Analysis of Slag-Steel Reactive of Fluorine-free Mold Fluxes Used for 3.52% Al-TRIP Steel

Abstract

With the improvement of the requirements for lightweight automobiles, the research and development of a higher Al content of TRIP steel have received extensive attention. However, Al in molten steel is easy to react with SiO₂ in mold slag, which is one of the restrictive factors of high Al-TRIP steel continuous casting quality control. Applying the double-film mass transfer theory, combined with the slag-steel reaction equilibrium test, a dynamic equilibrium model is established, and the crystalline phases composition before and after the reaction is analyzed. The results show that the melting process of mold flux samples A and B is uniform and meets the continuous casting control temperature. The main components of the two mold flux samples changed rapidly within 10 minutes of the initial reaction. When the reaction progressed to 20 minutes, the composition of the two mold flux samples had no obvious change trend, and the slag-steel reaction reached equilibrium. The main precipitated crystalline phases of the mold flux sample are LiAlO₂, Ca₃(BO₃)₂, NaAlO₂, ZrO₂, and CaZrO₃ crystals. Mold slag A identified the mass transfer of Al in the steel as the limiting part of the reaction, while mold slag B identified the mass transfer of SiO₂ in the slag as the limiting part of the reaction.

1. Introduction

Aluminum is added to steel as an alloying element, which is beneficial for improving the oxidation resistance of steel, improving the welding performance of steel, calming the molten steel, and preventing the formation of bubbles when the molten steel is solidified.^1,2,3) It can be widely used in aerospace,⁴⁾ automobile industry,⁵⁾ electrical industry⁶⁾ and other fields. However, TRIP steel (Transformation Induced Plasticity Steel), as a new generation of advanced high-strength steel, can be used in fields such as automobile manufacturing^7,8) and cryogenic containers.⁹⁾

Due to the high content of [Al] in molten steel, [Al] in steel is prone to secondary oxidation during continuous casting, and easily reacts with (SiO₂) in traditional mold flux (CaO–SiO₂ base),¹¹⁾ resulting in the chemical composition of slag, Al₂O₃, has increased dramatically. As a result, significant changes in the viscosity and crystallization properties of the mold flux will occur, resulting in poor flow of mold flux and reduced slag consumption, affecting its lubrication and heat transfer control functions, causing slab dents, excessive vibration marks, cracks, and slag inclusions and other quality issues.¹²⁾

Therefore, researchers are committed to the development of new mold fluxes, currently focusing on the research of reactive mold slag and non-(weak) reactive mold slag.¹³⁾ The reactive slag is made by increasing the SiO₂ content in the slag and reducing the Al₂O₃ content in the slag. However, a large amount of Al₂O₃ will enter the mold flux, and the physical and chemical properties of the mold flux will still undergo major changes,¹⁴⁾ which cannot fundamentally solve the quality problems faced by the continuous casting of high-aluminum steel. Non-(weak) reactive: by reducing the SiO₂ content in the slag or even without SiO₂, the quality problem caused by the slag-steel reaction during the continuous casting of high-aluminum steel can be fundamentally solved.^15,16) However, increasing the activity of Al₂O₃ in the original slag and reducing the activity of SiO₂ may not necessarily effectively inhibit the reaction between the mold slag and the high Al-TRIP steel. And its physical and chemical properties are significantly different from the traditional CaO–SiO₂ based mold flux, which causes many new problems.¹⁷⁾

Some researchers have begun to investigate the kinetics of non-(low) reactivity mold flux slag-steel reactions. Improving the continuous casting quality of high-aluminum steel by controlling the slag-steel interface reaction rate. Yu et al.¹⁸⁾ found that when the critical content of SiO₂ and B₂O₃ in TRIP steel with 1.0% Al content is 6%, the SiO₂ and B₂O₃ in the mold flux do not react with Al in the molten steel. Kim et al.¹⁹⁾ and Kang et al.¹²⁾ analyzed the reaction mechanism of slag steel from the perspective of reaction kinetics, and the results showed that: the content of [Al] in the steel is the main factor affecting the slag-steel reaction. The mass transfer of [Al] in the molten steel boundary layer is also the rate-controlling step of the entire reaction. Na₂O can also react with Al in the molten steel, and the generated Na may enter the molten steel or mold slag. Zhang et al.²⁰⁾ analyzed the slag-steel reaction mechanism from the dynamic equilibrium model. The results showed that the SiO₂ in the slag reacted by the slag-steel is not a limiting link, but the limiting link is not pointed out. Adjusting the diffusion coefficient and parameters such as the thickness of the liquid slag layer can control the loss of SiO₂. For high Al-TRIP steel, the main Al₂O₃ inclusions in the steel will enter the mold slag in a large amount and cause the viscosity to increase significantly, which seriously affects the lubrication function of the mold slag. Wang et al.²¹⁾ found that although the addition of CaF₂ is beneficial to reducing the viscosity of mold slag, fluorine will react to form oxyfluoride during the melting process of mold slag and easily precipitate CaF₂ crystals, which will cause equipment corrosion and slag ring problems during continuous casting. Park, Kim et al.^22,23,24) studied the viscosity of high-aluminum steel continuous casting mold flux and found that in high-aluminum steel grades, Al₂O₃ is mainly used as an amphoteric oxide to affect the viscosity. Generally speaking, the erosion of the continuous casting mold immerging-type nozzle at the interface between steel and slag is serious, which will reduce the service life of the nozzle. Kumar et al.²⁵⁾ found that mold flux, which will corrode ZrO₂–C intrusive nozzles, especially SiO₂ mold flux will corrode refractory materials more severely. If high melting point substances of CaZrO₃ are formed during the corrosion process, it will prevent the erosion of the mold flux, thereby prolonging the service life of the nozzle.

