Viscosity and Solidification Temperature of SiO₂–CaO–Na₂O Melts for Fluorine Free Mould Flux

Abstract

Fluorine-free mould flux has been required to reduce corrosion of continuous casters. In this paper, Na₂Ca₂Si₃O₉ has been applied to an alternative crystalline phase of Ca₄Si₂O₇F₂ (cuspidine) that crystallized in commercial flux films. Effects of basicity described as T.CaO/SiO₂ and additions of Li₂O, Na₂O, MgO and/or MnO into Na₂Ca₂Si₃O₉ composition mould flux on viscosity and solidification temperature were studied and the solidified specimens were examined by X-ray diffraction to identified crystalline phases.The viscosity of the mould flux is reduced with increaing of the basicity or the basic components contents. The solidification temperature decreased slightly as the basicity increased from 0.47 to 0.60, whereas it increased steeply with increasing of the basicity from 0.60 to 0.70. The solidification temperature also steeply increased when large amounts of these basic components were added. With increasing of the basic components amounts, the first peak intensity of the target phase, Na₂Ca₂Si₃O₉, decreased while that of Na₂Ca₂Si₂O₇, which had higher melting point than the target phase, increased. It is indicated that the solidification temperature is related to Na₂Ca₂Si₂O₇ crystallization.Carbon steel was cast with one of the developed fluorin-free mould fluxes and slab without any surface cracking was obtained.

1. Introduction

Fluorine is an important component to adjust physical properties of mould flux appropriately as well as to crystallize Ca₄Si₂O₇F₂ (cuspidine) in flux films to control a heat transfer for mild cooling.¹⁾ Therefore, the crystallization of cuspidine has been studied about phase relation,²⁾ primary field³⁾ and precipitation.⁴⁾ Gibbs energy of formation of cuspidine has been also determined.⁵⁾

On the other hand, a reaction between the mould flux and secondary cooling water causes fluorine dissolution into the water and generates hydrogen fluorine that corrodes plant equipment.⁶⁾ Thus, fluorine-free mould flux has been required. Recently, Fox et al.⁷⁾ and Nakada et al.⁸⁾ have submitted B₂O₃ and TiO₂ which are alternative substitutes for CaF₂ in the mould fluxes, respectively. These reports indicate possibility of fluorine-free mould fluxes.

In this study, SiO₂–CaO–Na₂O system is focused as a base system of the fluorine-free mould flux. Na₂Ca₂Si₃O₉, which has a melting point of 1280–1290 degree Celsius,^9,10,11) is thought to be an applicable crystalline phase which substitutes for cuspidine. Effects of additions of Li₂O, Na₂O, MgO and/or MnO to Na₂Ca₂Si₃O₉ composition of the mould flux mixed with law materials on solidification temperatures, viscosities and crystalline phase have been studied.

2. Experimental

2.1. Specimen Preparation

The equilibrium phase diagram of SiO₂–CaO–Na₂O system is shown in Fig. 1. The primary field of Na₂Ca₂Si₃O₉ consulted with the reference⁹⁾ is also illustrated in Fig. 1. Specimen components of SiO₂, CaO and Na₂O correspond to the shaded region in Fig. 1, and Li₂O, MgO and/or MnO were added into them. These specimens were prepared by mixing industrial raw materials. The compositions of these specimens are shown in Table 1. R_S in Table 1 named “soda ratio” in this paper is defined as a ratio of Na₂O mass divided by total mass of SiO₂, CaO and Na₂O.

Fig. 1.

Equilibrium phase diagram of Na₂O–CaO–SiO₂ system and primary field of Na₂Ca₂Si₃O₉.

Table 1.Basicity and compositions of specimens.

Basicity (T.CaO/SiO2)	0.47–0.70
Na2O	14.9–25.8 mass%
RS	0.18–0.28
Li2O	0–2.2 mass%
MgO	0.4–7.8 mass%
MnO	0–0.9 mass%
Al2O3	1.9–3.6 mass%

2.2. Measurements Procedure and Analysis

The viscosities and the solidification temperature of the mould fluxes were evaluated with the oscillating-plate viscometer.¹²⁾ 1 kg of the sample was melted at 1673 K in a carbon crucible under argon atmosphere. Then its temperature was decreased at a rate of 2 K/min, and the viscosity was measured continuously. When the sample began to solidify, the viscosity rose rapidly. A breakpoint found in relationship between the temperature and the viscosity was defined as the solidification temperature.

