Effects of BaO and B2O3 on the Absorption of Ti Inclusions for High Titanium Steel

In order to study the effect of BaO or B2O3 on the absorption of Ti inclusions, the effects of mold fluxes with different contents of BaO (0~15%) or B2O3 (0~15%) on the mass transfer coefficients of TiO2 or TiN were studied with the rotating cylinder method. The experimental results show that with the addition of BaO in the mold flux, the mass transfer coefficient of TiO2 increases from 4.58 × 10−4 m/s to 6.08 × 10−4 m/s, that of TiN increases from 3.09 × 10−4 m/s to 4.41 × 10−4 m/s, 2CaO·MgO·2SiO2 is transformed into BaO·2CaO·MgO·2SiO2, and the Ti inclusions combine with CaO to form CaTiO3. With the addition of B2O3 in the mold flux, the mass transfer coefficient of TiO2 increases from 4.58 × 10−4 m/s to 7.46 × 10−4 m/s, that of TiN increases from 3.09 × 10−4 m/s to 5.50 × 10−4 m/s, CaO and B2O3 combine to 2CaO·B2O3, and Ti inclusions exist in the form of TiO2. During the experiment, TiN will be transformed into titanium oxide.


Introduction
Incoloy825 alloy is a kind of austenitic Fe-Ni-Cr superalloy stabilized by Ti [1]. In the process of continuous casting, Ti inclusions float up to the slag-metal interface and the floating Ti inclusions are inevitably absorbed by mold flux. As a result, the compositions of mold flux are changed, and the properties (melting temperature, viscosity, etc.) of the mold flux are also changed [2][3][4]. Therefore, the performance of mold flux used in practice is different from that of the original design. Consequently, it is urgent to study the ability of mold flux to absorb Ti inclusions.
Due to the importance of absorbing inclusions, this topic has been studied by scholars, and the main research methods used are the rotating cylinder method [5][6][7] and the confocal scanning laser microscope (CSLM) method [8][9][10]. Zhong-shan Ren et al. [11] studied the dissolution and diffusion behavior of TiO 2 particles in molten slag with the CSLM method. Higher temperatures favored the dissolution and diffusion of TiO 2 , whereas a greater Al 2 O 3 content in the slag restrained the dissolution and diffusion. The results of Zhanquan Hao et al. [12] showed that the dissolution rate of TiO 2 in molten mold flux increased with increasing temperature and basicity, decreasing TiO 2 content, and increasing CaF 2 content. The effect of temperature and basicity was the most significant, and the diffusion process was the factor restricting TiO 2 dissolution.
Zhanjun Wang and Tae Sung Kim [13,14] believed that BaO was beneficial for the lubrication of liquid slag. Xiong Yu et al. [15] showed that, with an increase in the B 2 O 3 content, the viscosity, turning point temperature, and viscous flow activation energy of mold flux decreased. Therefore, the addition of BaO or B 2 O 3 into the mold flux can improve the deteriorating performance caused by the absorption of Ti inclusions.
As mentioned above, the experimental work on the effect of BaO or B 2 O 3 on the absorption of Ti inclusions by mold flux is still limited. In this paper, the influence of BaO or B 2 O 3 on the absorption rate of TiO 2 and TiN is studied by the rotating cylinder method. The interfaces between Ti inclusions and mold fluxes are observed and analyzed

Apparatus and Method
The dissolution experiment was carried out in the Vertical MoSi 2 resistance furnace, and a schematic diagram of the experimental apparatus is shown in Figure 1. The furnace body was a corundum tube with an inner diameter of 90 mm and a length of 1000 mm. The furnace temperature was measured by a type-B thermocouple (Pt-30%Rh/ Pt-6%Rh), where the deviation of temperature was maintained within ±2 K of the center of the invariable temperature zone (10 mm).
The interfaces between Ti inclusions and mold fluxes are observed and analyzed b ning electron microscopy (SEM) and energy-dispersive X-ray spectroscopy (EDS), dissolution mechanism is also discussed.

