig. 3 e. From the kinetic prospective, there were two ways to move most inclusions into low Melton point region: on the one hand, more Ca addition was needed in order to move all the inclusion into low Melton point region. On the other hand, longer soft blowing time was needed.

Change of Mg concentration in molten steel, slag in industrial production

Inclusions distribution of samples during LF refining. a Inclusions distribution of Sample 1 at the beginning of LF refining (S1); b inclusions distribution of Sample 2 at 200 s (S2); c inclusions distribution of Sample 3 at 350 s (S3); d inclusions distribution of Sample 4 at 1550 s (S4); e inclusions distribution of Sample 5 at 1740 s (S5).

Figure 3 also showed that the size of inclusions concentrated in the range of 2–5 µm during LF treatment, and the number of inclusions with diameter 2–5 µm gradually increased with time. The number of inclusions with diameter 5–10 µm gradually decreased with time.

To understand the effect of Mg in the inclusions on the size of inclusions in the slab, the variation of Mg concentration in the inclusions with different sizes was shown in Fig. 4, where the value of X-axis represents for the range of inclusion size, while the value of Y-axis for the Mg content in the inclusions. The results showed that the inclusion size tend to become smaller with the increase of Mg concentration in the inclusions. So addition of Mg into the molten steel is probably a good method to reduce the size of inclusions in the future.

The variation of inclusion compositions with different sizes

Refractory containing Mg was also an important factor. So the erosion rate of the slag line and the ladle wall was tallied. Statistics from 120 heats showed that the erosion rate of the slag line was 0.025 mm min⁻¹ which was twice as that of the ladle wall. The composition of the lining in tangsteel was shown in Table 1.

Table 1

Composition of lining and erosion rate

Position of the lining	Composition of lining (%)			Erosion rate (mm/min)
Position of the lining	MgO (%)	C (%)	Al₂O₃ (%)	Erosion rate (mm/min)
Slag line	80	20		0.025
Ladle wall	10	10	80	0.0125

Discussion

In this study, a two-film theory model for the composition change in the oxide inclusions was coupled to a kinetic model for the slag-steel reactions that occur in the ladle furnace under industrial operations. The composition change of the component in the slag and the steel was calculated by a coupled reaction model which was reported by previous scholars^8–12 and proven to predict accurately the composition variation of the slag and molten steel. It was assumed that the chemical reactions at the interface were rapid and equilibrated at any time, and the reaction rates are controlled by mass transfer in both metal and slag phase. The reactions considered in this study are illustrated schematically in Fig. 5. The following reactions are considered in this model:

Reactions between the molten steel and the slag

Reactions between the molten steel and inclusions

Refractory erosion including slag line and wall

Schematic of the calculated model

The initial chemical compositions of the molten steel and slag were set according to the results from practical production at the beginning of LF treatment as shown in Tables 2 and 3 respectively.

Table 2

Initial compositions of molten steel

Initial composition	C	Mn	S	P	Si	Al	Ca	Mg	O
(mass %)	0.027	0.242	0.013	0.012	0.007	0.0474	0	0	0.0005

Table 3

Initial compositions of slag

Initial composition	FeO	SiO₂	CaO	MgO	Al₂O₃	P₂O₅	MnO	CaS
(mass %)	1.44	4.35	54.16	8.03	30.92	0.05	0.235	0.837

Mass transfer behaviour of Mg through steel/slag reaction

The coupled reaction kinetic model was based on the slag system of CaO–Al₂O₃–MgO–SiO₂–FeO–P₂O₅ and steel system of Mn–Si–Al–Ca–Mg–S–P–O for the first time. The reactions between the slag and the steel were expressed by equation (1).^13,14 (1) The equilibrium constant K_M can be defined by equation (2). (2) The desulphurisation reaction can be given by equation (3), and the equilibrium constant can be defined by equation (4). (3) (4) The effective equilibrium constants E_m can be expressed by equation (5). (5) In addition, the effective equilibrium constants of oxidation of Fe and desulphurisation reaction can be expressed by equations (6) and (7). (6) (7) During the transfer process of component M from molten steel to slag, the flux-density should be equal. The relationship can be expressed by equation (8). (8) In addition, the flux-density of Fe, s, dissolved oxygen can be expressed by equations (9)–(11) respectively. (9) (10) (11)

