Effect of FeO Content on Foaming and Viscosity Properties in FeO-CaO-SiO₂-MgO-Al₂O₃ Slag System

Abstract

The foaming process is an important part of the electric arc furnace (EAF) steelmaking process. It can promote thermal efficiency and reduce refractory consumption. FeO is a key material used during the foaming process. Unlike with other components used in forming foaming slag, the amount of FeO can be controlled by oxygen or carbon injection. Therefore, adjusting the content of FeO is the most economical foaming mode adopted for the EAFs steelmaking process. In this study, the influence of FeO content on the physical properties of slag was discussed. The melting temperature of the slag was evaluated using three methods: viscosity experiment, using Thermo-Calc simulation software, and high-temperature optical method. The experimental results revealed that the viscosity of slag increases as FeO content decreases. The results also revealed that foam height ratio exhibited a positive correlation with the viscosity of slag.

1. Introduction

The electric arc furnace (EAF) steelmaking process involves the consumption of high amounts of electric energy. As scraps are being increasingly recycled, the demand for EAFs has gradually increased. Thus, it has become increasingly important to save energy during the EAF steelmaking process. Several important techniques such as scrap preheating, stirring, oxygen or carbon injection, usage of the air-tight system, usage of hot charge (e.g., hot metal), and slag foaming are used to reduce electric energy consumption in EAFs. Among these techniques, foaming slag is the most useful and economically beneficial.

The slag foaming process significantly affects the thermal efficiency and refractory lining/electrode consumption of EAFs. These effects have been confirmed by many researchers [1,2,3,4]. In addition, the foaming process can remove oxygen from FeO present in slag. It can also increase the recycling rate of iron. The main factor affecting the foaming process is CO gas generated during melting. It is produced when the carbon content of steel reacts with FeO in slag. The carbon content in molten steel can be controlled by injected oxygen and carbon (with FeO in the slag), as shown below:

$$\mathrm{C} + \left(\mathrm{FeO}\right) = \mathrm{CO} + \mathrm{Fe}$$

(1)

$${2\mathrm{C} + \mathrm{O}}_{2 }= 2\mathrm{CO}$$

(2)

The above reactions occur at the steel–slag interface. CO gas enters the slag and increases its volume for efficient foaming. The duration for which the CO gas remains in the slag affects its foaming ability. It can be explained by the foaming index (Σ) as follows [5,6,7,8,9]:

$$\mathsf{\Sigma }=\frac{\mathsf{\Delta }\mathrm{h}}{\mathrm{j}}=\mathrm{k}\frac{\mu }{\sqrt{\mathsf{\rho }·\mathsf{\sigma }}}$$

(3)

where Δh (cm) is foaming height, j (cm/s) is superficial gas velocity, k is a constant, μ represents viscosity of the slag, ρ denotes density of the slag, and σ is surface tension of the slag. Therefore, viscosity is one of the major factors affecting the slag foaming process. Many studies have been carried out with a focus on the effect of the foaming process or the viscosity. Luz et al. [10] reported the effect of viscosity on the foaming process. Bhoi et al. [11] reported that FeO reduction could increase the viscosity of slag and promote foaming ability. Kim et al. [12] evaluated the foaming index and foaming constant in different slag systems. Lee et al. [13] reported in detail the viscous behavior of slag with varying FeO content. Martinsson et al. [14] investigated the relationship between the viscosity and foaming ability of the CaO-SiO₂-FeO slag systems. Although the effects of some of these factors on the slag foaming process have been previously investigated, only a few studies have reported the influence of viscosity changes (with calcium silicate solidified phase particles) on the foaming process. Few researchers have analyzed the experimental data and described in detail the effect of viscosity and foaming on the slag foaming process [11,14,15].

Based on these aspects, this work aims to discuss the effects of viscosity changes and solidified particles on the slag foaming phenomenon.

2. Materials and Methods

2.1. Slag Sampling and Characterization

The slag samples used in this study were prepared from mixtures of analytical grade Al₂O₃, CaO, SiO₂, FeO, and MgO.

Table 1. The composition of the slag samples.

