COMPARISON OF THERMAL ANALYSIS RESULTS WITH THE THEORETICAL DETERMINATION OF SOLIDUS AND LIQUIDUS TEMPERATURES FOR SPECIFIC STEEL GRADE

Transcription

1 COMPARISON OF THERMAL ANALYSIS RESULTS WITH THE THEORETICAL DETERMINATION OF SOLIDUS AND LIQUIDUS TEMPERATURES FOR SPECIFIC STEEL GRADE Karel GRYC a, Bedřich SMETANA b, Markéta TKADLEČKOVÁ a, Monika ŽALUDOVÁ b, Ladislav SOCHA a, Karel MICHALEK a, Ladislav VÁLEK c a VŠB Technical University of Ostrava, Faculty of Metallurgy and Materials Engineering, Department of Metallurgy and Foundry, and Regional Materials Science and Technology Centre, Czech Republic, karel.gryc@vsb.cz, marketa.tkadleckova@vsb.cz, ladislav.socha@vsb.cz, karel.michalek@vsb.cz, b VŠB Technical University of Ostrava, Faculty of Metallurgy and Materials Engineering, Department of Physical Chemistry and Theory of Technological Processes, and Regional Materials Science and Technology Centre, Czech Republic, bedrich.smetana@vsb.cz, monika.zaludova@vsb.cz, jana.dobrovska@vsb.cz c ArcelorMittal Ostrava a.s., Research, Ostrava, Czech Republic, ladislav.valek@arcelormittal.com, radim.pachlopnik@arcelormittal.com Abstract Knowledge of solidus and liquidus temperature is especially critical for correct setting of processes associated with the production and especially casting of steel. Generally, steel is multicomponent melt containing a wide range of non-metallic phases. So, accurate determination of these temperatures is a considerable difficult. Nowadays, there are many ways to detect these temperatures. The utilisation of a wide range of existing empirically defined or thermo dynamical calculations are some of them. In addition to individual patterns reported in various literature and often focusing on specific steel grade, there are also sophisticated and often universal thermodynamic database in the form of professional software packages. This paper is devoted to the comparison of theoretically defined solidus and liquidus temperatures based on above mentioned theoretical methods with the results of high-temperature thermal analysis carried out on two devices (Netzsch STA 449 F3 Jupiter; Setaram SETSYS 18 TM ). On large samples (23 g), the method of direct thermal analysis was applied. On small samples (200 mg), the experiments using differential thermal analysis were realized. It was found out that the solidus and liquidus temperatures for the studied steel grade can vary in dependence on the used determination method. The used methodology of thermal analysis is fully reproducible, and the these thermo-analytical results can be considered as necessary for the correct setting of critical parameters in applied research on the process of steel casting. Keywords: steel, solidus temperature, liquidus temperature, thermal analysis, calculations 1. INTRODUCTION Steel production is one of the most important industries. Optimization of any crucial parameters of metallurgical technology should leads to significant lowering of production cost and to better quality of steel products. It is necessary to succeed in tough competition. In the refining processes, optimizing the slag regimes [1], thermal and chemical homogenization of the melt [2] or filtration of steel [3] is very important to solve. Works toward optimizing the process of solidification of heavy forging ingots [4] are currently being implemented in the casting and solidification of steel studies.

