Dissolution Mechanism of Carbon Brick into Molten Iron

Abstract

In order to investigate the dissolution mechanism of carbon brick into molten iron, cylindrical specimens were immersed into molten iron to carry out the experiments. The dissolution reaction of carbon was considered as the dominant reaction through thermodynamic analysis, the result of SEM revealed the hole diameter decrease from the reaction interface to the center position of carbon brick. The quantitative relationship between element content and erosion was obtained through the experimental results, the characteristic parameters were selected to compare the influence degree of element content on the erosion. The calculation model of mass transfer coefficient was established, the dissolution reaction of the sample is controlled by interfacial reaction and mass transfer of carbon when the phosphorus content up to 0.2% in molten iron. The adsorption of sulfur on the iron-carbon interface covers part of the effective surface, the degree of adsorption on the interface depends on the proportion of sites covered by sulfur.

1. Introduction

In recent years, many studies have been carried out to extend the campaign life of blast furnace (BF).^1,2) The hearth sidewall which is considered as the key area of BF has attracted more and more concerns.^3,4) The direct contact between carbon refractory and carbon-unsaturated molten iron is regarded as the main reason for the hearth sidewall erosion. The erosion is inevitable due to the carbon is easily dissolved into molten iron.^5,6,7) Therefore, the dissolution of carbon refractory into molten iron is an important reaction for the ironmaking process.

Many papers have been reported to describe the dissolution of carbon into molten iron. The dissolution rate of carbon into molten iron was investigated by Dahlke.⁸⁾ The effects of carbon structure and extent of porosity on the dissolution were studied in some literature.^9,10,11) Cham et al.¹²⁾ calculated the activation energy for the dissolution of carbon from graphite. Shigeno et al.¹³⁾ supposed that the dissolution rate of carbon from coke is lower than that of pure graphite. Mourao et al.¹⁴⁾ asserted that a viscous layer formed by ash in carbon source was found around the interface. Orsten et al.¹⁵⁾ observed that a gas film around the interface formed by volatile matter from carbon source will hinder the dissolution process. Kayama et al.¹⁶⁾ assumed a high rate of dissolution can be obtained when the coke with low porosity.

Wright and Badlock¹⁷⁾ investigated the increase of sulfur content in molten iron decreases the graphite dissolution at 1450°C. While, Kalvelage et al.¹⁸⁾ reported that 0.28% sulfur in molten iron helps the dissolution of graphite, the dissolution kinetics is not affected when the sulfur content above 0.28%. Oersten and Oeters¹⁹⁾ observed the effect of sulfur on non-graphitic carbon is more significant. Hisatsuna et al.¹⁰⁾ found silicon promotes the dissolution.

Most researchers concluded the carbon diffusion in molten iron is the rate limiting step.^20,21,22) Kalvelage et al.¹⁸⁾ found an increase in the mass transfer coefficient when molten iron containing sulfur, whereas a decrease when the sulfur content beyond 0.03% caused by mixed control. Ericsson et al.²³⁾ reported a decrease in the mass transfer coefficient with the increasing sulfur content at a low rotational speed because the diffusion coefficient of carbon in molten iron decreased. Shurigin and Kryk²⁴⁾ obtained the diffusion coefficient of carbon in Fe–C, Fe–Si, Fe–P and Fe–Ni alloys.

In spite of the many investigations on the dissolution of carbon, few scholars paid attention to the dissolution of NMA carbon brick (a kind of hot pressing small carbon brick made in USA) into molten iron although NMA carbon brick is often used on the hot surface of hearth sidewall. Moreover, the actual molten iron composition in BF was not considered in previous studies and the quantitative relationship between element content and erosion was not clarified, interfacial reactions were not taken into account to analysis the rate limiting step. Therefore, an attempt has been made to explore the dissolution of NMA carbon brick into molten iron, various iron containing carbon, silicon, manganese, phosphorus, sulfur and titanium were melt to understand the dissolution.

