Abstract
In consideration of great need for the physicochemical properties of slag systems in metallurgical process estimation, this work examined the possibility of predicting multi-properties by one model. The mass triangle model was applied to evaluate the density, viscosity, surface tension and sulfide capacity of CaO–“FeO”–SiO2 system at 1,673 K. Good agreements were achieved between calculated data and experimental data in the various properties. Meanwhile, the calculated contour lines successfully predicted the properties of slag within the limited solubility area. The new model thus is competitive and flexible when an integrated knowledge of a certain system is necessary.
Introduction
Process estimation in metallurgical industry calls for a thorough understanding and an integrated control for slag properties, and metallurgical slag is supposed to offer both good dynamic conditions and adequate capacities to absorb impurities in ironmaking and steelmaking process. Thus, calculation of a certain slag system is a symmetric task and must be based on integrated assessments of various properties, including thermophysical and thermochemical properties.
To solve such problem, it is not possible to solely depend on the experimental works due to the huge amount and difficulties of measurements at high temperatures. Another choice might be to combine the experimental with the modeling. Therefore, various models have been proposed to evaluate different properties. Usually, researchers addressed a model for one property. In the case of viscosity, Riboud et al. [1] and Iida et al. [2] proposed equations for the viscosities on the basis of relationships between the network parameter and the reciprocal of the basicity index. Seetharaman [3, 4] reported a model based on the absolute reaction-rate theory, utilizing the Gibbs energies of mixing of the silicate melts. Zhang and Jahanshahi [5, 6] employed a function of bridging oxygen and free oxygen ions to elucidate the network structure and viscosity of molten slag. To evaluate the density of slag, Persson [7] found a relation between the molar volume of slag and the relative integral molar mixing enthalpy. Zhang [8] also derived an equation between density and optical basicity of slag. The optical basicity could also be used to evaluate the sulfide capacity of slag, as well, which was put forward by Sommerville [9] and developed by Young [10]. Researchers from KTH [11] developed a model to describe the sulfide capacity of slag on the basis of Temkin theory [12]. Reddy et al. [13] presented a model relating oxide activity to sulfide in binary system, then Pelton [14] extended this model to multicomponent system later. The same situation could be seen for surface tension, some were on basis of the additivity rule [15] and other models [16, 17, 18, 19, 20] were based on Butler’s equation [21]. Meanwhile, Nakamoto et al. [22] also utilized neural network computation to estimate the surface tension of molten silicate slags.
Even though, lacking of a model which can predict various properties for one system makes the optimization of industrial processes challenged. Therefore, in the present study, an attempt has been made to evaluate several properties by Mass triangle model, proposed by Zhou et al. [23]. In view of the importance of CaO–“FeO”–SiO2 system in ferrous and non-ferrous metallurgy as well, this system has been selected as an example to perform the examination.
Theoretical basis
To orient the author, the brief introduction of mass triangle model has been given in this section. The detailed information can be found elsewhere [24, 25]. For metallurgical slag, they are usually limited solubility system at certain temperature, which means no intersect point with binary systems. As shown in Figure 1, the area within the boundary represents a homogenous system. O is a composition point within this area. Three lines that parallel to the three sides of the triangle were drawn from O. They intersect with the sides at A, B and C and intersect with the soluble boundary at A’, B’ and C’. PA, PB, and PC’ represent the property data of point A’, B’ and C’ respectively. WA’, WB’, and WC’ refer to the weight factors of these three points. The weight factor of A is determined by area ratios represented in eq. (1).
When the property data on the boundary are known, then the property of point O within the area PO could be expressed as the weight summation of properties of A’, B’ and C’, which is shown in eq. (2).

Schematic graph of a ternary system with limited solubility.
Results and discussion
The limited solubility boundary of CaO–“FeO”–SiO2 system at 1,673 K was calculated by Factsage 6.2. Due to the utilization condition of this new model, the estimation of several properties in this area has been performed according to the corresponding data available in literature [26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 40, 41, 42, 43, 44, 45, 4647]. And the liquid line was also drawn in the figures which shown the calculated contour lines.
Density of CaO–“FeO”–SiO2 system
The density of slag is of great importance because it directly affects the volume of slag. Therefore, knowing the density of the slag can give instruction for the many processes, such as electroslag remelting, molten-salt electrolysis, etc. There were many researchers [26, 27, 2829] having reported reliable density data of this system.