2 20 mm , Brno, Czech Republic, EU The methods of study of metallurgical processes are also based on knowledge of thermodynamic properties of materials occurring in a given technology nodes. Knowledge of solidus and liquidus temperatures of the studied steels is one of the most important factors - especially in dealing with the processes involved in the casting and solidification. These temperatures are critical parameters for proper adjustment of models (physical or numerical) or in the final stage of applied research of the real process. It is significantly affecting the final quality of the as-cast steel (billets or ingots). Therefore, this paper is devoted to discussion of findings obtained during the utilization of dynamic thermal analysis methods [5-8] to identify the solidus and liquidus temperatures of selected steel grade. Generally, it is not so easy to identify the phase transformations occurring in such multicomponent systems like steels [9-12]. 2. USED THERMO-ANALYTICAL METHODS AND ANALYSED STEEL New Laboratory for Modelling of Processes in the Liquid and Solid Phases within the project Regional Materials Science and Technology Centre (RMSTC) was formed at the Faculty of Metallurgy and Materials Engineering at the VŠB-Technical University of Ostrava in Czech Republic. This Laboratory has also acquired new equipment for high-temperature thermal analysis Netzsch STA 449 F3 Jupiter (Fig.1). The conditions for initiation of intensive research activities in the field of dynamic thermal analysis methods for steel are based on years of experience of team members with the issue of laboratory studies of metallurgical processes and the ability to use other equipment of this type Setaram SETSYS 18 TM (Fig.2) create. Fig.1 Netzsch STA 449 F3 Jupiter Fig. 2 Setaram SETSYS 18TM This paper discusses methods and results of thermal analysis of samples (Table 1) taken from the grain oriented electrical steel, with very low carbon content (0.04 wt. %) and high content of [Si] = 3 wt. %. Table 1 Steel samples with dimensions specified for each analysis Sample for NETZSCH STA 449 F3 Jupiter method: SETARAM Setsys 18 TM Ø14 mm Dimensions:

3 Two methods for dynamic thermal analysis were used to measure the solidus (TS) and liquidus (TL) temperatures: Differential Thermal Analysis (DTA) Setaram SETSYS 18 TM, Direct Thermal Analysis - Netzsch STA 449 F3 Jupiter. The principles of both methods are described for example in [13]. Both above mentioned dynamic thermal analysis methods were used to determine the TS and TL close to equilibrium in the frame of studied grain oriented electrical steel grade. Three direct thermal analysis was realized on the big steel samples (about 23 g) and the TS and TL were obtained by analysing of the heating curves (heating rate 30 Cmin -1 ). The four measurements for close to equilibrium TS and TL temperatures were also realized by differential thermal analysis on small steel samples (about 200 mg) during their heating by heating rate 10 Cmin -1 ). Liquidus temperatures were then corrected according to generally accepted methods [14]. Measured TS and TL are summarized in the Table 2. Table 2 Experimental settings and measured TS and TL for close to equilibrium conditions Heating rate Method, Measurement No. Sample mass [mg] T [ Cmin -1 S [ C] ] Direct thermal, T L [ C] Direct thermal, Direct thermal, DTA, DTA, DTA, DTA, Based on data in the Table 2, it can be stated that there is low variability between individual results for close to equilibrium temperatures (standard deviations: 2.5 C for TS and 1.6 C for TL) independently on used method and mass of samples. It means that methodology of measurement is set correctly and results are fully reproducible. Based on mean values, the TS and TL were for selected steel grade were identified: C and C. 3. CALCULATION OF LIQUIDUS AND SOLIDUS TEMPERATURES The calculation of liquidus (TL) and solidus (TS) temperatures of above mentioned steel grade were done by commonly used equations [15-24] and also CompuTherm software. The equations are based on the effect of contents of selected elements (% weight of element) in the steel on final liquidus/solidus temperatures. The different elements and their multiples are taken into account in individual calculations. Some equations include only effects of some elements. On the other hand, some equations include the higher number of elements. In some cases, this fact can lead to the substantially different calculated values of TS and TL. The discussion about limitations of mentioned and generally used equations is out of range of this paper. The CompuTherm SW is able to calculate both studied temperatures. It is possible to choose two microsegregation models (Scheil or Lever). In the case of Lever, the Lever Rule is applied, which corresponds to a complete mixing of the solute in the solid (i.e. very good diffusion in the solid). On the other hand, the Scheil model corresponds to no diffusion model at all in the solid phase (both model consider complete mixing of infinite diffusion in the liquid). The Back Diffusion model allows for some diffusion in the solid and corresponds thus to a situation in between the Lever Rule and Scheil. When the Back Diffusion model is