2. Experimental

2.1. Materials and Samples Preparation

The reduced iron powder (AR, purity greater than 98%), graphite powder (CP, purity greater than 99.85%), silica fume (purity greater than 99%), manganese powder (purity greater than 99.9%), FeS powder (AR, purity greater than 99%) and FeP (P=25.7%) and titanium powder (purity greater than 99%) were used to configure molten iron with different components. The mixture (200 g) at the desired mass percent was ground thoroughly in a mortar in order to increase the contact among the particles, ensure the mixture can melt completely. In order to ensure a certain liquid level of molten iron in the experiment, the mixture was packed into a dense alumina crucible (34 mm I.D.) and the dense alumina crucible was placed in the protection crucible (62 mm I.D.).

NMA carbon brick which was the main lining utilized in the large-scale BF hearths was taken from a commercial BF to carry out the experiment. The carbon brick was cut into cylindrical specimens with 12 mm in diameter and 50 mm in length through a special drill (Fig. 1(a)). The chemical composition of NMA carbon brick is presented in Table 1. The main components of NMA carbon brick are carbon and silica which is demonstrated by the XRD (Cu-Kα) pattern analysis (Fig. 2(a)). Before the experiment, the cylindrical specimen must be connected with the corundum rod through a high temperature binder in order to realize the stirring of molten iron (Fig. 1(b)).

Fig. 1.

Schematic diagram of cylindrical specimen.

Table 1. The chemical composition of NMA carbon brick.

Constituents (%)	C	SiO2	Al2O3	S	Fe2O3	Other
Numerical value	82.30	8.75	2.10	1.50	1.36	3.99

Fig. 2.

The XRD pattern analysis of NMA carbon brick. (Online version in color.)

2.2. Reactor

The high temperature tube furnace and the stirring equipment were employed to carry out the experiments. The internal structure of high temperature tube furnace was presented in Fig. 3. The U-shape MoSi₂ were used as heating elements to reach the desire temperature. A Pt-6%Rh/Pt-30%Rh thermocouple was placed under the protection crucible to measure the temperature of molten iron. The constant temperature zone was about 8 cm in length, and the highest accuracy area (±1°C) was located at 3 cm above the crucible supporter which was confirmed by a standard thermocouple before the experiment, so that the temperature measured by the Pt-6%Rh/Pt-30%Rh thermocouple matched the actual temperature of molten iron. The top of the high temperature tube furnace was motor, the corundum stirring rod was connected with the motor and the cylindrical specimen, respectively. The cylindrical specimen can be immersed in molten iron to carry out the experiment.

Fig. 3.

The internal structure of high temperature tube furnace. (Online version in color.)

The stirring equipment was motor combined electric stirrer. The stirring rod was connected with the motor by a joint (Fig. 4), and the electric motor was connected with an electric stirrer to control the rotational speed to complete the experiment under the condition of dynamic stirring to simulate the molten iron flow in the hearth.

Fig. 4.

Schematic diagram of the mixing equipment.

2.3. Experimental Procedure

The crucible assemble containing materials (200 g) was placed in the furnace, the chemical composition of molten iron and experimental conditions of the experiments are presented in Fig. 5. The furnace tube was purged with high-purity argon gas (99.999%) throughout the experiment at a flow rate of 3 L/min. The crucible assemble was heated from room temperature to the desired temperature at a rate of approximately 5°C/min. After the molten iron reached the desired temperature, the temperature was maintained for 60 minutes and the molten iron was stirred by a glass rod to ensure a uniform composition. The stirring rod was preheated before into the furnace, then, the stirring rod was slowly put into the furnace. The experiment was carried out for 60 minutes after the stirring rod was placed in contact with the molten iron. Samples of molten iron were withdrawn by using a glass tube (4 mm I.D.) at every 15 minutes and quenched quickly in water for chemical analysis. The XRD (X-ray Diffraction) analysis of NMA carbon brick was characterized after the experiment. SEM (Scanning Electron Microscope) and EDS (Energy Dispersive Spectrometer) were conducted on the iron-carbon interface after the experiment.