Sample	Chemical Composition (wt.%)
	CaO	SiO2	FeO	Al2O3	MgO	CaO/SiO2
A1	53.0	24.2	9.9	6.2	6.7	2.2
A2	45.0	21.3	20.0	6.6	7.1	2.1
A3	39.7	18.7	29.6	5.3	6.7	2.1
B1	49.5	19.3	20.6	5.0	5.7	2.6
B2	42.0	15.6	30.2	7.2	5.1	2.7
B3	32.6	12.4	43.2	7.1	4.7	2.6

shows the chemical composition of the slags. The slags were pre-melted in an induction furnace and quenched into water. The chemical compositions of the slags were analyzed using a WDXRF (Wavelength-dispersive X-ray Fluorescence Spectrometer, Rigaku supermini 200, Rigaku, Tokyo, Japan). The Fe²⁺ chemical titration method was followed for characterization. Thermo-Calc software (Slag 3.2 database, Thermo-Calc 4.0, Thermo-Calc Software AB, Stockholm, Sweden) was used to simulate the solid fraction and melting temperature of the slag samples.

The characteristics of the samples were analyzed at high temperatures using a viscometer (Brookfield DV IIIRV, AMETEK Brookfield, Middleboro, MA, USA), and high-temperature optical images were recorded. The data were analyzed using Thermo-Calc software, and the melting temperature and flow properties of the slags were compared.

The setup followed for viscosity analysis is shown in Figure 1. The experiments were conducted in a Si-Mo electric resistance furnace (SJ High Technology Company, Taipei, Taiwan). The furnace was heated at a rate of 10 °C/min until the temperature reached 1350 °C. Following this, the sample was heated at 2 °C/min until the temperature reached 1600 °C. Molten slag was prepared under a high-purity Ar atmosphere in a molybdenum crucible (inside diameter: 45 mm, outside diameter: 55 mm, height: 120, inside depth: 110 mm). The flow rate was maintained at 5 L/min. The molybdenum rotor was slowly put into the crucible after the slag melted (rotor diameter: 30 mm). The distance was set at 5 mm from the bottom of the crucible. The rotor speed was set at 20 rpm, and the viscosity was measured in the temperature range of 1350–1600 °C.

The high-temperature optical image was recorded in a Si-Mo electric resistance furnace (SJ High Technology Company, Taipei, Taiwan). The setup followed for the high-temperature optical image is shown in Figure 2. The slag samples were crushed and sieved to produce powdered samples (75 μm). The powder was pressed into a cylinder (radius: 3 mm and height: 3 mm) and placed on a platinum board. The furnace was heated at a rate of 10 °C/min until the temperature reached 1350 °C. Following this, the heating rate was maintained at 2 °C/min until the temperature of 1595 °C was reached. Optical imaging equipment was used to continuously record changes in sample height and temperature. The experiment was stopped at a temperature of 1595 °C. Sample height was analyzed to define melting behavior.

2.2. Foaming Experiment

The foaming experiment was performed in an induction furnace (Jia Shing Electricuty Construction Co., Kaohsiung, Taiwan). The experimental setup is shown in Figure 3. A slag sample (100 g) was put in the molybdenum crucible and heated to a specified temperature. Two B-type thermocouples, in contact with the crucible wall and the bottom of the sample, respectively, were used to record the temperature. After the sample melted, the electrode was used to measure the initial liquid height. Under a constant flow rate, the electrode was dropped again, and foaming height was measured.

3. Results and Discussion

3.1. Thermo-Calc Phase Diagram Simulation

In this study, the FeO-CaO-SiO₂-Al₂O₃-MgO slag system was simulated using Thermo-Calc software. The simulation results are provided in Figure 4, where the x-axis represents the CaO concentration, and the y-axis represents the temperature. MgO and Al₂O₃ contents were set at 5%. The temperature was maintained in the range of 1500–1800 °C. The results revealed that when C/S value was low, the major precipitate phase was Ca₂SiO₄. When C/S value increased, Ca₂SiO₄ transformed into Ca₃SiO₅. At high C/S values, purely solid CaO formed the major precipitate phase. The saturation level of MgO decreased as FeO weight fraction increased. In addition, according to the MgO saturation equation provided in the literature of Luz [16], the MgO content of this work in all samples were below the saturation level at temperatures between 1400 and 1600 °C.