Lee [26] studied the density and structure of CaO–“FeO”–SiO2 in the temperature range of 1,573 K~1,873 K by maximum bubble pressure method. The starting materials included wustite prepared by melting and reducing ferric oxide in iron crucibles. Meanwhile, argon gas was used as the bubbling gas. Based on his experimental data, densities of CaO–“FeO”–SiO2 at 1,673 K were calculated by the model described above. The results are shown in Figure 2.

Densities (g/cm3) of CaO–“FeO”–SiO2 system at 1,673 K.
As discussed by Lee, when the mole fraction of SiO2 increased, the average silicate anion size would increase as well. As shown in the figure, with increasing content of SiO2, the density value got decreased. That was identical with the atmosphere presented by Lee. Meanwhile, Lee also pointed out that the calcium silicate melts were less polymerized than the corresponding iron silicates. Again, the calculation gave the same trend, that is, the increasing content of “FeO” elevated the density, while the effect of CaO content to the density is not very obvious.
In order to measure the accuracy of this new model, two experimental data were taken to compare the calculation results, marked by a circle and a cross. The deviation is defined by eq. (3). Table 1 showed the calculation error rates of these two data were –0.346 % and 1.003 %, respectively.
It is demonstrated that the new model successfully predicted the trend of density change with different composition. When the composition is fixed, the model can also calculate the property using the boundary point with a lower deviation.
Viscosity
Viscosity is one of the most important properties in both ironmaking and steelmaking, in that slag reacted with both metal and gas. If the viscosity is too high, the slag cannot absorb detrimental element and gas in unable to get through the slag smoothly. However, if the viscosity is too low, foaming is likely to happen, and it could cause serious difficulty and harm. Therefore, viscosity is also a basic factor that is taken to consideration in process modeling. Previous papers [30, 31, 32, 33, 34, 35] made a great contribution to understanding the viscosity of CaO–“FeO”–SiO2 system.
Ji [30] reported his experimental study on viscosity of CaO–“FeO”–SiO2 system in temperature range of 1,423 K–1,753 K. In his experiment, iron crucible was employed to both mix the slag and measure the viscosity. And the oxygen partial pressure was kept between 2.5×10–10 and 1×10–6 atm. Based on the experimental data, the viscosity contour lines were obtained, shown in Figure 3. Generally, the viscosity increased as the SiO2 content increased. This was because the SiO2 always acted as a network former, which would increase viscosity of slag. On the contrary, increasing content of “FeO” and CaO will decrease the viscosity, due to their function of a network modifier. Ji also indicated that replacement of SiO2 by “FeO” caused a dramatically decrease in viscosity. The experimental data reported by Gudenau [35] was slightly higher than the calculation, partly due to the slag containing ferric oxide.

Viscosities(Pa × s) of CaO–“FeO”–SiO2 system at 1,673 K.
Surface tension
Information on the surface tension of slags is related to the interface behavior of slag phase and has combined effect with other properties. In sintering process, the surface tension plays an important role in pelleting of raw materials. In ironmaking and steelmaking, the decrease of ratio of surface tension to viscosity is necessary to obtain a stable foaming slag. In refining and continuous casting, this property affects the absorption capacities to inclusions and erosion to the refractory and mold channels. Valuable data can be found in previous work [36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46]. Lee et al. [43] studied the surface tension of CaO–“FeO”–SiO2 system in different temperatures and based on his experimental data, iso-surface tension lines were calculated, presented in Figure 4.

Surface tensions (mN/m) of CaO–“FeO”–SiO2 system at 1,673 K.
As can be seen, the calculation results satisfactorily predicted trend of surface tension with altered composition. The increasing content of CaO severely elevated the surface tension of this system while the increasing content of SiO2 slightly lowered the surface tension. However, the effect of “FeO” content was not as notable as other oxides. The point marked by a dot was calculated to have an error rate of 1.04 %, as presented in Table 1. Therefore, the new model can also estimate the surface tension of CaO–“FeO”–SiO2 system successfully.
Sulfide capacity
Sulfur removal in ironmaking and steelmaking processes is now of even more importance than in the past because of the increasing demand for extremely clean steel. Sulfide capacity is regarded as the capacity of a slag to hold sulfur, which is of great importance to the industry application. As for the system concerned, Nzotta et al. [47] have studied the sulfide capacity of CaO–“FeO”–SiO2 system at 1,673 K and 1,773 K by gas–slag equilibrium. FeO used in his studied was examined by X-ray Diffraction for confirmation. Thus, these data obtained by experiment work were used as “boundary data”, and then the calculation performed. The sulfide contours are exhibited in Figure 5.