Fig. 5.

Schematic diagram of technical route.

3. Results and Discussion

3.1. Thermodynamic Analysis of Carburization

Firstly, the thermodynamic analysis of carburization was studied due to the carbon content in molten iron has a significant effect on carburization. The carburization of molten iron can be expressed by the following two chemical reactions:   

$$C=\left[C\right]\hspace{1em}\mathrm{\Delta }{G}^{\mathrm{\Theta }}=22\mathrm{\hspace{0.17em} }590-42.26T,\mathrm{\hspace{0.17em} }\text{kJ}/\text{mol}$$(1)

  

$$C+3Fe=F{e}_{3}C\hspace{1em}\mathrm{\Delta }{G}^{\mathrm{\Theta }}=10\mathrm{\hspace{0.17em} }530-10.20T,\mathrm{\hspace{0.17em} }\text{kJ}/\text{mol}$$(2)

According to the above two reactions, the carburization is composed of two processes: one process is the solid carbon source dissolved into the molten iron, the other process is carbon reacts with iron atoms to form Fe₃C, the XRD pattern analysis of carbon brick after the experiment proved this (Fig. 2(b)). The standard Gibbs free energy of solid carbon source dissolved into the molten iron is always lower than the other one in the experimental temperature range (Fig. 6). It is proved that the dissolution reaction of carbon is the dominant reaction.

Fig. 6.

The change trend of standard Gibbs free energy of carburization with temperature.

3.2. Reaction Interface Analysis

The micro morphology of carbon brick was observed by SEM after the experiment (Fig. 7). There are a lot of holes near the reaction interface. As presented in Fig. 7(d), the average hole diameter near the reaction interface is 164 μm, while the average hole diameter near the center position of carbon brick is 76 μm, the hole diameter decreases from the reaction interface to the center position of carbon brick. This can be explained in two ways: Firstly, there are some volatiles in NMA carbon brick since it is hot pressing brick, the volatiles disappear and the holes are formed at high temperature. The hole diameter increases due to some of the carbon is oxidized at high temperature. Therefore, holes were observed both near the reaction interface and the center position of carbon brick. Secondly, a chemical reaction takes place at the reaction interface:   

$$\begin{array}{l}Si{O}_{2(s)}+2C=S{i}_{(s)}+2CO\\ \mathrm{\Delta }{G}^{\mathrm{\Theta }}=729\mathrm{\hspace{0.17em} }794-379.34T\left(J/mol\right)\end{array}$$(3)

Fig. 7.

The micro morphology of carbon bricks after the experiment. (Online version in color.)

The starting temperature of reaction (3) is 1923 K, and the experimental temperature is 1773 K which is lower than the starting temperature. However, the Gibbs free energy of reaction (3) can be reduced since Si and Fe in molten iron produce a variety of stable compounds, for example:   

$$Fe+Si=FeSi\hspace{1em}\mathrm{\Delta }{G}^{\mathrm{\Theta }}=-80\mathrm{\hspace{0.17em} }390-4.19T\left(J/mol\right)$$(4)

Therefore, reaction (3) can be expressed by reaction (5):   

$$\begin{array}{l}Si{O}_{2(s)}+2C+Fe=FeSi+2CO\\ \mathrm{\Delta }{G}^{\mathrm{\Theta }}=649\mathrm{\hspace{0.17em} }404-383.53T\left(J/mol\right)\end{array}$$(5)

The starting temperature of reaction (5) is 1693 K, therefore the reaction can occur under experimental condition. CO gas generated by the reaction diffused along the hole at the reaction interface in carbon brick result in the hole diameter near the reaction interface is larger.