The dotted lines in Figure 4 represent the melting points of the different samples under different C/S values. The results are shown in

Table 2. The solid fraction and melting point of each sample simulated by Thermo-Calc software.

Sample	Solid Fraction (%)							Melting Point (°C)
	1350 °C	1400 °C	1450 °C	1500 °C	1550 °C	1600 °C	1650 °C
A1	-	-	-		28.8	21.6	12.7	1704
A2	-	25.6	21.6	14.2	5.1	-	-	1575
A3	19.7	11.9	3	-	-	-	-	1465
B1	-	-	-	-	22.6	11.1	-	1601
B2	-	25.7	16.6	6.5	-	-	-	1503
B3	16.4	6.5	0.2	-	-	-	-	1489

. In samples with higher FeO content (A2, A3, B2, and B3), the solid fraction decreased as FeO content increased. A1 and B1 maintained a solid fraction at contents >20%.

3.2. Viscosity Analysis and High-Temperature Optical Image Observation

Figure 5 and

Table 3. Viscosity of each sample at different temperatures.

Sample	Viscosity (cP)
	1400 °C	1450 °C	1475 °C	1500 °C	1550 °C	1575 °C
A1	-	-	-	-	1700	1150
A2	-	1139	975	562	82	36
A3	28	28	27	25	22	-
B1	-	-	-	-	1358	870
B2	-	1134	746	183	28	26
B3	454	64	59	43	36	34

present the viscosity analysis results. In this study, the slag with C/S = 2.1–2.2 and C/S = 2.6–2.7 was used to control the viscosity at a temperature of 1550 °C with varying FeO contents. According to the results of Thermo-Calc simulation software, both FeO and Basicity will affect the solidified phase content, which was also reflected in viscosity data. The viscosity decreased with increasing FeO content at a constant temperature and constant basicity (C/S). The results also revealed that viscosity increased as the C/S value increased at constant FeO content. The viscosity of A1 and B1 were above 1300 cP at 1550 °C. The viscosities of A2, A3, B2, and B3 at 1460 °C were <100 cP. Figure 6 presents the results of studying the FeO-CaO-SiO₂-Al₂O₃-MgO slag system under different C/S values and varying FeO contents. The viscosities tended to decrease with increasing FeO content (under all C/S values investigated). According to viscosity measurement studies reported by Seok et al. [17] for the FeO-CaO-SiO₂-Al₂O₃-MgO slag system, the addition of FeO can promote the formation of low melting point compounds such as fayalite (Fe₂SiO₄). This is because free Fe²⁺ and O²⁻ were introduced into the silicate melts. The mixing entropy increased when FeO was added. Lee et al. [18] reported that FeO content influences the Gibbs free energy of FeO-CaO-SiO₂-Al₂O₃-MgO slags.

High-temperature optical images were recorded. DIN 51,730 was used to determine the melting behavior of the slag samples. The 1/2 height and 1/3 height in Figure 7 represent the half ball temperature and flow temperature, respectively. The results are presented in Figure 7 and

Table 4. Results of the high-temperature optical image observation, temperature of slags at viscosity 100 cp, and melting temperature simulated by Thermo-Calc software.

Sample	Melting Temperature (°C)
	Half Ball Temp.	Flow Temp.	Fully Liquid	Thermo-Calc	Vis-100 cP
A1	1555	1560	>1595	1704	>1600
A2	1420	1465	1535	1575	1538
A3	1355	1370	1405	1465	<1370
B1	1570	1605	>1608	1601	>1600
B2	1435	1475	1500	1503	1507
B3	1335	1350	1420	1489	1431

. The melting temperature was simulated using Thermo-Calc software, and the temperature at which viscosity was 100 cP was recorded following the viscosity analysis procedure outlined herein. The temperature at which liquid slag samples were fully formed is labelled as the “flat” temperature. At this temperature, precipitates were absent in the slag samples. At the “flat” or melting temperature, the viscosity of the slag sample was ≈100 cP.

