Acta Metall. Slovaca Conf. 165 HIGH TEMPERATURE RHEOMETRY OF MOLTEN AND SEMI-SOLID IONIC SOLUTIONS BEARING ALUMINIUM AND TITANIUM OXIDES Piotr Migas 1)*, Marta Korolczuk-Hejnak 1) 1) Department of Ferrous Metallurgy, Faculty of Metals Engineering and Industrial Computer Science, AGH-University of Science and Technology, Poland Received: 10.12.2013 Accepted: 18.04.2014 * Corresponding author: e-mail: pmigas@agh.edu.pl, Tel.:+48 12 617 38 08, Department of Ferrous Metallurgy, Faculty of Metals Engineering and Industrial Computer Science, AGH- University of Science and Technology, 30 Mickiewicza Av., 30-059 Krakow, Poland Abstract Rheological tests of molten and semi-solid ionic solutions are difficult and demanding from the point of view of instrumentation and testing methodology. Reports that do exist in references often contain measured values of dynamic viscosity coefficient of analysed slag systems. However, these are based on the assumption that the analysed systems are liquids similar to what would be regarded as the Newtonian perfect body, and few papers contain analyses of viscosity problems from the standpoint of rheological characteristics of the systems examined. So the authors of this article have undertaken the subject of rheological characteristics of liquid ionic solutions of the blast furnace type. A prototype, high-temperature, rotary, multi-point viscometer with a rotating internal cylinder has been used for these tests. The article presents selected results of tests and rheological analyses of completely molten and semi-solid slag systems with additions of TiO 2 and Al 2 O 3 within a temperature range of 1310-1490 o C. Keywords: viscosity, blast furnace slags, rheology 1 Introduction The hearth and bottom lining of the blast furnace is one of the main factors that enables blast furnace operations to continue without interruption. The coefficient of dynamic viscosity of actual iron and BF slag systems has been monitored for a considerable period, and any deviation of viscosity from the standards set by the practice would result in effects deemed undesirable in the blast furnace operation [1-10]. Usage of burden materials with titanium compounds have begun in order to extend the life of the refractory lining and campaine of the blast furnace, with a very high thermal stability of the produced compounds that ensures effective protection of the lining. However, the use of titanium materials and the processes of formation of titanium compounds directly influence the physical and chemical properties of slag and iron in the blast furnace, and consequently the process, its productivity and economy. The rheological properties of liquid and semi-solid products of the process are important for the correct charge material movement mechanics and for the correct flow of reduction gases through the coke layers (cohesive zone) in the blast furnace. Based on the conducted tests of industrial slags that contained high concentrations of TiO 2, the authors [1] have found that regardless of the FeO content, slags are fluid liquids and their viscosity is almost constant even when there is an increase in temperature. The content of the solids in the system volume is the main factor influencing the viscosity of the slags.
Acta Metall. Slovaca Conf. 166 Many liquids, including most liquid synthetic and industrial slags, show a dependency of viscosity on the temperature in accordance with the Arrhenius-Guzman equation [2]: Eη R T η = A e (1.) where: E η [ev] - activation energy of viscous flow, A, R [-] experimental indexes and constants, T [K] - temperature. The authors [3] also have found that with the addition of 20% Ti 2 O 3 to the slag, the activation energy decreases by around 57% when compared to the initial value. In contrast, for TiO 2 this decrease is at a level of 27%. These obtained graphs have irregular shapes and are not rectilinear dependencies. The slags of this type do not show the Arrhenius dependency, which may reflect a continuous change in their internal structure along with the temperature change. The precipitation of finely dispersed solids in the system volume is another possibility as this changes the nature of flow and the nature of the liquid [2]. After analysis, the authors [4] advanced a thesis that the viscosity depends