3.3. The Quantitative Relationship Between Element Content and Erosion

The diameter of carbon brick before and after the experiment was measured, the erosion rate of carbon brick was defined based on the results of the measurement:   

$$v=\frac{\pi \times l\times \left(d_0^2-d_f^2\right)\times \rho \times w_C}{4\times t\times s}$$(6)

where v is the erosion rate of carbon brick, g·h⁻¹·cm⁻²; l is the immersion depth of carbon brick into the molten iron, cm; d₀ is the diameter of carbon brick before the experiment, cm; d_f is the diameter of carbon brick after the experiment, cm; ρ is the density of carbon brick, g·cm⁻³; w_c is the carbon content of the carbon brick, 82.3%; t is the reaction time, h; s is the reaction area, cm².

The erosion rate of carbon brick can be calculated based on the results of the measurement of the diameter. It is evident from Fig. 8 that the erosion rate of carbon brick was dependent on the element content in molten iron. The quantitative relationship between the erosion rate and single element content was obtained by fitting curve with experimental data. The exponential value and slope of the curve were selected as the characteristic parameters to compare the influence degree of element content on the erosion rate as presented in Table 2. Carbon, silicon and titanium decreased the erosion rate, the order of influence degree: carbon > silicon > titanium, whereas manganese, phosphorus and sulfur increased the erosion rate, the order of influence degree: sulfur > phosphorus > manganese.

Fig. 8.

The erosion rate of carbon brick under different element content in molten iron.

Table 2. The quantitative relationship between the erosion rate and single element content.

Element	Quantitative relationship	Characteristic parameter	The erosion rate	Influence degree
C	y=64.39×[C]−4.74	−4.74	–	☆☆☆
Si	y=0.03×[Si]−0.64	−0.64	–	☆☆☆
Mn	y=0.001241×exp([Mn]/0.11672)+0.03863	0.12	+	☆
P	y=0.00668+0.4843[P]	0.48	+	☆
S	y=0.03473+0.51946[S]	0.52	+	☆☆
Ti	y=0.11704−0.57233[Ti]	−0.57	–	☆☆

3.4. Calculation of Mass Transfer Coefficient and Rate Limiting Step Analysis

3.4.1. Establishment of Theoretical Model

(1) Calculation of Mass Transfer Coefficient

The dissolution rate can be expressed as Eq. (7) or (8) if the dissolution of NMA carbon brick is controlled by the carbon diffusion in molten iron:²⁵⁾   

$$n_D=\frac{{\rho }_{L}A}{1\mathrm{\hspace{0.17em} }200}{k}_D\left(C_s-C_b\right)$$(7)

  

$$\frac{d{C}_b}{dt}=\frac{A}{V}{k}_D\left(C_s-C_b\right)$$(8)

The amount of dissolved carbon should be balance with the increase in carbon content in molten iron:   

$$-A{\rho }_S\frac{dr}{dt}=\frac{{\rho }_{L}V}{100}\frac{d{C}_b}{dt}$$(9)

In the short time of slight change in carbon content, Eq. (10) can be obtained from Eqs. (8) and (9):   

$$k_D=100{\rho }_s\mathrm{\Delta }r/{\rho }_L(C_s-\overline{C_b})\mathrm{\Delta }t$$(10)

where n_D is the dissolution rate, mol·m⁻²·s⁻¹; ρ_L is the density of molten iron, g·m⁻³; A is the area of the solid/liquid interface, m²; k_D is the mass transfer coefficient of carbon in molten iron, m·s⁻¹; C_s is the saturated carbon content in molten iron [C_S]=1.34+2.54×10⁻³(T−273)−0.35[%P]+0.17[%Ti]−0.54[%S]+0.04[%Mn]−0.30[%Si], %; C_b is the carbon content in the bulk, %; t is time, s; V is the volume of molten iron, m³; ρ_S is the density of carbon brick, g·m⁻³; r is the radius of cylindrical specimen, m; Δr is the decrease in the radius, m; $\overline{C_b}$ is the arithmetic mean of initial and final carbon content in molten iron, %; Δt is the reaction time, s.