3.3. Viscosity Analysis and High-Temperature Optical Image Observation

The foaming index can also be expressed as follows [19]:

$$\mathsf{\Sigma }\equiv \frac{\mathsf{\Delta }\mathrm{V}}{\mathrm{Q}},$$

(4)

where ΔV denotes the increased volume of foaming slag (m³) and Q represents the rate of gas production (m³/sec). As the cross-sectional area of the furnace body remains almost constant, the foaming index is divided by the cross-sectional area. Foaming can be described as Formula (3). The formula of the foaming index reveals a positive correlation between viscosity and foaming height. The slag samples A1 and B1 were characterized by high viscosity (compared to the other slag samples). Thus, A1 and B1 could potentially exhibit better foaming ability. Figure 8 and Figure 9 reveal the foaming height of A1 and B1, respectively. The X index represents voltage, and the Y index represents the internal height of the crucible. Initially, the electrode was placed in a nonconductive area where the voltage was 800 mV. As the electrode dropped to contact the liquid surface, the voltage rose to 5000 mV. The black dotted line represents the initial liquid surface height. The red and blue lines represent the foaming height measured for the first and second times, respectively.

According to the work of Ito et al. [5,6], foaming index and foaming height are positively correlated. In this study, gas velocity was constant for each foaming experiment. Therefore, the foaming ability in this study could show as foam height ratio (△H/H₀) [10].

Table 5. Foam height ratio of each sample at different temperatures.

Sample	Foam Height Ratio (△H/H0)
	1500 °C	1550 °C	1560 °C	1575 °C	1585 °C	1600 °C
A1	-	-	1.021	0.965	-	0.742
A2	0.615	0.719	-	-	-	-
A3	0.731	-	-	-	-	-
B1	-	-	1.097	0.927	0.851	-
B2	0.696	0.598	-	-	-	-
B3	0.377	-	0.318	0.318	-	-

shows the foam height ratio of slag samples in the temperature range of 1500–1600 °C. The foam height ratio of A1 and B1 was 1.021 and 1.097, respectively, at 1560 °C. This was because solid particles with a high melting point formed precipitates in the slag. The particles increased the viscosity of the slag samples and promoted the foaming ability of slags. When FeO was used as a co-solvent, the viscosity increased with increasing FeO content. However, temperature also affected the foam height ratio. When the temperature increased, more precipitated particles melted, decreasing the foaming height ratio value.

The correlation between viscosity and foam height ratio for the slag samples used in this study are shown in Figure 10. The foam height ratio was proportional to viscosity in the range of 25–1810 cP. The relation can be expressed as follows:

$$\raisebox{1ex}{$\mathsf{\Delta }\mathrm{h}$}\!\left/ \!\raisebox{-1ex}{${\mathrm{h}}_0$}\right.= 2\times {10}^{-4}\mu +0.6552$$

(5)

Figure 11 shows the foaming mechanism. The bubbles in the low-viscosity slag easily overflowed because of the lack of solid particles. As the number of solid particles increased, the resistance of the bubbles increased, making it difficult for the bubbles to escape. This enhanced foaming ability.

4. Conclusions

Based on the results, the following conclusions can be drawn:

• The viscosity of the slag sample increased as FeO content decreases. At a slag composition of 9.9% FeO, 53% CaO, 24.2% SiO₂, 6.2% Al₂O₃, and 6.7% MgO, the maximum viscosity was recorded at 1550 °C. When FeO content was >29.6%, the viscosity rapidly dropped to <100 cP.
• The results from the foaming experiment revealed that for the FeO-CaO-SiO₂-MgO-Al₂O₃ slag system in the viscosity range of 25–1810 cP, the foam height ratio and viscosity exhibited a linear relationship which can be expressed as: $\raisebox{1ex}{$\mathsf{\Delta }\mathrm{h}$}\!\left/ \!\raisebox{-1ex}{${\mathrm{h}}_0$}\right.= 2\times {10}^{-4}\mu +0.6552$
• Viscosity increased with increasing solid phase fraction in the slag. The experimental results revealed that the slag-containing solid fraction exhibited better foaming ability. In the six slag samples studied, the slag sample B1 (20%FeO-49.5% CaO-19.3% SiO_2–5% Al₂O_3–5.7% MgO) exhibited the best foaming ability, and the foam height ratio was 109.7%.