After the physical properties were measured, a solidified specimen in the crucible was crushed into powder and analysed by X-ray diffraction. From peak patterns of the X-ray diffraction, crystalline phase was identified. An intensity of the X-ray diffraction at the first peak of each crystalline phase was used as an index of the crystallization.

3. Results and Discussion

3.1. Solidification Temperature and Crystalline Phase

The relationship between the basicity and the solidification temperature is illustrated in Fig. 2. The solidification temperature decreased slightly as the basicity increased from 0.47 to 0.60, whereas it increased steeply with increasing of the basicity from 0.60 to 0.70. The effect of an amount of Li₂O, Na₂O or MgO into Na₂Ca₂Si₃O₉ composition fluxes on the solidification temperature is shown in Fig. 3. The solidification temperature decreased as the amount of Li₂O or MgO increased. When the content of Na₂O was less than 6 mass%, it had little influence on the solidification temperature. On the other hand, the solidification temperature increased a little at around 10 mass% Na₂O.

Fig. 2.

Relationship between the basicity and the solidification temperature.

Fig. 3.

Effect of an amount of Li₂O, Na₂O or MgO on the solidification temperature.

The peak patterns of the X-ray diffraction showed that Na₂Ca₂Si₃O₉ crystallized in all specimens and Na₂Ca₂Si₂O₇ crystallized in several specimens. The relationship between the basicity and intensities of the first peaks of these two kinds of crystalline phases is illustrated in Fig. 4. The first peak intensity of Na₂Ca₂Si₃O₉ decreased as the basicity increased from 0.60 to 0.70, whereas the first peak intensity of Na₂Ca₂Si₂O₇ increased with increasing of the basicity from 0.60 to 0.70. Effects of Na₂O content on the first peak intensities of Na₂Ca₂Si₃O₉ and Na₂Ca₂Si₂O₇ were shown in Fig. 5. With increasing of Na₂O content, the first peak intensity of Na₂Ca₂Si₃O₉ decreased and the intensity of Na₂Ca₂Si₂O₇ increased. While the addition of Li₂O or MgO caused the first peak intensity of Na₂Ca₂Si₃O₉ to decrease, Na₂Ca₂Si₂O₇ was little observed.

Fig. 4.

Relationship between the basicity and the first peak intensities of Na₂Ca₂Si₃O₉ and Na₂Ca₂Si₂O₇.

Fig. 5.

Effect of an amount of Na₂O on the first peak intensities of Na₂Ca₂Si₃O₉ and Na₂Ca₂Si₂O₇.

From Figs. 2, 3, 4, 5, it is indicated that the solidification temperature increased with increasing the first peak intensity of Na₂Ca₂Si₂O₇. A “compositional length”, L, which is defined in Eq. (1), is applied to investigate the effect of the composition at the basicity of around 0.60 on the first peak intensities of these two kinds of crystalline phases and the solidification temperatures.   

$$L=\sqrt{\sum _i{(X_i-Y_i)}^2}$$(1)

i: SiO₂, CaO, Li₂O, Na₂O, MgO, MnO, Al₂O₃

Here, X and Y are the mole fractions of specimens and Na₂Ca₂Si₃O₉, respectively. This length represents a compositional distance between specimens and the target phase. Relationships between the compositional length and the first peak intensities of Na₂Ca₂Si₃O₉ and Na₂Ca₂Si₂O₇ and the solidification temperature are shown in Fig. 6. The first peak intensity of Na₂Ca₂Si₃O₉ steeply decreased and that of Na₂Ca₂Si₂O₇ gradually increased with the compositional length increase. The solidification temperature increased steeply as the compositional length increased from 0.10 to 0.17, where the first peak intensity of Na₂Ca₂Si₂O₇ was relatively high. This result indicates that the solidification temperature is increased with crystallization of Na₂Ca₂Si₂O₇, which has about 170 degree Celsius higher melting point than Na₂Ca₂Si₃O₉. At the constant basicity, increasing of amounts of Li₂O, Na₂O, MgO and/or MnO, which are obviously basic components, cause both activities of SiO₂ to decrease and of CaO to increase. Accordingly, it is considered that Na₂Ca₂Si₂O₇, which has higher basicity than Na₂Ca₂Si₃O₉, crystallizes with increasing of the compositional length.