In conclusion, the slag-steel reaction of non-(weak) reactive mold flux with high Al-TRIP steel will degrade mold flux performance and cause refractory material corrosion. In this paper, a high-aluminum, fluorine-free, and environmentally friendly CaO–Al₂O₃ based mold flux is used to analyze the mold flux after being pre-melted into molten steel for the slag-steel reaction test, and the samples before and after the reaction are analyzed to determine their reactivity, composition, melting point, and crystallization changes in degree. Analyze the crystalline phases composition through the X-ray diffraction (XRD, MXP21VAHF, MAC Science, Japan) test, and finally establish a kinetic model to study the restrictive link of the molten steel reaction mass transfer.

2. Theoretical Fundamental and Experimental

2.1. Kinetic Mechanism of the Slag-steel Reaction

During the continuous casting of high-aluminum steel, the main reaction at the interface is a strong redox reaction between Al in the steel and SiO₂ in the mold slag. According to the dual-mode theory, there are two concentration boundary layers at the slag-steel interface, and the schematic diagram is shown in Fig. 1. where δ_M and δ_S are the thickness of the boundary layer on the slag side and the boundary layer on the molten steel side, respectively. The reaction is divided into the following steps: ① The mass transfer of the reactant Al from the inside of the molten steel to the boundary layer on the molten steel side; ② The mass transfer of the reactant SiO₂ from the liquid slag to the liquid slag side boundary layer; ③ Interfacial chemical reaction, that is, the breaking of the Si–O bond and the formation of the Al–O bond; ④ The product Si is transported from the slag-steel interface to the molten steel mass; ⑤ The product, Al₂O₃, is transferred from the slag-steel interface to the liquid slag.

Fig. 1.

Schematic diagram of interface slag-steel reaction steps. (Online version in color.)

Since the driving force of the slag-steel reaction is strong enough at the continuous casting temperature, step ③ is not a reaction-limiting link. Therefore, in this study, steps ①, ②, ④, and ⑤ are assumed to be the kinetic control conditions for the slag-steel reaction, respectively. According to the double-film theory, the following assumptions need to be made:

(1) The phases of molten steel and liquid slag are homogeneous;

(2) The chemical reaction at the slag-steel interface is in an instantaneous equilibrium state at any time;

(3) Ignore the effect of inclusions and bubbles floating in the slag-steel reaction’s mold slag reaction;

(4) The slag-steel reaction only considers the reaction between Al and SiO₂ by default.

Assuming that Al mass transfer is the limiting link, according to the double-film theory, then:   

$$\frac{d{n}_{A\mathrm{l}}}{dt}=-A\frac{D_{\mathrm{A}\mathrm{l}}}{{\delta }_{\mathrm{m}}}(C_{A\mathrm{l}}-C_{A\mathrm{l}}{}^{\mathrm{i}})$$(1)

In the formula, $c_{A\mathrm{l}}^{\mathrm{i}}$ — the concentration of Al in the interface, mol/m³; c_Al — the concentration of Al in the molten steel, mol/m³; n_Al — the number of moles of Al, mol; A — the slag-steel reaction interface area, m²; D_Al — Diffusion coefficient of Al, m²/s; δ_m — Boundary layer thickness of molten steel, m.   

$$c_{\mathrm{A}\mathrm{l}}=\frac{[\%\mathrm{A}\mathrm{l}]}{100}\cdot \frac{{\rho }_{\mathrm{m}}}{M_{\mathrm{A}\mathrm{l}}}$$(2)

In the formula, [Al%] — the percentage content of Al in the molten steel; M_Al — the relative atomic mass of Al, 27.

but:   

$$n_{\mathrm{A}\mathrm{l}}=c_{\mathrm{A}\mathrm{l}}\cdot V_m=\frac{[\%\mathrm{A}\mathrm{l}]}{100}\cdot \frac{{\rho }_{\mathrm{m}}}{M_{\mathrm{A}\mathrm{l}}}\cdot V_{\mathrm{m}}=\frac{[\%\mathrm{A}\mathrm{l}]}{100}\cdot \frac{{\rho }_{\mathrm{m}}}{M_{\mathrm{A}\mathrm{l}}}\cdot A\cdot h_{\mathrm{m}}$$(3)

In the formula V_m — volume of molten steel, m³; h_m — height of molten steel, m; ρ_m — density of molten steel, kg/m³.