The parameter such as argon stirring rate was also considered. The mass transfer coefficient of the metal phase was calculated by an empirical equations reported by Kitamura et al.,¹⁵ which was shown in equations (12)–(14). (12)

(13)

(14)

where k_m and k_s represent the mass transfer coefficients in the film layer of the liquid metal and slag phase respectively, m/s;

is the stirring energy, w/t;

is the Ar gas flow rate, Nm³ min⁻¹; T_L and T_n are the temperatures of the molten steel and the gas respectively, K; W is the weight of the molten steel, t; h ₀ and h_v are the injection and the bath depth respectively, m; Pa is the atmospheric pressure, Pa; d_v is the diameter of the bath, m. The parameters for model calculation were shown in Table 4.

Table 4

Parameters for model calculation

Parameter	Value
	1873
	4
	1.7
	3.42
Q (Nm³ min⁻¹)	0.05, 0.4, 0.8
	7000
	101 400
	15 000

Based on the above equations, the interfacial concentrations of species of both molten steel and the slag can be deduced as equations (15)–(20). (15) (16) (17) (18) (19) (20)

The interfacial concentration of all the elements dissolved in molten steel can be expressed as a function of bulk concentrations and the interfacial oxygen activity. So the interfacial oxygen activity ao* can be obtained by solving the equation (21). (21) With the value of interfacial oxygen activity, the mass transfer of each component in the slag and molten steel can be calculated during a time step (Δt) according to the following equations (22) and (23). In the reaction, the cross-sectional area of the bath is used as the interfacial area between the metal and slag phase. (22) (23)

Mass transfer behaviour of Mg through refractory erosion

The refractory dissolved into the steel or slag would affect the chemical composition of molten steel indirectly via the reaction between the slag and the molten steel. So the erosion of refractory including slag line and ladle wall was taken into consideration. The reaction between slag and slag line was considered as MgO from ladle wall dissolution into the slag which reacted with the molten steel. The reaction between ladle wall and the molten steel was considered as equations (24) and (25). The total Mg content in the molten steel came from not only the slag/metal reaction but also the refractory erosion. (24) (25)

Mass transfer behaviour of Mg through inclusion/steel model

From the above analysis, [Mg] and [Ca] existed in Melton steel through slag-metal reaction, even though Mg and Ca were not added in the early stage of the treatment. As a result, the deoxidised product Al₂O₃ was modified to Al₂O₃–MgO–CaO inclusions. A two-film theory model was employed to analyse the mechanism of the process of Al₂O₃ modification by Ca and Mg in the molten steel, under the assumption that diffusion of Mg and Ca in the molten steel was rate-determination. It was also assumed that the shape of inclusion particles was spherical and the outer diameter (Ro)of the particles did not change over time.

According to the balance of Mg in the metal side and inclusion side, the equation can be expressed by equation (26). It is then possible to compute the compositions of MgO in inclusions according to equations (26) and (27). (26) (27)

Calculation results

Figures 6 and 7 showed the calculation results and industrial results for the Mg concentration change in the molten steel and in the slag with time. Mg concentration increased with time while MgO in the slag almost unchanged. The calculated results agreed well with the results of the industrial results.

Calculation results and industrial results for the Mg concentration change in the molten steel

Calculation results and industrial results for the concentration change in the slag

Figure 8 showed comparison of Mg concentration from slag/steel reaction with that from refractory erosion. The results showed that, the concentration of Mg from the refractory erosion was higher compared with the concentration of Mg from slag/steel reaction. So it was concluded that the refractory erosion was the main resource of Mg in the molten steel.

Comparison of Mg concentration from slag/steel reaction with that from refractory erosion

Figure 9 showed calculated results and industrial results for the concentration of MgO in inclusions with time. And the MgO concentration reached 6.8% at the end of LF refining as shown in Fig. 9. The calculated results agreed with the results of the industrial results shown in Fig. 3.

Calculated results and industrial results for the concentration of MgO in inclusions

Conclusion

Under no Mg addition condition, the Mg concentration in the molten steel and the inclusions increased with time. The Mg concentration reached 3.7 ppm and the MgO concentration in inclusions reached 8.9% respectively. The Mg concentration in the slag almost unchanged. The erosion rate of refractory was tallied. The erosion rate of the slag line was 0.025 mm min⁻¹ which was twice as that of the ladle wall.