Sulfide capacities in logarithm of CaO–“FeO”–SiO2 system at 1,673 K.
As seen in the figure, the increase of SiO2 content considerably hindered the desulfurization capacity. That was because SiO2 was an acid oxide which could not offer much O2– so it had a lower affinity to sulfur. Generally, increasing content of “FeO” and CaO favored desulfurization. However, the effect of “FeO” and CaO could not be distinguished easily. Nzotta also mentioned the marginal difference observed in CS values between “FeO”–SiO2 slag and CaO–SiO2 slag. Comparison between the reference data and the calculated data was shown in Table 1. An error rate of –7.3 % was obtained when evaluating sulfide capacity of the point marked by a round dot. Similarly, the new model worked well with the sulfide capacities of CaO– “FeO” –SiO2 both in contour lines derivation and calculation for fixed composition.
Comparison between reference data and calculated data.
Property | Composition (mass%) | Reference data | Calculated data | Err. % | ||
---|---|---|---|---|---|---|
CaO | SiO2 | “FeO” | ||||
Density | 0.211 | 0.314 | 0.475 | ρ=3.18 g/cm3 | ρ=3.169 g/cm3 | −0.346 |
0.281 | 0.331 | 0.389 | ρ=3.09 g/cm3 | ρ=3.121 g/cm3 | 1.003 | |
Surface tension | 0.207 | 0.320 | 0.473 | σ=385 mN/m | σ=389 mN/m | 1.04 |
Sulfide capacity | 0.303 | 0.403 | 0.294 | logCS=–3 | logCS=–2.78 | −7.3 |
Conclusion
To gain a better estimation of thermophysical and thermochemical properties, it is necessary to make modeling according to the limited experimental data. Despite the widely development of various models, efforts for using one model to predict various properties for a certain system are very limited.
In this work, mass triangle model was employed to evaluate the density, viscosity, surface tension and sulfide capacities of CaO–“FeO” –SiO2 system at 1,673 K. The calculated contour lines revealed the reasonable trend as experimental results, and calculation results for fixed compositions are very close to the experimental data within a low error rate. Thus, the results have proved a good reliability and flexibility of the mass triangle model. This model might open a door to optimize multi-properties of one system integrally for industry purposes.
Funding statement: The authors would like to express their appreciation to National Nature Science Foundation of China (No. 51104013, No. 51174022), and SRF for ROCS, SEM (No. 44) as well as the state key laboratory of advanced metallurgy for their financial supports.
References
[1] P.V. Riboud, Y. Roux, L.D. Lucas and H. Gaye, Fachber. Hüttenprax. Metallweiter-Verarb., 19 (1981) 859–869.Search in Google Scholar
[2] T. Iida, H. Sakai, Y. Kita and K. Murakami, High Temp. Mater. Proc., 19 (2000) 153–164.10.1515/HTMP.2000.19.3-4.153Search in Google Scholar
[3] S. Du, J. Bygdén and S. Seetharaman, Metall. Mater. Trans. B, 25B (1994) 519–525.10.1007/BF02662770Search in Google Scholar
[4] S. Seetharaman, S. Du and F.-Z. Ji, Metall. Mater. Trans. B, 31 (2000) 105–109.10.1007/s11663-000-0135-7Search in Google Scholar
[5] L. Zhang and S. Jahanshahi, Metall. Mater. Trans. B, 29 (1998) 177–186.10.1007/s11663-998-0020-3Search in Google Scholar
[6] L. Zhang and S. Jahanshahi, Metall. Mater. Trans. B, 29 (1998) 187–195.10.1007/s11663-998-0021-2Search in Google Scholar
[7] M. Persson, Investigations of Slag Properties and Reactions, KTH Royal Institute of Technology, Sweden (2007).Search in Google Scholar
[8] G.H. Zhang and K.C. Chou, J. Iron and Steel Res. Int., 17 (2010) 1–4.Search in Google Scholar
[9] D.J. Sosinsky and I.D. Sommerville, Metall. Trans. B, 17 (1986) 331–337.10.1007/BF02655080Search in Google Scholar
