Kinetic study on melt crystallization behaviors of CaO-SiO2-CaF2 based mold fluxes

Abstract

In the continuous casting process of steel, understanding crystallization behavior mold flux is important because it has a huge impact on heat transfer and lubrication which are directly related to the surface quality of slab. Accordingly, it should be required to study the characteristics of crystalline for the purpose of designing proper mold flux according to the different types of steel grade. Since the crystallization behavior of mold flux such as morphology and the quality of crystals is determined by the mechanism of nucleation and growth, the first priority for studying crystallization behavior is to understand the mechanism of crystallization formation. Therefore, the present doctoral dissertation focuses on studying crystallization mechanism for commercial CaO-SiO2-CaF2 based mold flux during the cooling process. In order to meet the main purposes, the crystallization temperature and volume fraction as a function of the temperature and time were evaluated by the DSC thermal analysis. Subsequently, the products from DSC experiments were observed by SEM in order to identify the morphology of crystalline. Furthermore, SHTT (Single Hot Thermocouple Technique) is adapted to confirm the actual crystallization behavior. In the Chapter 4.1 of the present doctoral dissertation, it is investigated that how the basicity and additives (B2O3 and Li2O) has an influence on crystallization behavior. Due to the fact that mold flux experiences non-isothermal history in actual continuous casting process, the crystallization behavior of mold flux is evaluated by DSC with non-isothermal condition. It was found that the mold flux basicity has a tremendous influence on crystallization behavior. In the mold flux A with low basicity, Wollastonite and Cuspidine were competitively formed. On the other hands, in the mold flux B, C and D with high basicity, Cuspidine is solely precipitated as a main phase of crystalline. In particular, out of four different mold fluxes, SEM and SHTT analysis confirms that mold flux B and D show big differences in crystal morphology and formation behavior of crystallization. In Chapter 4.2, crystallization mechanism of Cuspidine during the cooling process is investigated with non-isothermal DSC experiment in order to study Cuspidine formation behavior with increasing basicity. In order to understand Cuspidine formation behavior, Ozawa and Friedman equation which are the most prevalent way to study under non-isothermal condition is applied for the present study. Ozawa exponent, no, representing nucleation mechanism and the dimensionality of crystal growth, increases with an increase on temperature. This could possibly imply that Ozawa analysis is not suitable for studying crystallization mechanism under non-isothermal condition. Instead, Friedman method is applied for the present study in order to obtain effective activation energy for Cuspidine formation. Effective activation energy for Cuspidine formation of mold flux B and D is revealed as negative value. This could mean that Cuspidine formation under the cooling process obey anti-Arrhenius behavior. Such explanation strongly suggests that Cuspidine formation is determined by thermodynamic driving force of nucleation related to the degree of undercooling. Although non-isothermal crystallization mechanism of Cuspidine under cooling process is successfully studied in Chapter 4.2, there has been still long limit in obtaining Avrami exponent, representing nucleation mechanism and the dimensionality of crystal growth, and quantitative value of crystallization rate. Hence, in Chapter 4.3, DSC experiment is conducted under iso-thermal condition for the purpose of evaluating quantitative parameters. In this case, well known JMAC equation is employed for studying crystallization mechanism under iso-thermal condition. According to the results, Avrami exponent for Cuspidine formation of mold flux B is found to be 3.1, meaning that the Cuspidine formation mechanism is governed by 3-dimensional instantaneous nucleation. On the other hands, Avrami exponent for Cuspidine formation of mold flux D equals to 3.4, implying that Cuspidine formation mechanism is governed by 3-dimensional continuous nucleation. Also, for the mold flux B and D, having different basicity, crystallization rate constant decreases as a function of temperature. Such explanation is firmly suggestive that Cuspidine crystallization obeys anti-Arrhenius behaviors. Namely, the entire crystallization rate is controlled by the nucleation which is closely relevant to undercooling. Such results are quite reasonable because they are in well constituent with non-isothermal experimental results. Moreover, the effective activation energy also shows negative values. As it is previously mentioned, the negative activation energy could be measured when crystallization formation is governed by nucleation. As a results of observing microstructure of mold flux B and D shortly after the DSC measurement, the morphology and the number of crystalline were different each other. This phenomenon could be well understood by Avrami value under iso-thermal condition.