So we get:   

$$\frac{d[\%\mathrm{A}\mathrm{l}]}{dt}=-\frac{D_{\mathrm{A}\mathrm{l}}}{{\delta }_{\mathrm{m}}{h}_{\mathrm{m}}}([\%\mathrm{A}\mathrm{l}]-[{\%}^{\mathrm{i}}\mathrm{A}\mathrm{l}])$$(4)

Due to the strong Al–O binding energy, Al in the molten steel enters the reaction interface and immediately combines with O, and it is assumed that the mass transfer of Al is the reaction-limiting link, so [%ⁱAl] in the interface can be regarded as zero, there are:   

$$\frac{d[\%\mathrm{A}\mathrm{l}]}{dt}=-\frac{D_{\mathrm{A}\mathrm{l}}}{{\delta }_{\mathrm{m}}{h}_{\mathrm{m}}}[\%\mathrm{A}\mathrm{l}]$$(5)

After integral calculation, there are:   

$$ln\frac{{[\%\mathrm{A}\mathrm{l}]}_t}{{[\%\mathrm{A}\mathrm{l}]}_0}=-\frac{D_{\mathrm{A}\mathrm{l}}}{{\delta }_{\mathrm{m}}{h}_{\mathrm{m}}}t$$(6)

Similarly, it can be obtained that [Si], (Al₂O₃), (SiO₂) mass transfer is the limiting link, there are:   

$$ln\frac{{[\%\mathrm{S}\mathrm{i}]}_t}{{[\%\mathrm{S}\mathrm{i}]}_0}=-\frac{D_{\mathrm{S}\mathrm{i}}}{{\delta }_{\mathrm{m}}{h}_{\mathrm{m}}}t$$(7)

  

$$ln\frac{{[\%\mathrm{A}{\mathrm{l}}_2{\mathrm{O}}_3]}_t}{{[\%\mathrm{A}{\mathrm{l}}_2{\mathrm{O}}_3]}_0}=-\frac{D_{\mathrm{A}{\mathrm{l}}_2{\mathrm{O}}_3}}{{\delta }_{\mathrm{f}}{h}_{\mathrm{f}}}t$$(8)

  

$$ln\frac{{[\%\mathrm{S}\mathrm{i}{\mathrm{O}}_2]}_t}{{[\%\mathrm{S}\mathrm{i}{\mathrm{O}}_2]}_0}=-\frac{D_{\mathrm{S}\mathrm{i}{\mathrm{O}}_2}}{{\delta }_{\mathrm{f}}{h}_{\mathrm{f}}}t$$(9)

In the formula [%Si] — the percentage content of Si in the molten steel, %; [%ⁱSi] — the percentage content of Si at the slag-steel interface, %; D_Si — the diffusion coefficient of Si, m²/s; [%Al₂O₃] — the percentage content of Al₂O₃ in the molten steel, %; [%ⁱAl₂O₃] — The percentage content of Al₂O₃ at the slag-steel interface, %; D_Al2O3 — Diffusion coefficient of Al₂O₃, m²/s; h_f — Liquid slag thickness, m; δ_f — Liquid slag boundary layer thickness, m. D_SiO2 — Diffusion coefficient of SiO₂, m²/s; [%SiO₂] — Percentage content of SiO₂ in molten steel, %; [%ⁱSiO₂] — Percentage content of SiO₂ at slag-steel interface, %.

Therefore, the slag-steel reaction test can be carried out according to the pre-melted mold flux poured into molten steel. The kinetic models of [Al], [Si] in molten steel and (Al₂O₃), (SiO₂) in slag and time t were established respectively. If this component is the limiting link in the slag-steel reaction, the logarithm of the ratio of the real-time content to the initial content of this component has a linear relationship with t.