4 used, an average cooling rate (corresponding to a representative cooling rate of the casting to be modelled) should be specified in order to determine the amount of back diffusion. For iron and carbon steel, the Lever rule is still recommended [25]. The CompuTherm Fe-rich alloys database has defined limitations in the chemical composition and recommended composition limits for them [26]. It was selected the Lever rule (without Pb, Sn, As, Zr, Bi, Ca, Sb, B, N content) for determination of TL and/or TS in the frame of this paper. 4. RESULTS AND DISCUSSION Based on above described methods of thermal analysis, equations and calculation realized in CompuTherm SW, the liquidus temperatures for selected industrially produced steel were determined (Fig.3). It can be seen that some variability between obtained TL values exists. Because, the TL values obtained by thermal analysis (TA) - DTA/Direct TA measurements - were determined from the real steel samples based on standardized methodology, it can be stated that thermal analysis results are the most accuracy to the real TL of studied industrially produced steel. So, the thermal analysis results are used like core values for the next comparing. Source/ Method Calculated / Measured For T L [ C] Calculated versus TA [15] [16] [17.1] [17.2] [18] [19] [20] [17.3] [17.4] [21.1] [22.1] [23] [24] SW TA 1496 X Fig.3 Liquidus temperatures for the studied steel grade determined by equations, CompuTherm and thermoanalytical methods Data (TL) based on selected equations differs from the thermal analysis results from 0 to 35 C, and there is also one extreme difference 118 C when equation from [23] was used. Although, it is possible to determine the best fit of TL base on calculation by equation [24] for this steel grade, it would not be possible without performing experiments of thermal analysis. Moreover, as presented in previous work [27], for other steel grades, which are not subject of this paper, the agreement of calculated results with TA measurements varying and no ideal equation was found for all 5 studied steel grades with [27] dissimilar chemical compositions. If the CompuTherm calculation (SW symbol in Fig.3) is compared with thermal analysis data, the difference in TL is 4 C. It can be postulated that result reached by CompuTherm is very close (unlike most of equations) in comparison with the thermal analysis core measurements for studied steel.

5 The solidus temperatures for studied steel grade were obtained by solving only equations from [21, 22] and by CompuTherm SW thermo dynamical calculation (SW) and by thermal analysis methods results and comparison in Fig.4. Source/ Method Calculated / Measured For T S [ C] Calculated versus TA [21.2] [22.2] SW TA 1479 X Fig.4 Solidus temperatures for the studied steel grade determined by equations, CompuTherm and thermoanalytical methods Opposed to the liquidus temperatures, Fig.4 shows that there are bigger differences between TS calculated by equations [21, 22] and by CompuTherm SW with thermo analysis results, from 3 to 48 C. The reason of such great differences between TA measurements and SW result could be connected with specific (high) content of silicon in this steel grade. 5. CONCLUSIONS The methods for determination of solidus and liquidus temperatures for selected steel grades were compared in the frame of this paper. Two methods of thermal analysis were used (DTA and direct thermal analysis). Generally known equations as well as thermodynamic database software CompuTherm for solidus and liquidus temperatures calculations were used too. Based on obtained results and its comparison can be stated next conclusions: Both applied thermo-analytical methods led to fully comparable and reproducible results. It is not easy to find the best equation from the literature to compute solidus and liquidus temperatures those should be equivalent to thermo-analytical measurements of real steel samples. The realisation of thermo-analytical measurements is necessary for choice the best fit equation. The thermodynamically calculated (by CompuTherm) liquidus temperature is comparable with measured one. Generally, the determination of correct solidus temperature by calculations and by thermo-analytical methods is harder than for liquidus temperature. The best possible way of optimisation of metallurgical process of steel casting could be the obtaining of theoretical values of crucial parameters such as temperatures of liquidus and solidus temperatures and their verification by experimental measurements, followed by simulations and afterwards the implementation to the real casting process. Based on presented results and huge interdisciplinary cooperation between university and industrial partner, the new more complex project Research and development in the field of numerical and material analysis of steel solidification with application output in the optimization of the continuous casting technology at innovative dimensions of billets was prepared and will be realised in period