[10] R.W. Young, J.A. Duffy and G.J. Hassall, Ironmaking & Steelmaking, 19 (1992) 201–219.Search in Google Scholar
[11] R. Nilsson, S. Du and S. Seetharaman, Scand. J. Metall., 25 (1996) 128–134.Search in Google Scholar
[12] M. Temkin, Acta Physicochim. URSS, 20 (1945) 411–420.Search in Google Scholar
[13] R.G. Reddy and M. Blander, Metall. Trans. B, 18 (1987) 591–596.10.1007/BF02654272Search in Google Scholar
[14] A.D. Pelton, G. Eriksson and A. Romero-Serrano, Metall. Trans. B, 24 (1993) 817–825.10.1007/BF02663142Search in Google Scholar
[15] R.E. Boni and G. Derge, J. Metals, 8 (1956) 53–59.10.1007/BF03377643Search in Google Scholar
[16] J.Y. Choi and H.G. Lee, ISIJ Int., 42 (2002) 221–228.10.2355/isijinternational.42.221Search in Google Scholar
[17] T. Tanaka, K. Hack, T. Iida and S. Hara, Zeitschrift Für Metallkunde, 87 (1996) 380–389.10.1515/ijmr-1996-870509Search in Google Scholar
[18] T. Tanaka, T. Kitamura and I.A. Back, ISIJ Int., 46 (2006) 400–406.10.2355/isijinternational.46.400Search in Google Scholar
[19] M. Nakamoto, A. Kiyose, T. Tanaka and L. Holappa, ISIJ Int., 47 (2007) 38–43.10.2355/isijinternational.47.38Search in Google Scholar
[20] M. Nakamoto, T. Tanaka, L. Holappa and M. Hämäläinen, ISIJ Int., 47 (2007) 211–216.10.2355/isijinternational.47.211Search in Google Scholar
[21] J.A.V. Butler, Proc. R. Soc. London. Ser. A, 135 (1932) 348–375.10.1098/rspa.1932.0040Search in Google Scholar
[22] M. Nakamoto, M. Hanao, T. Tanaka and M. Kawamoto, ISIJ Int., 47 (2007) 1075–1081.10.2355/isijinternational.47.1075Search in Google Scholar
[23] K.C. Chou, Calphad, 19 (1995) 315–325.10.1016/0364-5916(95)00029-ESearch in Google Scholar
[24] L.J. Wang, K.C. Chou and S. Seetharaman, Calphad, 32 (2008) 49–55.10.1016/j.calphad.2007.11.006Search in Google Scholar
[25] M. Hino and T. Fuwa, Sulfide Capacity of Liquid Manganese Silicate, 3rd Inter. Iron Steel Congress (1978) 321–326.Search in Google Scholar
[26] Y.E. Lee and D.R. Gaskell, Metall. Trans., 5 (1974) 853–860.10.1007/BF02643138Search in Google Scholar
[27] S.I. Popel and O.A. Esin, Measurement of the density of fused silicates, Soviet Res. Fused Salts, (1956) 255.Search in Google Scholar
[28] K.C. Mills and B.J. Keene, Inter. Mater. Rev., 32 (1987) 1–120.10.1179/095066087790150296Search in Google Scholar
[29] D.R. Gaskell, A. McLean and R.G. Ward, Trans. Faraday Soc., 65 (1969) 1498–1508.10.1039/tf9696501498Search in Google Scholar
[30] F.-Z. Ji, S. Du and S. Seetharaman, Metall. Mater. Trans. B, 28 (1997) 827–834.10.1007/s11663-997-0010-xSearch in Google Scholar
[31] P. Kozakevitch, Revue De Metallurgie, 46 (1949) 505–516.10.1051/metal/194946080505Search in Google Scholar
[32] P. Williams, M. Sunderland and G. Briggs, Trans. Inst. Mining Metall., Sec. C, 92 (1983) 105–109.Search in Google Scholar
[33] L. Bodnar, Š. Cempa, K. Tomášek and L. Bobok, Chem. Papers, 32 (1978) 798–809.Search in Google Scholar
[34] K. Seki and F. Oeters, Trans. Iron Steel Inst. Jpn., 24 (1984) 445–454.10.2355/isijinternational1966.24.445Search in Google Scholar
[35] H.W. Gudenau and J. Petry, TIZ Fachber, 105 (1981) 552.Search in Google Scholar
[36] K.C. Mills, Slag Atlas, VDEh 2nd ed., Verlag Stahleisen GmbH, Düsseldorf (1995).Search in Google Scholar
[37] K. Tomosek, L. Bodnar and J. Holeczy, Hutn. Listy, 29 (1974) 882. (from Ref 36, P. 439).Search in Google Scholar
[38] E.A. Yamshchikova, L.V. Butuzova and I.A. Montil’O, Tr. Ural’sk. Nauch-Issled. Proekt. Inst. Mednoi. Prom., 12 (1969) 163. (from Ref 36, P. 439).Search in Google Scholar
[39] V. Vanyukov and N.I. Utkin, Dokl. Nauk. Vys. Shkoly. Met., 3 (1958) 22. (from Ref 36, P. 439).Search in Google Scholar