2.2. Materials Preparation

The chemical compositions of the high-aluminum TRIP steel samples used in this study are shown in Table 1. The high-aluminum steel samples were smelted in a vacuum induction furnace (VIM, Wujin Company, UK). The industrial pure iron is kept at 1873 K in a vacuum induction furnace for 30 minutes and filled with argon gas to protect the molten steel, which is smelted by adding alloy components after ensuring complete melting. Samples A and B are low-reactivity mold flux, and mold flux samples are prepared from analytically pure (MacLean) CaCO₃, SiO₂, Al₂O₃, MgO, B₂O₃, Na₂CO₃, Li₂CO₃. These reagents were first mixed by mechanical stirring, then pre-melted in an induction furnace at 1873 K for 10 minutes, cooled, and ground to powder. The composition of mold flux was determined by the XRF analysis method (XRF-1800, Shimadzu Co., Ltd, Japan) and the ICP-MS Inductively Coupled Plasma Mass Spectrometer (Plasma MS 300, Steel Research Nanogram), the design and actual values are shown in Table 2. The ratio S/A of the percentage content of SiO₂ and Al₂O₃ in the mold flux can simply reflect the reactivity of the mold flux sample; The basicity R of the mold powder sample is the percentage content ratio of CaO and SiO₂; Grind the prepared mold flux into 200-mesh powder, reconcile the sample with absolute ethanol, and use a mold to make a Φ3×3 (mm) cylindrical sample. After drying, use a slag melting point and melting rate tester (MTS-1, National Engineering Research Center for Continuous Casting Technology) to measure the melting points of the two mold slags at a heating rate of 15°C/minute, and the T_m represents the melting point of the mold slag sample.

Table 1. Main chemical compositions of 3.52Al-TRIP steel (wt%).

steel	Al	C	Si	Mn	N	S	O	Fe
S1	3.52	0.37	0.39	1.29	0.00099	0.0032	0.00066	Bal

Table 2. Main chemical compositions of mold flux slag samples (design and actual values) (wt%).

Sample	CaO	Al2O3	SiO2	B2O3	MgO	Na2O	Li2O	R	S/A	Tm°C
A (design values)	40	31	7	3.5	3	12	3.5	5.7	0.22	/
B (design values)	42.5	30	7	2	3	12	3.5	6.1	0.23	/
A (actual values)	39.15	30.28	7.61	3.75	3.23	11.77	3.45	5.14	0.25	1140
B (actual values)	41.15	30.94	7.29	1.8	2.88	11.35	3.38	5.64	0.23	1106

2.3. Experimental Process and Analytical Methods

Figure 2 is a diagram of the slag-steel reaction equilibrium test device. The ZrO₂ crucible used for the test was filled with a 300 g steel sample jacketed graphite crucible and placed in a tube furnace. The steel sample was heated to 1550°C at a heating rate of 10°C/minute and then kept for 30 minutes to ensure that the steel sample was fully melted, during which argon was vented at a rate of 5 L/minute to prevent the molten steel from oxidizing. 60 g of mold flux was poured into molten steel for the reaction, and the whole sample was taken out and poured into the water for cooling after 10 minutes, 20 minutes, and 120 minutes of reaction respectively. After the reaction, the mold flux sample was pulverized and ground, and the composition of the sample was determined by XRF and ICP-MS. Each group of slag-steel reaction equilibrium tests was conducted twice to establish the test’s reliability, and the average value was used to determine the experiment’s final composition. The crystalline phases composition was tested using XRD and the crystallinity of the mold flux was calculated using the Profile Fitting (Peak Decomposition) module of the MDI Jade6 (California, America) software.

Fig. 2.

Schematic of the experimental apparatus. (Online version in color.)

3. Results

3.1. Analysis of Melting Point of Initial Composition Mold Flux

The melting point results for samples A and B are shown in Fig. 3. It can be found that the melting points of mold slag A and mold slag B are 1140°C and 1106°C, respectively. During continuous casting, the melting temperature of mold slag is usually controlled below 1200°C. Both samples A and B meet the requirements of continuous casting. However, compared with the melting rate of mold flux B of 95 s, the melting rate of mold flux A is 169 s, which may lead to problems such as an increased thickness of the liquid slag layer and too low slag consumption.

Fig. 3.

Melting temperature of mold flux sample: (a1–a3) mold flux sample A; (b1–b3) mold flux sample B. (Online version in color.)

3.2. Analysis of Evolution Process of the Chemical Composition of Mold Flux

As shown in Table 2, the mold flux samples A and B selected in this study had similar S/A and viscosity, Na₂O and Li₂O were added as fluxes, and B₂O₃ was added to enhance the ability to absorb Al₂O₃. The higher content of Al₂O₃ in mold flux samples A and B is considered to reduce S/A and weaken the reactivity, while the increase of CaO content can weaken the crystallization ability of mold flux. Therefore, mold flux samples A and B may be suitable for continuous casting of high-aluminum steel in terms of reactivity and heat transfer.