6 ACKNOWLEDGEMENTS This paper was created in the frame of projects: No. CZ.1.05/2.1.00/ "Regional Materials Science and Technology Centre" within the frame of the operation programme "Research and Development for Innovations" financed by the Structural Funds and from the state budget of the Czech Republic; TIP programme, project No. FR-TI3/053 Improvement of magnetic and quality properties of strips from grain oriented electrical steels ; No. P107/11/1566 of the Czech Science Foundation; SGS (student s university project) project No. SP2013/64, SP2013/62. REFERENCES [1] SOCHA, L. et al. Assessment of Briquetting Fluxing Agent Influence on Refining Effects of Slag during Steel Processing at the Secondary Metallurgy Unit. In METAL 2011: 20th int. metal. conference, p [2] MICHALEK, K. et al. Physical Modelling of bath Homogenization in Argon Stirred Ladle. Metalurgia, 2009, y 48, nr. 4, p [3] JANISZEWSKI, K. Influence of Slenderness Ratios of a Multi-Hole Ceramic Filters at the Effectiveness of Process of Filtration of Non-Metallic Inclusions from Liquid Steel, Archives of Metallurgy and Materials. Vol.57, issue 1/2012. p [4] TKADLEČKOVÁ, M. et al. Setting of Numerical Simulation of Filling and Solidification of Heavy Steel Ingot Based on Real Casting Conditions. Materiali in tehnologije, 2012, vol. 46 no.3, Slovenia. p. 7-10, ISSN [5] ŠESTÁK J. Measurement of thermophysical properties of solids. 1 st ed. Prague: ACADEMIA, (in Czech) [6] BLAŽEK A. Thermal analysis. 1 st ed. Prague: SNTL, (in Czech) [7] MIETTINEN J. Calculation of solidification-related thermophysical properties for steels. Metallurgical and Materials Transactions B, 28B (1997), [8] BOETTINGER W.J., KATTNER U.R. On differential thermal analyzer curves for the melting and freezing of alloys. Metallurgical and Materials Transaction A, 33A (2002), [9] GRÜNBAUM G. et al. A Guide to the Solidification of Steels. Stockholm: Jernkontoret, [10] PORTER D.A. et al. Phase Transformations in Metals and Alloys. 3 rd ed. Boca Raton: CRC Press, [11] ZHAO J.C. Methods for Phase Diagram Determination. Oxford: Elsevier, [12] BANDA W. et. al. Liquidus temperature determination of the Fe-Co-Cu system in the Fe-rich corner by thermal analysis. Journal of Alloys and Compounds, 461 (2008), [13] GALLAGHER P.K. Handbook of Thermal Analysis and Calorimetry: Principles and Practice. 2 nd ed. Oxford: Elsevier, [14] SMETANA, B., ZLÁ, S., DOBROVSKÁ, J., KOZELSKÝ, P. Phase Transformation Temperatures of Pure Iron and Low Alloyed Steels in the Low Temperature Region Using DTA. International Journal of Materials Research, 2010, vol. 101, no. 3, p [15] MYSLIVEC T. Physico-chemical fundamentals of steelmaking. 2 nd ed. Prague: SNTL, 1971, 445 p. (in Czech) [16] ŠMRHA L. Solidification and crystallization of steel ingots. Prague: SNTL, 1983, 305 p. (in Czech) [17] Unpublished information from the industrial partners ( ) [18] DUBOWICK W. Thermal Arrest Measurements and their application in the investment casting, In Proc. of the 2 nd Foundry Congress, Dusseldorf [19] AYMARD J. P., DETREY P. Industiral method for determining liquidus temperatures of steels - evaluation of solidification intervals, Fonderie, Paris, 29(1974)330, 14. [20] ROESER W., WENSEL H. T. Stahl und Eisen, Dusseldorf, (1951), nr. 8. [21] ŠTĚTINA J. Dynamic model of temperature field of continuously cast slabs. Ph.D. thesis, VŠB-TU Ostrava, 2007 [cited ]. Available from World Wide Web: (in Czech) [22] VOLKOVA, O. Mathematical Modelling and Experimental Investigation of the Rapid Solidification of Steels, Dissertation Thesis, Freiberg, Germany, 2002, 138 p. (in German) [23] CSICHOS, H. Fundamentals of Engineering. Berlin: Springer, p. ISBN (in German) [24] MEYER, G., SCHIFFNER, E. Technical Thermodynamics. Leipzig: Fachbuchverlag, p. (in German) [25] ProCast , User s manual (released: Feb-09), ESI Group 2009, p. 163 [26] ProCast , User s manual (released: Feb-09). ESI Group 2009, 175 [27] GRYC, K. et al. Determination of solidus and liquidus temperatures of the real steel grades by dynamic thermal analysis methods. Materiali in Tehnologije, 2013, vol. 47, issue 5, in print