Apparatus and Method
The dissolution experiment was carried out in the Vertical MoSi2 resistance f and a schematic diagram of the experimental apparatus is shown in Figure 1. The body was a corundum tube with an inner diameter of 90 mm and a length of 10 The furnace temperature was measured by a type-B thermocouple (Pt-30%Rh/ Pt where the deviation of temperature was maintained within ±2 K of the center of variable temperature zone (10 mm). The experimental procedures can be described as follows. Firstly, 100 g of pr mold flux sample was placed in a graphite crucible with an inner diameter of 40 m a length of 80 mm, then it was heated to 1623 K in an argon atmosphere (argon The experimental procedures can be described as follows. Firstly, 100 g of premelted mold flux sample was placed in a graphite crucible with an inner diameter of 40 mm and a length of 80 mm, then it was heated to 1623 K in an argon atmosphere (argon purity ≥99.999%). Secondly, when the temperature reached 1623 K, the premelted mold flux should be kept stable for 30 min to ensure uniform melting. The rod with Ti inclusions was preheated for 10 min above the mold flux surface, then immersed in the mold flux for 30 mm and rotated at a fixed speed of 200 rpm/min. At the end of the experiment, the rod with Ti inclusions was quickly withdrawn from the mold flux and cooled in air. The mold flux adhering to the side surface of the rod was removed with HCl dilute solution, as shown in Figure 2, and the rod diameter was measured by a micrometer to an accuracy of 0.02 mm. The morphology of the cross section was subjected to SEM observation with EDS analysis.
Metals 2021, 11,165 ≥99.999%). Secondly, when the temperature reached 1623 K, the premelted mo should be kept stable for 30 min to ensure uniform melting. The rod with Ti inc was preheated for 10 min above the mold flux surface, then immersed in the mold 30 mm and rotated at a fixed speed of 200 rpm/min. At the end of the experiment, with Ti inclusions was quickly withdrawn from the mold flux and cooled in air. Th flux adhering to the side surface of the rod was removed with HCl dilute solu shown in Figure 2, and the rod diameter was measured by a micrometer to an accu 0.02 mm. The morphology of the cross section was subjected to SEM observation w analysis.

Mass Transfer Coefficient
In the process of the rotational cylinder method, the rod (M1) was continuou solved in the slag, its internal component (B) formed an intermediate reaction laye surrounding dissolution process, and component (B) diffused into the slag ag shown in Figure 3. In the process of rod dissolution, the mass of component disso the rod ( The dissolution rate of rod was obtained as:

Mass Transfer Coefficient
In the process of the rotational cylinder method, the rod (M 1 ) was continuously dissolved in the slag, its internal component (1) The dissolution rate of rod was obtained as: where w(B) 1 , w(B) s , and w(B) 2 are the concentration of each component (B); β is the mass transfer coefficient; A is the contact area between inclusion rod and slag; ρ 2 is the density of the slag.
should be kept stable for 30 min to ensure uniform melting. The rod with Ti in was preheated for 10 min above the mold flux surface, then immersed in the mold 30 mm and rotated at a fixed speed of 200 rpm/min. At the end of the experiment with Ti inclusions was quickly withdrawn from the mold flux and cooled in air. T flux adhering to the side surface of the rod was removed with HCl dilute solu shown in Figure 2, and the rod diameter was measured by a micrometer to an acc 0.02 mm. The morphology of the cross section was subjected to SEM observation w analysis.