[40] K.K. Mori, Y. Shiraishi and N. Yamada, Tetsu-to-Hagané, 62 (1976) 53. (from Ref 36, P. 439).10.2355/tetsutohagane1955.62.1_53Search in Google Scholar
[41] P. Kozakevitch, Rev. Metall., 46 (1949) 572. (from Ref 36, P. 439 and 443).10.1051/metal/194946090572Search in Google Scholar
[42] S.I. Popel and O.A. Esin, Zh. Fiz. Khim., 30 (1956) 1193. (from Ref 36, P. 439).Search in Google Scholar
[43] Y.E. Lee and D.R. Gaskell, Trans. Iron Steel Inst. Jpn., 11 (1971) 564. (from Ref 36, P. 439).Search in Google Scholar
[44] S.I. Popel, V.I. Sokolov and V.G. Korpachev, Sb. Nauch. Tr. Ural’sk, Politekhn, Inst., 126 (1963) 24. (from Ref 36, P. 439).Search in Google Scholar
[45] A.V. Vanyukov and N.I. Utkin, Izv. V.U.Z. Tsvetn. Metall., 6 (1958) 43. (from Ref 36, P. 439).Search in Google Scholar
[46] D. Skupien and D.R. Gaskell, Metall. Mater. Trans. B, 31 (2000) 921–925.10.1007/s11663-000-0068-1Search in Google Scholar
[47] M.M. Nzotta, S. Du and S. Seetharaman, Metall. Mater. Trans. B, 30 (1999) 909–920.10.1007/s11663-999-0096-4Search in Google Scholar
© 2017 Walter de Gruyter GmbH, Berlin/Boston
This article is distributed under the terms of the Creative Commons Attribution Non-Commercial License, which permits unrestricted non-commercial use, distribution, and reproduction in any medium, provided the original work is properly cited.
Articles in the same Issue
- Frontmatter
- Research Articles
- Estimation for Iron Redox Equilibria in Multicomponent Slags
- The Effect of Multi-inclined Holes on the Creep Properties of Nickel-Based Superalloy
- Estimation of Various Properties of CaO–“FeO”–SiO2 System at 1,673 K by Mass Triangle Model
- The Enhancing Effect of Microwave Irradiation and Ultrasonic Wave on the Recovery of Zinc Sulfide Ores
- The Self-assembled Deposition on the Surface of Mono-crystalline Silicon Induced by High-Current Pulsed Electron Beam
- Numerical Model of Dephosphorization Reaction Kinetics in Top Blown Converter Coupled with Flow Field
- Morphological Evolution of Low-Grade Silica Fume at Elevated Temperature
- Discussion of Carbon Emissions for Charging Hot Metal in EAF Steelmaking Process
- Predictive Models for Modulus of Rupture and Modulus of Elasticity of Particleboard Manufactured in Different Pressing Conditions
- Photoluminescence Properties of Eu3+-activated Silicate Phosphors
- Synthesis, Acidity and Catalytic of the Rare Earth Ce Loaded on the Composite Pore Zeolite Catalyst for Hydrogenation Cracking
Articles in the same Issue
- Frontmatter
- Research Articles
- Estimation for Iron Redox Equilibria in Multicomponent Slags
- The Effect of Multi-inclined Holes on the Creep Properties of Nickel-Based Superalloy
- Estimation of Various Properties of CaO–“FeO”–SiO2 System at 1,673 K by Mass Triangle Model
- The Enhancing Effect of Microwave Irradiation and Ultrasonic Wave on the Recovery of Zinc Sulfide Ores
- The Self-assembled Deposition on the Surface of Mono-crystalline Silicon Induced by High-Current Pulsed Electron Beam
- Numerical Model of Dephosphorization Reaction Kinetics in Top Blown Converter Coupled with Flow Field
- Morphological Evolution of Low-Grade Silica Fume at Elevated Temperature
- Discussion of Carbon Emissions for Charging Hot Metal in EAF Steelmaking Process
- Predictive Models for Modulus of Rupture and Modulus of Elasticity of Particleboard Manufactured in Different Pressing Conditions
- Photoluminescence Properties of Eu3+-activated Silicate Phosphors
- Synthesis, Acidity and Catalytic of the Rare Earth Ce Loaded on the Composite Pore Zeolite Catalyst for Hydrogenation Cracking