(2) Relationship between k_D and Carbon Content

Equation (7) can be accepted if k_D is independent of the concentration driving force. In order to confirm the assumption, the relationship between k_D and carbon content is presented in Fig. 9, k_D remained basically unchanged with the change of carbon content, it is proved that k_D is independent of carbon content and the concentration driving force, and the assumption is confirmed.

Fig. 9.

Relationship between k_D and carbon content.

(3) Relationship between k_D and Fluid Motion

The fluid motion around the rotating cylinder is complicated, the vortex around the cylinder is in the transition zone from laminar to turbulent. k_D is a function of several parameters expressed by Eq. (11), The relationship between dimensionless numbers is presented as Eq. (12):   

$$k_D=f\left(\mu ,\mathrm{\hspace{0.17em} }D,\mathrm{\hspace{0.17em} }U,\mathrm{\hspace{0.17em} }L\right)$$(11)

  

$$ShS{c}^{-b}=c{Re}^a$$(12)

where μ is the dynamic viscosity of molten iron, m²·s⁻¹; D is the diffusion coefficient, m²·s⁻¹; U is the circumferential velocity of the cylinder, m·s⁻¹; L is the characteristic size which refers to the diameter of the cylinder, m; Sh=k_DL/D, Sc=μ/D, Re=LU/μ; a, b, c are experimental constants, respectively, b=1/3 according to reference.²⁶⁾

3.4.2. Calculation of Mass Transfer Coefficient

The relationship between logk_D and logU is presented in Fig. 10(a), the linear relationship proves that the mass transfer process of carbon is the controlling step of the dissolution reaction. The logarithm of Eq. (12) is taken, the linear relationship is obtained based on the experimental data (Fig. 10(b)). The relationship between the dimensionless numbers is obtained:   

$$ShS{c}^{-1/3}=5.46{Re}^{3.01}$$(13)

Fig. 10.

Relationship between the dimensionless numbers.

The parameters of molten iron are summarized in Table 3.²⁷⁾ When Re<200, a positive deviation from the linear relationship was found by Kosaka et al.,²²⁾ they explained that the deviation was caused by a natural convection owing to the density change of molten iron after the dissolution. In the current study, the deviation of the results is very small due to the higher rotational speed in high Reynolds number region.

Table 3. The parameters of molten iron.

T, K	1573	1673	1773
μ×107, m2·s−1	6.5	5.0	4.0
D×108, m2·s−1	0.71	1.1	1.6
ρL×10−6, g·m−3	7.5	6.99	6.52

3.4.3. Effect of Phosphorus and Sulfur Content on The Reaction

(1) Effect of Phosphorus on The Reaction

When the phosphorus content up to 0.2% in molten iron, the relationship between the dimensionless numbers can be obtained, a negative deviation is found in Fig. 10(b).

The dot should coincide with the straight line if the dissolution of the sample is controlled only by the diffusion of carbon. The dissolution reaction may become mixed control caused by the adsorption of phosphorus on the interface.

The dissolution rate at the interface between iron and carbon:   

$$n_c=\frac{{\rho }_L}{1\mathrm{\hspace{0.17em} }200}A{k}_c\left(C_s-C_i\right)$$(14)

The mass transfer rate of carbon in molten iron:   

$$n_D=\frac{{\rho }_L}{1\mathrm{\hspace{0.17em} }200}A{k}_D\left(C_i-C_b\right)$$(15)

Equations (14) and (15) are added to eliminate C_i, the overall mass transfer rate of carbon is obtained:   

$$n=\frac{{\rho }_L}{1\mathrm{\hspace{0.17em} }200}Ak\left(C_s-C_b\right)$$(16)

where n_c is the dissolution rate at the interface, mol·s⁻¹; k_c is the reaction rate constant, m·s⁻¹; C_i is the carbon content at the interface, %; n_D is the mass transfer rate of carbon in molten iron, mol·s⁻¹; n is overall mass transfer rate of carbon, mol·s⁻¹; k is the overall mass transfer coefficient of carbon, m·s⁻¹.