largely on the quantity, size and shape of the solid particles and free ions presented in the system. In addition, the degree of polymerisation of the slag influences the viscosity value. According to the authors [4], the coefficient of dynamic viscosity of slag is not affected by any changes in the rotational speed of the use bob. The authors [5] embarked on studies on creating a model determining the dynamic viscosity indexes of slags with the additions of high titanium contents in a range of 0-40%. It was found that the addition of TiO 2 and Ti 2 O 3 causes a decline in viscosity for a given temperature, although this concerns temperatures above the liquidus. The occurrence of solids in the slag system at temperatures under the liquidus causes a rapid increase in viscosity, whereas the addition of TiO 2 and Ti 2 O 3 in siliceous liquid results in a decrease. According to the authors [5] it is reasonable to assume that titanium compounds act largely as network modifiers. Few research canters [6, 7] have tackled such a complex subject of analysis of changes in the dynamic viscosity coefficient from the perspective of rheological properties. But it seems reasonable to define rheological nature of slags not only on the basis of their chemical composition and temperature, but also by rheological parameters: the time of force impact on the system, shear stress, shear rate [8]. To describe viscosity of slags containing a solid phase, the Einstein-Roscoe equation is most often used. The following equation may be used for the estimation of viscosity of slags containing up to 30% of the solid fraction within the system volume: S L n S η = η (1 RΦ ) (2.) where: η S [Pas] general apparent viscosity index (solids and liquid), η L [Pas] viscosity of residual liquid phase, Φ S [-] - volume fraction of solids, R,n [-] experimental constants. For spherical particles of equal size the constants R and n in the equation are 1.35 and 2.5 respectively. The inverse of the R value has physical meaning of the maximum amount of the solid phase that can be accommodated by the liquid before the viscosity has reached an infinitely large value. Viscosity is a measure of flow capability when a shear stress is applied. Most fully
Acta Metall. Slovaca Conf. 167 fluid slags and metallic liquids show properties of Newtonian liquids [9]. Therefore, their rheological charakter is defined by the Newton equation as a constant of proportionality between the shear stress and the shear rate: η τ & γ = (3.) where: η [Pas] - dynamic viscosity coefficient, τ [Pa] - shear stress, γ& [s -1 ] - shear rate. Ionic liquids within certain ranges of chemical composition, temperature and existing strains show the properties of pseudo Newtonian liquids, for example, viscoelastic body [8]. In order to verify the Einstein-Roscoe equation, A. Kondratiev et al. [11] conducted tests of four various, partially crystallised, triple slag systems. Among others, such slag system: Al 2 O 3 -FeO-SiO 2 during cooling from 1773 to 1633K and heating from 1633 to 1723K. The parameters adjusted to the model are R=1.29 and n=2.04, and are comparable with the Roscoe model values. Despite considerable research on slag system viscosity, the data available in the literature is still too fragmentary and incomplete to be able to understand the structure and to anticipate the properties of viscosity of liquid and semi-solid slags that commonly occur in metallurgical processes. But high temperature rheometric tests do bring in new information and extend the knowledge of rheological properties and the internal structure of liquid slags. 2 Measurement apparatus Fig. 1 presents a schematic diagram of a high-temperature rheometer used for rheological tests of liquid metals and slags. Fig. 1 Scheme presenting a high-temperature rheometer FRS1600 [12,13] The main part of the FRS1600 rheometer is its head, along with its cooling system. The head, like the furnace, is computer controlled. A thermocouple for temperature measurement is placed in the furnace, and the tubular resistance furnace with a mullite tube is controlled by a Eurotherm controller. An inert gas - argon with purity of 5.0 - is inject into the furnace tube to enable a protective atmosphere to be maintained during rheological measurements that last several hours.