Figure 4 shows the component contents of the two mold fluxes after reacting with 3.52% Al-TRIP steel for 0 minutes, 10 minutes, 20 minutes, and 120 minutes, respectively. At the beginning of the slag-steel reaction, the reaction driving force was strong, the main components of the two mold slag samples changed rapidly within the initial 10 min of the reaction, SiO₂ was rapidly consumed and the content of Al₂O₃ increased rapidly. In the following 10 minutes, the driving force of the reaction was weakened due to the rapid decrease of S/A in the mold slag sample, so the slag-steel reaction rate gradually slowed down. After the reaction was carried out for 20 minutes, the composition of the two mold slag samples had no obvious change trend, which means that the slag-steel reaction gradually slowed down. Under this test condition, the reaction equilibrium endpoint was about 20 minutes. Finally, the content of SiO₂ in mold slag samples A and B is only 0.83% and 0.93%, respectively, and the content of B₂O₃ is 0.24% and 0.14%, respectively, both of which become SiO₂-free slag systems. However, mold flux A still has a weak slag-steel reaction after 20 minutes of slag-steel reaction, while mold flux B has almost no slag-steel reaction. It shows that there is still a slag-steel reaction when the mold slag samples A and B with lower S/A react with 3.52% Al-TRIP steel. This shows that for the continuous casting process of TRIP steel with an Al content exceeding 1.0%, the high Al content in the steel has a higher influence on the slag-steel reaction than the reactivity of the mold flux itself. In addition, little Na₂O content was detected in mold slag samples A and B after the slag-steel reaction, which may be due to volatilization or as a co-solvent. At the same time, during the reaction process, the mold slag corrodes the ZrO₂ crucible, which causes the ZrO₂ to enter the mold slag.

Fig. 4.

Changes in the main components of the mold flux sample: (a1– a2) mold flux sample A; (b1– b2) mold flux sample B. (Online version in color.)

3.3. Research on the Composition and Crystallinity of Crystalline Phases

As shown in Fig. 5, the main crystalline phases precipitated in mold flux sample A are mainly LiAlO₂ crystals, Ca₃(BO₃)₂ crystals, NaAlO₂ crystals, ZrO₂ crystals, and CaZrO₃ crystals. The main crystalline phases precipitated in mold slag sample B are LiAlO₂ crystals, Ca₃(BO₃)₂ crystals, NaAlO₂ crystals, ZrO₂ crystals, 3CaO Al₂O₃ crystals, and Ca_0.2Zr_0.8O_1.8 crystals. Due to the volatile phenomenon, there are no NaAlO₂ crystals in the mold fluxes after the reaction. B₂O₃ will promote the formation of the glass phase in the slag film and prevent the precipitation of high melting point crystals, which is beneficial for reducing the melting temperature and improving the lubrication of the slab. However, B₂O₃ is also easy to react with [Al] in molten steel, and the content of B₂O₃ after the slag-steel reaction is very small, so the crystals phase of Ca₃(BO₃)₂ will also decrease. With the increase of the reaction time, the erosion of the ZrO₂ crucible gradually increases, and some ZrO₂ will penetrate the mold slag, resulting in the formation of ZrO₂ and CaZrO₃ crystals phases. The initial crystalline phases of mold flux sample B contains 3CaO Al₂O₃, which also exists stably after the reaction. Ca_0.2Zr_0.8O_1.8 is generated due to the infiltration of ZrO₂ into the mold flux.

Fig. 5.

X-ray diffraction patterns of different reaction times: (a) mold flux sample A; (b) mold flux sample B. (Online version in color.)

The crystallinity of mold flux samples A and B is shown in Fig. 6. The initial crystallinity of mold flux sample A is 82.75%, and the crystallinity begins to drop to 61.8% with the increase of reaction time. When the reaction time is 120 minutes, the crystallinity rapidly increases to 83.37%. The crystallinity of the initial sample of mold flux sample B was 91.32%. When the reaction time was 10 minutes, the crystallinity began to drop sharply to 69.78%. When the reaction time was 120 minutes, the crystallinity rapidly increased to 89.13%. This may be because, with the increase of reaction time, a large amount of high melting point Al₂O₃ is formed, which enters into the mold flux, causing the viscosity to rise and the crystallization performance to drop rapidly. At the same time, the corrosion of the ZrO₂ crucible increases with the increase of reaction time, and ZrO₂ penetrates the mold flux, resulting in the formation of ZrO₂ and CaZrO₃ crystals phases. ZrO₂ is beneficial for promoting the crystallinity of mold flux and inhibiting the formation of the Al₂O₃ crystals phase.

Fig. 6.

The crystallinity of mold flux sample A and B at different reaction times. (Online version in color.)

Fig. 7.

CaO/Al₂O₃ of mold flux samples A and B. (Online version in color.)