The inclusion size tended to become smaller with the increase of Mg concentration in the inclusions.

Form kinetics prospective, a predicted model based on a two-film theory model were established in order to explain the mass transfer of Mg between slag, steel and inclusions during LF treatment and validated by the industrial results in tangsteel plant. The results showed that the refractory erosion was the main resource of Mg in the molten steel.

References

Guo

Cheng

S. S.

Cheng

Z. J.

: ‘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, (12), 2142–2151. doi: 10.2355/isijinternational.53.2142

Zhao

D.-W.

H.-B.

Bao

C.-L.

Yang

: ‘Inclusion evolution during modification of alumina inclusions by calcium in liquid steel and deformation during hot rolling process’, ISIJ Int., 2013, 53, (12), 2115–2124.

Jiang

Wang

X.-H.

Chen

Wang

: ‘Laboratory study on evolution mechanisms of non-metallic inclusions in high strength alloyed steel refined by high basicity slag’, ISIJ Int., 2010, 50, (1), 95–104. doi: 10.2355/isijinternational.50.95

Jansson

Brabie

Jönsson

: ‘Magnesia–carbon refractory dissolution in Al killed low carbon steel’, Ironmak. Steelmak., 2006, 33, (5), 389–397. doi: 10.1179/174328106X113977

Harada

Miyano

Maruoka

Shibata

Kitamura

S.-Y.

: ‘Dissolution behavior of Mg from MgO into molten steel deoxidized by Al’, ISIJ Int., 2014, 54, (10), 2230–2238. doi: 10.2355/isijinternational.54.2230

Kasimagwa

Brabie

Jönsson

P. G.

: ‘Slag corrosion of MgO–C refractories during secondary steel refining’, Ironmak. Steelmak., 2014, 41, (2), 121–131. doi: 10.1179/1743281213Y.0000000110

Harada

Maruoka

Shibata

Zeze

Asahara

Huang

Kitamura

S.-Y.

: ‘Kinetic analysis of compositional changes in inclusions during ladle refining’, ISIJ Int., 2014, 54, (14), 2569–2577. doi: 10.2355/isijinternational.54.2569

Robertson

D. G. C.

Deo

Ohguchi

: ‘Multicomponent mixed-transport-control theory for kinetics of coupled slag/metal and slag/metal/gas reactions’, Ironmak. Steelmak., 1984, 11, (1), 45–55.

Ohguchi

Roberson

D. G. C.

Deo

: ‘Simultaneous dephosphorisation and desulphurization of molten pig iron’, Ironmak. Steelmak., 1984, 11, (4), 202–213.

10.

Harada

Maruoka

Shibata

Kitamura

S.-Y.

: ‘A kinetic model to predict the compositions of metal, slag and inclusions during ladle refining: Part 1. Basic concept and application’, ISIJ Int., 2013, 53, (12), 2110–2117. doi: 10.2355/isijinternational.53.2110

11.

Harada

Maruoka

Shibata

Kitamura

S.-Y.

: ‘A kinetic model to predict the compositions of metal, slag and inclusions during ladle refining: Part 2. Condition to control the inclusion composition’, ISIJ Int., 2013, 53, (12), 2118–2125. doi: 10.2355/isijinternational.53.2118

12.

Zhang

Xie

H. Y.

Diao

C. Q.

: ‘Coupled reaction kinetics of duplex steelmaking process for high phosphorus hot metal’, Ironmak. Steelmak., 2013, 40, (4), 282–289. doi: 10.1179/1743281212Y.0000000036

13.

Chen

J. X.

: ‘Handbook of commonly used diagrams ad data in steelmaking process’, 1984, Beijing, Metallurgical Industry Press.

14.

Huang

X. H.

: ‘Principles of iron and steel metallurgy’, 2002, Beijing, Metallurgical Industry Press.

15.

Kitamura

S.-Y.

Kitamura

Shibata

Mizukami

Mukawa

Nakagawa

: ‘Effect of stirring energy, temperature and flux composition on hot metal dephosphorization kinetics’, ISIJ Int., 1991, 31, (12), 1322–1328. doi: 10.2355/isijinternational.31.1322