Acta Metall. Slovaca Conf. 168 Fig. 2 presents measurement tools/systems used for rheometric tests of liquid ionic solutions, fluxes, mould powders or glasses. In this case graphite was used as the material for measurement systems. Fig. 2 Measurements systems [14,15,16] To test slag systems described in this article, a system with a perforated side surface of the spindle was used in order to eliminate possible slippage between the spindle and the tested medium. 3 High-Temperature Rheometric Tests The rheological research focused on multi-component slags of the blast furnace type in the systems: CaO SiO 2 Al 2 O 3 MgO-TiO 2 and CaO SiO 2 Al 2 O 3 MgO. These slag were obtained by means of a synthesis of pure components in liquid form: CaO calcined powder, SiO 2 - analytically pure quartz, Al 2 O 3 analytically pure, MgO analytically pure calcined powder (made by a E.Merck company), TiO 2 - rutile. Before the components were weighed, they had been dried in the temperature of 120 C for 5 hours. Then they were carefully mixed. Slag was melted in an induction furnace (in a graphite crucible). The chemical composition of slag was analysed using an XFR spectrometer TWIN-X. The results of this analysis can be found in Table 1 and 2. Table 1 Slags samples bearing TiO 2 chemical compositions N o CaO MgO Al 2 O 3 SiO 2 TiO 2 Na 2 O K 2 O Sample [%] 1 43,29 7,35 11,03 32,91 4,37 0,13 0,221 2 37,12 7,70 11,12 29,81 12,80 0,15 0,141 3 35,04 8,46 11,44 25,13 18,76 0,13 0,240 4 33,51 8,34 10,20 25,40 21,74 0,06 0,224 5 32,05 8,13 9,30 21,95 27,93 0,03 0,221 Rutile was added to the base slags on the assumption that this would obtain titanium oxide concentration in the systems at a level of 6, 15, 21, 25, 30%. After conducting a rheological measurement, a chemical analysis was performed with a TWIN-X spectrometer; the chemical compositions are presented in Table 1. Another analysed slag type was a CaO SiO 2 Al 2 O 3 MgO system, with chemical compositions presented in Table 2. All analysed slags were systems with increased concentrations of Al 2 O 3 and TiO 2, which were above the levels used in the blast furnace process.
Acta Metall. Slovaca Conf. 169 Table 2 Slags samples bearing Al 2 O 3 chemical compositions N o CaO MgO Al 2 O 3 SiO 2 Sample [%] 1 44,23 6,46 7,07 42,24 2 41,61 6,07 12,76 39,55 3 40,34 5,88 15,60 38,18 4 37,66 5,49 21,29 35,56 5 35,56 5,18 25,65 33,61 Fig. 3 presents flow curves of the samples tested at the selected temperatures. The presented rheological characteristics reveal that at specific temperatures, and for specific concentrations of TiO 2 in the system, all samples tested are similar to the Newtonian ideal body. However, at a temperature of 1490 ºC, the curve shows deviations from what would be an perfect viscose fluid. Fig. 3 Flow curves for slags with an increased TiO 2 content In an almost completely liquid system in which only modest amounts of solids Ti(C, N) exist, titanium plays the role of a network modifier by diametrically decreasing the dynamic viscosity, and as a result the slag system shows small deviations from the perfect Newtonian body. However, when the temperature decreases, a substantial quantity of solids precipitate in the form of perovskite, intensively raising the coefficient of dynamic viscosity of the system but without any significant influence on its rheological nature. Fig. 4 presents flow curves for the temperatures of 1400, 1330 and 1310 o C for the selected samples. A change in the rheological nature of the liquid resulting from a temperature decrease - and thus a change in the system internal structure - can be observed. The amount and nature of the precipitated solids/particles influences the shown deviations from similarities to the Newtonian liquid.
Acta Metall. Slovaca Conf. 170 Fig. 4 Flow curves for slags with an increased Al 2 O 3 content It can be concluded from the nature of the flow curves that the analysed system with solids is not a Newtonian body. It may be supposed that this is a system that has been diluted by shearing. This effect is characteristic to polymer solutions or suspensions of solid particles in liquids. 4 Conslusions Tests conducted with an FRS1600 rheometer for two types of slag systems with a cross-linking agent - Al 2 O 3 and a network modifier - TiO 2, revealed that at measurement temperatures up to 1490 o C (and above 1400 o C), the addition of TiO 2 slightly changes the rheological nature of the analysed slag system. The liquids show deviations from the Newtonian pefect body. In particular, within temperatures of 1490 down to 1430 o C, samples 1 and 2 show minimum deviations from the Newtonian nature. However, for temperatures from 1400 to 1310 o C, the biggest deviations from the Newtonian nature of the system are revealed in samples 3-5. These show changes in the coefficient of viscosity in the same direction. A decrease in the value of the coefficient of dynamic viscosity was observed with an increase in the TiO 2 content in the system at temperatures up to 1490 o C, while for the lowest temperature (1310 o C) this tendency is the reverse - the highest viscosity is featured in slags with high TiO 2 concentrations, and this is consistent with the findings presented in world literature. For the liquid slag system CaO-SiO 2 -MgO-Al 2 O 3, at a temperature of 1400 o C the increase in the Al 2 O 3 content raises the value of the dynamic viscosity coefficient. The examined liquid ionic solutions are similar to a Newtonian ideally viscous liquid. In the analysed systems in which solid particles occur, the nature changes from a Newtonian liquid to a system diluted by shearing (pseudoplastic), in which the shear rate influences the value of the dynamic viscosity coefficient, and an increase in the shear rate causes a decrease in viscosity.