Equation (17) is obtained from Eqs. (14), (15) and (16):   

$$1/k=1/k_c+1/k_D$$(17)

The value of k can be obtained from Eq. (10) with k_D replaced by k. Equation (7) can be used to calculate k due to the approximation of Eqs. (16) and (7). Take the reciprocal of Eq. (13):   

$$DS{c}^{1/3}/k_{D}L=0.183{Re}^{-3.01}$$(18)

By combining Eqs. (17) and (18):   

$$DS{c}^{1/3}/kL=DS{c}^{1/3}/k_{c}L+0.183{Re}^{-3.01}$$(19)

The linear relationship between DSc^1/3/kL and Re^−3.01 is found. k_c can be obtained by calculating the intercept of the line when the phosphorus content is 0.2% and the temperature is 1773 K (Fig. 11). The little difference between k_c and k_D demonstrate that the dissolution reaction of the sample is controlled by interfacial reaction and mass transfer of carbon.

Fig. 11.

Effect of phosphorus on the reaction.

(2) Effect of Sulfur on The Reaction

The same method can be used to analyze the effect of sulfur, the dissolution reaction of the sample is also controlled by interfacial reaction and mass transfer of carbon. The adsorption of sulfur on the iron-carbon interface covers part of the effective area. Considering the adsorption of sulfur, the following equilibrium exists at the interface of iron and carbon:   

$$K=a_{Sb}/a_{Sad}$$(20)

where K is the reaction equilibrium constant; a_sb is the activity of sulfur in molten iron; a_Sad the activity of sulfur at the interface.

Langmuir adsorption isotherm equation:   

$$\theta /\left(1-\theta \right)=K{\gamma }_s\left[mass\%S\right]$$(21)

where θ is the proportion of sites covered by sulfur; γ_s is the activity coefficient of sulfur, mass%⁻¹.

When $k_c^0$ is defined as the reaction rate without sulfur:   

$$k_c=\left(1-\theta \right)k_c^0$$(22)

Equation (23) is obtained from Eqs. (17), (21) and (22):   

$$1/k=\left(1/k_D+1/k_c^0\right)+\left(K{\gamma }_s/k_c^0\right)\left[mass\%S\right]$$(23)

The linear relationship between 1/k and sulfur content is found, 1/$k_c^0$ and K can be obtained from the slope and intercept of the line. The value of $k_c^0$ calculated at 1773 K is 9×10⁻⁴ m·s⁻¹, the value of θ increases with the higher sulfur content (Fig. 12), the degree of adsorption of sulfur on the surface depends on θ.

Fig. 12.

Effect of sulfur on the reaction.

4. Conclusions

In order to investigate the dissolution mechanism of carbon brick into molten iron, cylindrical specimens were immersed into molten iron to carry out the experiments, the calculation model of mass transfer coefficient was established, the fallowing results were found:

(1) The dissolution reaction of carbon was considered as the dominant reaction through thermodynamic analysis. The result of SEM revealed the hole diameter decrease from the reaction interface to the center position of carbon brick, CO gas generated by the reaction diffused along the hole at the reaction interface in carbon brick result in the hole diameter near the reaction interface is larger.

(2) The quantitative relationship between element content and erosion was obtained through the experimental results, carbon, silicon and titanium decreased the erosion rate, the order of influence degree: carbon > silicon > titanium, whereas manganese, phosphorus and sulfur increased the erosion rate, the order of influence degree: sulfur > phosphorus > manganese.

(3) The dissolution reaction of the sample is controlled by interfacial reaction and mass transfer of carbon when the phosphorus content up to 0.2% in molten iron. The adsorption of sulfur on the iron-carbon interface covers part of the effective surface, the degree of adsorption on the interface depends on the proportion of sites covered by sulfur.

Acknowledgments

This work was financially supported by the National Science Foundation for Young Scientists of China (51704019), Major Science and Technology Program for Water Pollution Control and Treatment (2017ZX07402001), supported by the Fundamental Research Funds for the Central Universities (FRF-BD-17-010A) and (FRF-TP-17-040A1).