Mass Transfer Coefficient
In the process of the rotational cylinder method, the rod (M1) was continuo solved in the slag, its internal component (B) formed an intermediate reaction lay surrounding dissolution process, and component (B) diffused into the slag a shown in Figure 3. In the process of rod dissolution, the mass of component diss the rod ( The dissolution rate of rod was obtained as:  During the dissolution process, the surface area of the rod decreased conti and presented a cylindrical shape, as shown in Figure 2. Thus, the dissolution ra rod with Ti inclusions is as follows: During the dissolution process, the surface area of the rod decreased continuously and presented a cylindrical shape, as shown in Figure 2. Thus, the dissolution rate of the rod with Ti inclusions is as follows: where ρ 1 is the density of the rod, r is the diameter of rod, l is the length of the rod in the slag. Equations (2) and (3) are the same, so it is concluded that: When t = 0~t and r = r 0~r , the mass transfer rate of rod dissolution is obtained as follows: According to Equation (5) where 1 ρ is the density of the rod, r is the diameter of rod, l is the length of the rod in the slag. Equations (2) and (3) are the same, so it is concluded that: When t = 0~t and r = r0~r, the mass transfer rate of rod dissolution is obtained as follows: According to Equation (5)  The mass transfer coefficients of TiO2 and TiN increased with the addition of BaO or B2O3 in the mold flux, which indicates that the addition of BaO or B2O3 in the mold flux is beneficial to the absorption of TiO2 and TiN inclusions. The mass transfer coefficients of TiO2 and TiN in mold fluxes containing B2O3 are higher than those of BaO, which indicates that mold flux containing B2O3 has a more obvious effect on the mass transfer coefficient. In addition, the mass transfer coefficient of TiO2 in mold fluxes is greater than that of TiN under the same conditions. This is due to the addition of BaO or B2O3 in the mold flux, which greatly changes the physical properties of the mold flux. With the addition of BaO into the mold flux, the ability to provide basic cations in mold flux is enhanced, and it is easier to absorb Ti inclusions and enhance mass transfer. In the process of mass transfer, the decrease in viscosity enhances the relative movement between the rod with Ti inclusions and mold flux The mass transfer coefficients of TiO 2 and TiN increased with the addition of BaO or B 2 O 3 in the mold flux, which indicates that the addition of BaO or B 2 O 3 in the mold flux is beneficial to the absorption of TiO 2 and TiN inclusions. The mass transfer coefficients of TiO 2 and TiN in mold fluxes containing B 2 O 3 are higher than those of BaO, which indicates that mold flux containing B 2 O 3 has a more obvious effect on the mass transfer coefficient. In addition, the mass transfer coefficient of TiO 2 in mold fluxes is greater than that of TiN under the same conditions. This is due to the addition of BaO or B 2 O 3 in the mold flux, which greatly changes the physical properties of the mold flux. With the addition of BaO into the mold flux, the ability to provide basic cations in mold flux is enhanced, and it is easier to absorb Ti inclusions and enhance mass transfer. In the process of mass transfer, the decrease in viscosity enhances the relative movement between the rod with Ti inclusions and mold flux and increases the mass transfer rate. It can be seen from Table 1 that, with the increase in B 2 O 3 in the mold flux, the viscosities of the mold flux decreased obviously more than those of BaO. Therefore, the mass transfer rate of TiO 2 and TiN in the mold fluxes containing B 2 O 3 is higher than that of BaO.

Phase of Mold Flux
XRD analyses were carried out on the mold flux after adsorbing Ti inclusions, and the results are shown in Figure 5. As shown in Figure 5a, when mold fluxes containing BaO absorb Ti inclusions, the corresponding phase is CaTiO 3 . The peak value of 2CaO·MgO·2SiO 2 is decreased and the peak value of BaO·2CaO·MgO·2SiO 2 increased. With the addition of BaO in the mold flux, BaO combined with 2CaO·MgO·2SiO 2 to form BaO·2CaO·MgO·2SiO 2 . As shown in Figure 5b of BaO. Therefore, the mass transfer rate of TiO2 and TiN in the mold fluxes contain B2O3 is higher than that of BaO.

Phase of Mold Flux
XRD analyses were carried out on the mold flux after adsorbing Ti inclusions, a the results are shown in Figure 5. As shown in Figure 5a, when mold fluxes contain BaO absorb Ti inclusions, the corresponding phase is CaTiO3. The peak value 2CaO·MgO·2SiO2 is decreased and the peak value of BaO·2CaO·MgO·2SiO2 increas With the addition of BaO in the mold flux, BaO combined with 2CaO·MgO·2SiO2 to fo BaO·2CaO·MgO·2SiO2. As shown in Figure 5b, with the addition of B2O3 in the mold fl the properties of the mold fluxes changed greatly. When the content of B2O3 in the m flux increases to 10%, 6# and 7# mold fluxes change from crystalline to glass. This is d to the combination of B2O3 and CaO to form a large amount of 2CaO·B2O3 with a l melting point, which promotes the vitrification of the mold flux and reduces the viscos of the mold flux. The addition of B2O3 to the mold flux reduces the activity of CaO a inhibits the binding ability of CaO and TiO2. Therefore, when mold fluxes containing B absorb TiO2 and TiN, the corresponding phase is still TiO2.