4. Discussion

4.1. Changes of Slag Properties during the Reaction of Slag to Steel

The reasons for the increased viscosity of mold slag may be the following: (1) In the CaO–Al₂O₃ slag composition system, the slag-steel reaction leads to a decrease in the ratio of CaO/Al₂O₃, as shown in Fig. 8. With the production of Al₂O₃ with a larger crystallization factor, the SiO₂ in the slag decreases.²⁶⁾ More O²⁻ participates in the formation of Al–O–Al oxygen bridges, and the network structure formed by [AlO₄]⁵⁻ tetrahedral units is more stable; (2) Since B₂O₃ participates in the slag-steel reaction, B₂O₃ fails to destroy the Al–O–Al oxygen bridge in the slag, which increases the degree of polymerization of the slag; (3) It is also possible that the ZrO₂ crucible is gradually eroded with the slag-steel reaction, resulting in the introduction of ZrO₂ into the mold slag to increase its viscosity. The reason is that the Na₂O, B₂O₃, and SiO₂ components in the mold slag promote the dissolution of ZrO₂ in the slag.

Fig. 8.

ZrO₂–CaO–Al₂O₃ Phase Diagram. (Online version in color.)

According to the phase diagram in Fig. 8, at a certain temperature, ZrO₂ and calcium oxide will react to form a CaZrO₃ high melting point substance. The melting point is above 2000°C. From XRD experiments, it is found that samples A and B both generate CaZrO₃ or derivatives, which can theoretically prevent the erosion of mold slag into the refractory material, thereby prolonging the service life of the mold nozzle.

4.2. Research on the Restrictive Link of Slag-steel Reaction Kinetics

The composition analysis shows that the composition of the mold slag does not change after the slag-steel reaction for 20 minutes, which provides a key point for the restrictive link of the slag-steel reaction kinetics.

Figure 9 shows the relationship between [%Al₂O₃], [%SiO₂] in samples A and B, and [%Al], [%Si] in molten steel and time t after the slag-steel reaction. It can be found that in the reaction process of 0–20 minutes, the $ln\frac{{[\%\mathrm{A}{\mathrm{l}}_2{\mathrm{O}}_3]}_t}{{[\%\mathrm{A}{\mathrm{l}}_2{\mathrm{O}}_3]}_0}$ and $ln\frac{{[\%\mathrm{S}\mathrm{i}{\mathrm{O}}_2]}_t}{{[\%\mathrm{S}\mathrm{i}{\mathrm{O}}_2]}_0}$ in mold flux sample A and $ln\frac{{[\%\mathrm{S}\mathrm{i}]}_t}{{[\%\mathrm{S}\mathrm{i}]}_0}$ in molten steel have a nonlinear relationship with t, which is inconsistent with the relationship in (7)–(9). Therefore, it is judged that the mass transfer of Al₂O₃ and SiO₂ in mold slag and Si in molten steel is not the limiting link of sample A slag-steel reaction within 0–20 minutes of slag-steel reaction under the conditions of this study. The $ln\frac{{[\%\mathrm{A}{\mathrm{l}}_2{\mathrm{O}}_3]}_t}{{[\%\mathrm{A}{\mathrm{l}}_2{\mathrm{O}}_3]}_0}$ in the mold flux sample B and the $ln\frac{{[\%\mathrm{S}\mathrm{i}]}_t}{{[\%\mathrm{S}\mathrm{i}]}_0}$ and $ln\frac{{[\%\mathrm{A}\mathrm{l}]}_t}{{[\%\mathrm{A}\mathrm{l}]}_0}$ in the molten steel has a nonlinear relationship with the time t, which is inconsistent with the relationship in (6)–(8). Therefore, the mass transfer of Al₂O₃ in mold slag B and Si and Al in molten steel are not the limiting links of slag-steel reaction under the conditions of this study.

Fig. 9.

The relationship between the composition of mold flux samples A and B (0–20 minutes) and time t: [Al₂O₃%]; (b) mold flux sample [Si%]; (c) mold flux sample [SiO₂%]; (d) mold flux sample [Al%]. (Online version in color.)

At the same time, the relationship between [%Al] in molten steel and time t after the slag-steel reaction of sample A can be obtained. It can be found that during the reaction process, there is a linear relationship between the molten steel and the t. Therefore, it is determined that the limiting link of the slag-steel reaction of sample A in this study is the mass transfer of Al in the molten steel. The relationship between [%SiO₂] in the slag-steel reaction of sample B and time t. From the slag-steel reaction of 0–20 minutes, it can be found that the relationship between $ln\frac{{[\%\mathrm{S}\mathrm{i}{\mathrm{O}}_2]}_t}{{[\%\mathrm{S}\mathrm{i}{\mathrm{O}}_2]}_0}$ and t is linear. Therefore, it is determined that the mass transfer of SiO₂ in mold slag B is the limiting link of the slag-steel reaction under the conditions of this study. For on-site production, Al in molten steel is an invariable factor. When the loss of SiO₂ is the limiting link, by adjusting the loss of SiO₂, the slag-steel reaction can be reduced and the quality of continuous casting can be improved.