Acta Metall. Slovaca Conf. 171 References [1] G. Handfield, G.G. Charette: Canadian Metallurgical Quarterly, Vol. 10, 1971, No. 3, p. 235-243, DOI: http://dx.doi.org/10.1179/cmq.1971.10.3.235 [2] G. Handfield, G.G. Charette, H.Y. Lee: Titanium Bearing Ore And Blast Furnace Slag Viscosity, Journal Of Metals, 1972, p. 37 40 [3] S. Seetharaman K. Mukai, Du Sichen: Steel Research International, Vol. 76, 2005, No. 4, p. 267-278 [4] A. Shankar, Mårten Görnerup, A.K. Lahiri, And S. Seetharaman: Metallurgical and Materials Transactions B, Vol. 38, 2007, No. 6, p. 911-915, DOI: 10.1007/s11663-007- 9087-5 [5] L. Zhang, S. Jahanshahi: Modelling Viscosity Of TiO x Containing Silicate Melts, In.: International Conference On Molten Slags Fluxes And Salts, Johannesburg, The South African Institute Of Mining and Metallurgy, 2004, p. 51-56 [6] A. Kondratiev, P. C. Hayes And E. Jak.: ISIJ International, Vol. 48, 2008, No. 1, p. 7 16, DOI: 10.2355/isijinternational.48.7 [7] A. Kondratiev, E. Jak.: Metallurgical and Materials Transactions B, Vol. 32, 2001, No. 1, p. 1015-1025, DOI: 10.1007/s11663-001-0090-y [8] Y. Sasaki, H. Urata, M. Iguchi, M. Hino: ISIJ International, Vol. 46, 2006, No 3, p. 385 387, DOI: 10.2355/isijinternational.46.385 [9] N. N. Viswanathan, F.-Z. Ji, Du Sichen, S. Seetharaman: ISIJ International, Vol. 41, 2001, No. 7, p. 722 727, DOI: 10.2355/isijinternational.41.722 [10] Z. Kembłowski: Rheometry of non-newtonian fluids, WNT, Warszawa, 1973 (in Polish) [11] A. Kondratiev, P. C. Hayes And E. Jak: ISIJ International, Vol. 46, 2006, No. 3, p. 359 367, DOI: 10.2355/isijinternational.46.359 [12] K. Sołek, M. Korolczuk-Hejnak, W. Ślęzak: Archives of Metallurgy and Materials, Vol. 57, 2012, No. 1, p. 333 338, DOI: 10.2478/v10172-012-0031-6 [13] M. Korolczuk-Hejnak, P. Migas: Archives of Metallurgy and Materials, Vol. 57, 2012, No. 4, p. 963 969, DOI: 10.2478/v10172-012-0107-3 [14] M. Korolczuk-Hejnak, P. Migas: Archives of Metallurgy and Materials, Vol. 57, 2012, No. 2, p. 583 591, DOI: 10.2478/v10172-012-0062-z [15] P. Migas, M. Korolczuk-Hejnak, M. Karbowniczek: Hutnik, Vol. 79, No. 4, p. 196 202 [16] P. Migas, M. Korolczuk-Hejnak, M. Karbowniczek: Rheological analysis of blast furnace synthetic slag admixtures of Al 2 O 3, In.: Ninth International Conference on Molten Slags, Fluxes and Salts (Molten12), Beijing, China, 2012, p. 1 15 Acknowledgements Financial support of National Science Center for research project Developing an empirical model of the rheological properties of liquid metals on the example of iron solutions no. 2011/01/N/ST8/07368 is gratefully acknowledged. The research was conducted with the financial support from the statutory project no. 11.11.110.081 AGH University.