Fig. 6.

Relationships between the compositional length and the first peak intensities of Na₂Ca₂Si₃O₉ and Na₂Ca₂Si₂O₇ and the solidification temperature.

3.2. Viscosity

The relationship between the basicity and the viscosity at 1300 degree Celsius is illustrated in Fig. 7. The viscosity decreased with increasing of the basicity. The effect of the addictive amount of Li₂O, Na₂O or MgO on the viscosity at 1300 degree Celsius is shown in Fig. 8. Figure 8 expresses that these basic components, especially Li₂O, affect on decreasing of the viscosity. The viscosity behaviors including Li₂O or MgO with different soda ratio are shown in Figs. 9 and 10. From these two figures, it is demonstrated that the Li₂O or MgO addition conducts the viscosity reducing and it is more effective with lower contents of Na₂O. The relationship between the compositional length and the viscosity at 1300 degree Celsius at the basicity of aronud 0.60 is shown in Fig. 11. The viscosity was indicated to decrease as the compositional length increased because the basic components, except for Al₂O₃, were added.

Fig. 7.

Relationship between the basicity and the viscosity at 1300 degree Celsius.

Fig. 8.

Effect of the amount of Li₂O, Na₂O or MgO on the viscosity at 1300 degree Celsius.

Fig. 9.

Effect of the amount of Li₂O at each constant soda ratio on the viscosity at 1300 degree Celsius.

Fig. 10.

Effect of an amount of MgO at each constant soda ratio on the viscosity at 1300 degree Celsius.

Fig. 11.

Relationship between the compositional length and the viscosity at 1300 degree Celsius.

Viscosity is considered to depend on the network of silicate, and activity of SiO₂ is applied as an index of the viscosity. The activity of SiO₂ in molten flux decreases with increasing of the basic components, as same as the viscosity decrease. The relationship between the viscosity at 1300 degree Celsius and the SiO₂ activity at same temperature calculated by using FactSage thermodynamic software and FToxide Database¹³⁾ is shown in Fig. 12. Because FToxide does not support a solution database including Li₂O, the results seen in Fig. 12 are comprised without Li₂O. Figure 12 demonstrates that the viscosity at 1300 degree Celsius decreases lineally as the SiO₂ activity decreases in the range of composition applied in this study.

Fig. 12.

Relationship between the SiO₂ activity and the viscosity at 1300 degree Celsius.

3.3. Prospective Composition

As described above, the addition of basic components into Na₂Ca₂Si₃O₉ composition flux mixed with industrial materials causes Na₂Ca₂Si₃O₉ crystallization to decrease while it is required to reduce the viscosity. From Figs. 6 and 11, however, there are available specimens as the mould fluxes in the range of compositional length from 0.05 to 0.10. In this range, the first peak intensity of Na₂Ca₂Si₃O₉ is much higher than that of Na₂Ca₂Si₂O₇, and it is indicated that Na₂Ca₂Si₃O₉ mainly crystallizes in flux films and the compositions in this range prevent steep increases of the solidification temperature caused by Na₂Ca₂Si₂O₇ crystallization. Tsutsumi et al.¹⁴⁾ have observed that the critical cooling rate for glass formation of Li₂O–SiO₂, Na₂O–SiO₂ and Na₂O–CaO–SiO₂ systems by hot-thermocouple methods with DTA. From the report, it is estimated that the critical cooling rate of mould flux whose basicity is 0.62 achieves around 20 mass% Na₂O and shows little change with more Na₂O added, and that Li₂O has more effect on the critical cooling rate than Na₂O. These are important information on the crystallization of the present mould flux.

In the components range used in the present study, the lowest viscosities at 1300 degree Celsius are around 3 poise, which are thought to be an available viscosity for casting since the viscosities of commercial mold fluxes^15,16) are approximately from 1 to 4 poise at that temperature.

4. Casting Test

4.1. Casting Condition

One of the developed fluorine-free mould fluxes shown in Table 2 and 3, whose composition was given in the above discussion, was applied to steel casting with our test continuous caster for slab casting. The casting test with the conventional mould flux seen in Tables 2 and 3 was also carried out. 0.47%C steel, whose composition was shown in Table 4, was casted under the condition of velocity of 0.70 m/min, oscillation stroke of 7 mm, and negative strip ratio of 40%.