Figure 10 shows the equilibrium test for the slag-steel reaction from 0 to 120 minutes. The relationship between [%Al₂O₃], [%SiO₂] in samples A and B and [%Si], [%Al], and t in molten steel was found. Within 0-120 minutes of the slag-steel reaction, $ln\frac{{[\%\mathrm{A}{\mathrm{l}}_2{\mathrm{O}}_3]}_t}{{[\%\mathrm{A}{\mathrm{l}}_2{\mathrm{O}}_3]}_0}$, and $ln\frac{{[\%\mathrm{S}\mathrm{i}{\mathrm{O}}_2]}_t}{{[\%\mathrm{S}\mathrm{i}{\mathrm{O}}_2]}_0}$ in mold slag A and B, $ln\frac{{[\%\mathrm{S}\mathrm{i}]}_t}{{[\%\mathrm{S}\mathrm{i}]}_0}$, and $ln\frac{{[\%\mathrm{A}\mathrm{l}]}_t}{{[\%\mathrm{A}\mathrm{l}]}_0}$ in molten steel are nonlinearly related to time t. The research shows that there is an applicable range for the slag-steel kinetic model, which is generally before the slag-steel reaction gradually slows down. At the same time, the limiting kinetic link can be the composition of molten steel or the composition of mold slag.

Fig. 10.

The relationship between the composition of mold flux sample A and B and time t: (a) mold flux sample [Al₂O₃%]; (b) mold flux sample [Si%]; (c) mold flux sample [SiO₂%]; (d) mold flux sample [Al%]. (Online version in color.)

To sum up, this study found that the mass transfer of SiO₂ in the mold slag was the limiting kinetic link. Therefore, the content of SiO₂ should be strictly controlled when designing a low-reactivity mold flux.

5. Conclusions

In this paper, the changes in slag composition, melting point, and crystallinity during the slag-steel reaction were studied. Based on the double-film mass transfer theory, the restrictive links of the slag-steel reaction in the continuous casting of high-aluminum steel are analyzed, and the restrictive links of the slag-steel reaction. The conclusions are as follows:

(1) Due to the strong reaction driving force at the beginning of the slag-steel reaction, the main components of the two mold slag samples changed rapidly within the initial 10 minutes of the reaction. When the reaction was carried out for 20 minutes, the composition of the two mold slags had no obvious change trend, and the slag-steel reaction was balanced.

(2) The main crystalline phases precipitated from the mold slag are mainly LiAlO₂ crystals, Ca₃(BO₃)₂ crystals, NaAlO₂ crystals. With the increase of reaction time, the erosion of the ZrO₂ crucible gradually increases, and some ZrO₂ will penetrate into the mold slag, resulting in the formation of ZrO₂, CaZrO₃ crystals phases, while the formation of ZrO₂ crystals inhibits the formation of Al₂O₃ crystals phases.

(3) Mold slag A identified the mass transfer of Al in the steel as the limiting part of the reaction, while mold slag B identified the mass transfer of SiO₂ in the slag as the limiting part of the reaction.

(4) The melting point of sample B is lower than that of sample A, the melting rate is faster. By strictly controlling the content of SiO₂, the continuous casting quality of high aluminum steel can be improved.

Acknowledgement

The authors are grateful for the financial support of this work from the National Natural Science Foundation of China (No. U1860106).