Activity Model of a(CaO)
In order to understand the form of absorbing Ti inclusions in the slag, the activity CaO was calculated by the ion and molecule coexistence theory (IMCT) [17]. On the ba of IMCT, the mole fraction of each oxide could be assigned as  Tables 2 and 3, respectiv [18][19][20][21].

Activity Model of a(CaO)
In order to understand the form of absorbing Ti inclusions in the slag, the activity of CaO was calculated by the ion and molecule coexistence theory (IMCT) [17]. On the basis of IMCT, the mole fraction of each oxide could be assigned as m 1 = n 0 CaO , m 2 = n 0 SiO 2 , m 3 = n 0

Mass Action Concentration of Structural Unit or Ion Couple N i (-)
Na 2 O·MgO·4SiO 2 n c42 = n Na 2 O·MgO·4SiO 2 N c42 = N Na 2 O·MgO·4SiO 2 = n c42 / ∑ n i -Na 2 O·2MgO·6SiO 2 n c43 = n Na 2 O·2MgO·6SiO 2 N c43 = N Na 2 O·2MgO·6SiO 2 = n c43 / ∑ n i Table 3. Chemical reaction formulas of possibly formed complex molecules. −124,683 + 0.766T The equilibrium constant equation of the reaction is established by the structural unit of the slag, and then the activity calculation model of the slag is established according to the mass balance. Finally, the mathematical model is constructed.

Reactions
According to the material balance, it can be concluded that: N 1 , N 2 , N 3 , N 4 , N 5 , N 6 , N 7 could be calculated by solving the algebraic equation groups of Equation (7) to Equation (14) with Matlab, and N 1 represents a(CaO). The activity of CaO in the mold flux containing BaO increased from 0.022 to 0.039.
Similarly, the a(CaO) of the mold flux containing B 2 O 3 was calculated, and activity of CaO in different mold fluxes is shown in Figure 6. The activity of CaO in the mold flux containing B 2 O 3 decreased from 0.022 to 0.003.
Metals 2021, 11,165 (7) to Equation (14) with Matlab, and N1 represents a(CaO). The ac CaO in the mold flux containing BaO increased from 0.022 to 0.039.
Similarly, the a(CaO) of the mold flux containing B2O3 was calculated, and ac CaO in different mold fluxes is shown in Figure 6. The activity of CaO in the m containing B2O3 decreased from 0.022 to 0.003.   With the increase in BaO in the mold flux, the activity of CaO increased, which promoted the combination of CaO and TiO 2 to form CaTiO 3 , thus increasing the mass transfer capacity of TiO 2 , as shown in Equation (15).

Intermediate Reaction Layer
The increase in B 2 O 3 in the mold flux reduces the activity of basic oxides and inhibits the binding ability of TiO 2 and CaO.

Intermediate Reaction Layer
Through line scanning Ti element, it is determined that there is an intermediate reaction layer between the mold flux and the rod in the process of absorbing inclusions by the mold flux (2#), as shown in Figure 7. According to EDS analysis of absorbing Ti inclusions by the mold flux (2#), the main phase is CaTiO 3 , as shown in Table 4. Under the same conditions, the thickness of the intermediate reaction layer for absorbing TiO 2 is about 120 µm, and that for absorbing TiN is about 35 µm, which indicates that the mass transfer coefficient of TiO 2 to the mold flux is greater than that of TiN. This is due to the reaction of TiN and O 2 to produce N 2 , which takes away heat, reduces local temperature, increases viscosity, and slows down the mass transfer process, as shown in Equation (16) [22].

Conclusions
The dissolution of TiO2 and TiN inclusions in mold fluxes containing BaO or B2O3 for steel continuous casting has been investigated by employing the rotating cylinder method. The results are presented as follows:

Conclusions
The dissolution of TiO 2 and TiN inclusions in mold fluxes containing BaO or B 2 O 3 for steel continuous casting has been investigated by employing the rotating cylinder method. The results are presented as follows:

Conflicts of Interest:
The authors declare no conflict of interest.