Table 2.Compositions of mould flux (mass%).

mould flux	T.CaO/SiO2	T.Na2O+Li2O	T.F	others
conventional	0.80	5	5	8
developed	0.62	22	–	4

Table 3.Physical properties and crystalline phase.

mould flux	Viscosity at 1300°C (poise)	Solid. temp. (°C)	Main Crystalline Phase
conventional	2.5	1150	Ca4Si2O7F2
developed	3.2	1100	Na2Ca2Si3O9

Table 4.Composition of casted steel (mass%).

C	Si	Mn	P	S	Cr
0.47	0.23/0.24	0.72/0.77	0.010/0.011	0.020/0.024	0.15/0.16

4.2. Casting Results

There were no cracks found on the surface of each slab. Several thermocouples were installed longitudinally-arranged in the centre position in width of the casting mould. The temperatures measured by the thermocouples during casting are shown in Fig. 13. The temperatures decreased as the distances from meniscus increased. While the each temperature shown in Fig. 13(b) is a little higher than ones in Fig. 13(a), it is an acceptable difference. If a poor flow of mould flux between the mould and the initial solidified shell or unstable heat transfer in the flux film occurred, the temperatures measured by the thermocouples would show a large change though the casting. Figure 13, however, expresses that the temperature changes seen in the casting with the developed mould flux are as small as with the conventional one. It is demonstrated that stable casting is achieved by using the fluorine-free mould flux.

Fig. 13.

Temperatures measured by thermocouples installed the casting mould during casting with (a) conventional and (b) developed mould fluxes.

5. Conclusions

In the present study, the fluorine-free mould flux has been developed. Na₂Ca₂Si₃O₉ turned to be available for main crystalline phase in the flux film in the mould, instead of cuspidine for the conventional. The reasonable viscosity for lubrication is obtained by adjusting the contents of the basic components without fluorine. In the actual casting, the developed flux brings stable lubrication, heat transfer and surface quality of without any cracking.

References

• 1)   M.  Hanao,  M.  Kawamoto and  T.  Watanabe: ISIJ Int., 44 (2004), 827.
• 2)   T.  Watanabe,  H.  Fukuyama and  K.  Nagata: ISIJ Int., 42 (2002), 489.
• 3)   H.  Nakada,  H.  Fukuyama and  K.  Nagata: ISIJ Int., 46 (2006), 1660.
• 4)   M.  Hayashi,  T.  Watanabe,  H.  Nakada and  K.  Nagata: ISIJ Int., 46 (2006), 1805.
• 5)   H.  Fukuyama,  H.  Tanabe,  T.  Oshita and  K.  Nagata: ISIJ Int., 44 (2004), 1488.
• 6)   T.  Omoto,  Y.  Iwamoto and  H.  Yamaji: Shinagawa Tech. Rep., 45 (2002), 85.
• 7)   A. B.  Fox,  K. C.  Mills,  D.  Lever,  C.  Bezerra,  C.  Valadares and  I.  Unamuno: ISIJ Int., 45 (2005), 1051.
• 8)   H.  Nakada and  K.  Nagata: ISIJ Int., 46 (2006), 441.
• 9)  VDEh: Slag Atlas, 2nd ed., Verlag Stahleisen GmbH, Düsseldorf, (1995), 138.
• 10)   E. M.  Levin,  C. R.  Robbins and  H. F.  McMurdie: Phase Diagram for Ceramists, Vol. 1, The American Ceramic Society, Inc., Columbus, OH., (1964), 175.
• 11)   G. K.  Moir and  F. P.  Glasser: J. Am. Ceram. Soc., 59 (1976), 199.
• 12)   T.  Iida,  M.  Kawamoto,  S.  Fujimoto and  Z.  Morita: Tetsu-to-Hagané, 71 (1985), 1490.
• 13)  FactSage, www.factsage.com
• 14)   K.  Tsutsumi,  J.  Ohtake,  T.  Nagasaka and  M.  Hino: Tetsu-to-Hagané, 84 (1998), 464.
• 15)   H.  Nagano,  K.  Koyama,  T.  Nakano,  T.  Mukai,  T.  Komai and  N.  Kaneko: Tetsu-to-Hagané, 70 (1984), S145.
• 16)   Y.  Yamaoka,  T.  Matsuzaki,  M.  Osame,  H.  Tanabe,  H.  Fujiwara and  H.  Sakai: CAMP-ISIJ, 10 (1997), 895.