References

• 1)   U.  Brüx,  G.  Frommeyer and  J.  Jimenez: Steel Res., 73 (2002), 543. https://doi.org/10.1002/srin.200200026
• 2)   J. E.  Jin and  Y. K.  Lee: Acta Mater., 60 (2012), 1680. https://doi.org/10.1016/j.actamat.2011.12.004
• 3)   Z.  Cheng,  S.  Gao,  J.  Wang,  H.  He,  Z.  Liu and  M.  Wu: COMPEL- Int. J. Comput. Math. Electr. Electron. Eng., 17 (1998), 347. https://doi.org/10.1108/03321649810203260
• 4)   W. P.  Tong,  Z.  Han,  L.  Wang,  J.  Lu and  K.  Lu: Surf. Coat. Technol., 202 (2008), 4957. https://doi.org/10.1016/j.surfcoat.2008.04.085
• 5)   O.  Grässel,  L.  Krüger,  G.  Frommeyer and  L. W.  Meyer: Int. J. Plast., 16 (2000), 1391. https://doi.org/10.1016/S0749-6419(00)00015-2
• 6)   W.  Yan,  W. Q.  Chen,  Y. D.  Yang,  C.  Lippold and  A.  Mclean: Ironmaking Steelmaking, 42 (2015), 698. https://doi.org/10.1179/1743281215Y.0000000024
• 7)   Y. X.  Li,  Z. Q.  Lin,  A. Q.  Jiang and  G. L.  Chen: Mater. Des., 24 (2003), 177. https://doi.org/10.1016/S0261-3069(03)00021-9
• 8)   G. M.  Liu,  J. B.  Wang,  Y. F.  Ji,  R. Y.  Hao,  H. Y.  Li,  Y. G.  Li and  Z. Y.  Jiang: Materials, 14 (2021), 2478. https://doi.org/10.3390/ma14102478
• 9)   G.  Frommeyer,  U.  Brüx and  P.  Neumann: ISIJ Int., 43 (2003), 438. https://doi.org/10.2355/isijinternational.43.438
• 10)   C.-x.  Ji,  Y.  Cui,  Z.  Zeng,  Z.-h.  Tian,  C.-l.  Zhao and  G.-s.  Zhu: J. Iron Steel Res. Int., 22 (2015), 53. https://doi.org/10.1016/S1006-706X(15)30138-2
• 11)   H.  Cui,  K. T.  Zhang,  Z.  Wang,  B.  Chen,  B. S.  Liu,  J.  Qing and  Z. J.  Li: Metals, 9 (2019), 204. https://doi.org/10.3390/met9020204
• 12)   Y.-B.  Kang,  M.-S.  Kim,  S.-W.  Lee,  J.-W.  Cho,  M.-S.  Park and  H.-G.  Lee: Metall. Mater. Trans. B, 44 (2013), 309. https://doi.org/10.1007/s11663-012-9769-5
• 13)   W. L.  Wang,  B. X.  Lu and  D.  Xiao: Metall. Mater. Trans. B, 47 (2016), 384. https://doi.org/10.1007/s11663-015-0474-z
• 14)   K. T.  Zhang,  J. H.  Liu,  H.  Cui and  C.  Xiao: Metall. Mater. Trans. B, 49 (2018), 1174. https://doi.org/10.1007/s11663-018-1236-5
• 15)   J.  Li,  Q.  Shu,  X.  Hou and  K.  Chou: ISIJ Int., 55 (2015), 830. https://doi.org/10.2355/isijinternational.55.830
• 16)   J.  Gao,  G.  Wen,  T.  Huang,  B.  Bai,  P.  Tang and  Q.  Liu: J. Non-Cryst. Solids., 452 (2016), 119. https://doi.org/10.1016/j.jnoncrysol.2016.08.036
• 17)   Q.  Wang,  J.  Yang,  J.  Zhang,  O.  Ostrovski,  C.  Zhang and  D.  Cai: Steel Res. Int., 93 (2022), 2100193. https://doi.org/10.1002/srin.202100193
• 18)   X.  Yu,  G. H.  Wen,  P.  Tang and  H.  Wang: Ironmaking Steelmaking, 36 (2009), 623. https://doi.org/10.1179/174328109X461419
• 19)   M. S.  Kim,  S. W.  Lee,  J. W.  Cho,  M. S.  Park,  H. G.  Lee and  Y. B.  Kang: Metall. Mater. Trans. B, 44 (2013), 299. https://doi.org/10.1007/s11663-012-9770-z
• 20)   K. T.  Zhang,  J. H.  Liu and  H.  Cui: ISIJ Int., 59 (2019), 2256. https://doi.org/10.2355/isijinternational.ISIJINT-2019-291
• 21)   W.  Wang,  K.  Blazek and  A.  Cramb: Metall. Mater. Trans. B, 39 (2008), 66. https://doi.org/10.1007/s11663-007-9110-x
• 22)   H. S.  Park,  S. S.  Park and  I.  Sohn: Metall. Mater. Trans. B, 42 (2011), 692. https://doi.org/10.1007/s11663-011-9512-7
• 23)   H.  Kim,  H.  Matsuura,  F.  Tsukihashi,  W.  Wang,  D. J.  Min and  I.  Sohn: Metall. Mater. Trans. B, 44 (2013), 5. https://doi.org/10.1007/s11663-012-9759-7
• 24)   G. H.  Zhang and  K. C.  Chou: ISIJ Int., 53 (2013), 177. https://doi.org/10.2355/isijinternational.53.177
• 25)   A.  Kumar,  R.  Khanna,  J.  Spink and  V.  Sahajwalla: Steel Res. Int., 85 (2014), 1185. https://doi.org/10.1002/srin.201300336
• 26)   J.  Guo,  M.-D.  Seo,  C.-B.  Shi,  J.-W.  Cho and  S.-H.  Kim: Metall. Mater. Trans. B, 47 (2016), 2211. https://doi.org/10.1007/s11663-016-0697-7