References

• 1)   T. J.  Yang: Ironmaking, 33 (2014), 1.
• 2)   Z. J.  Liu,  J. L.  Zhang and  T. J.  Yang: ISIJ Int., 55 (2015), 1146.
• 3)   K.  Takatani,  T.  Inada and  K.  Takata: ISIJ Int., 41 (2001), 1139.
• 4)   K. C.  Zheng,  Z.  Wen,  X. L.  Liu,  Y. Q.  Ren,  W. F.  Wu and  H. B.  Qiu: ISIJ Int., 49 (2009), 1277.
• 5)   F. M.  Zhang: J. Iron Steel Res. Int., 20 (2013), 53.
• 6)   S. R.  Zhang: 5th Int. Cong. on Science and Technology of Ironmaking (ICSTT09), Chinese Society of Metals (CSM), Beijing, (2009), 13.
• 7)   Z. J.  Liu,  J. L.  Zhang,  T. J.  Yang and  H. B.  Zuo: ISIJ Int., 52 (2012), 1713.
• 8)   V. O.  Dahlke and  O.  Knacke: Arch. Eisenhuttenwes., 26 (1955), 373.
• 9)   R.  Krzeszewski: Gie Berei, 46 (1959), 417.
• 10)   Ch.  Hisatsuna,  T.  Shinodd and  Sh.  Sae: Imonology, 37 (1965), 27.
• 11)   C. F.  Loper,  S. L.  Liu,  S.  Shirvani and  T. W.  Witter: Trans. AFS, 92 (1984), 323.
• 12)   S. T.  Cham,  R.  Sakurovs,  H.  Sun and  V.  Sahajwalla: ISIJ Int., 46 (2006), 652.
• 13)   Y.  Shigeno,  M.  Tokuda and  M.  Ohtani: Trans. JIM, 26 (1985), 33.
• 14)   M. B.  Mourao,  G. G.  Krishnamurthy and  J. F.  Elliot: Metall. Trans. B, 24B (1993), 629.
• 15)   S.  Orsten and  F.  Oeters: Proc. 7th Process Technology Conf., Vol. 7, ISS, Warrendale, PA, (1988), 31.
• 16)   N.  Kayama and  O.  Ishikawa: Imonology, 37 (1965), 278.
• 17)   J. K.  Wright and  B. R.  Baldock: Metall. Trans. B, 19B (1988), 375.
• 18)   L.  Kalvelage,  J.  Markert and  J.  Poetschke: Arch. Eisenhuttnwes., 50 (1979), 107.
• 19)   S.  Orsten and  F.  Oeters: Proc. 5th Int. Iron and Steel Cong. on Process Technology, Book 3, ISS, Warrendale, PA, (1986), 143.
• 20)   O.  Dahlke and  O.  Knacke: Arch. Eisenhüttenwes., 7 (1955), 374.
• 21)   R. G.  Olsson,  V.  Koump and  T. F.  Perzak: Trans. Metall. Soc. AIME, 236 (1996), 426.
• 22)   M.  Kosaka and  S.  Minowa: Tetsu-to-Hagané, 53 (1967), 1468.
• 23)   S.-O.  Ericsson and  P.-O.  Mellberg: Scand. J. Metall., 10 (1981), 15.
• 24)   P. M.  Shurygin and  V. I.  Kryk: Izv. Vyssh. Uchebn. Zaved., Chern. Metall., 6 (1963), 14.
• 25)   Y.  Shigeno,  M.  Tokuda and  M.  Ohtani: J. Jpn. Inst. Met., 46 (1982), 713.
• 26)   M.  Eisenberg,  C.  Tobias and  C.  Wilke: Chem. Eng. Prog. Symp. Ser., 51 (1955), 1.
• 27)   A.  Adachi and  Z.  Morita: Handbook of Physical Properties of Molten Iron and Slags, ISIJ, Tokyo, (1972), 8.