Influence of Blast Furnace Burden with Different TiO₂ Contents on the Process of Reduction and Slag Formation in Cohesive Zone

Abstract

Vanadium and titanium magnetite (VTM) possesses considerable value in comprehensive mining and utilization. The varying TiO₂ content in the blast furnace burden during smelting directly influences the formation, location, thickness, permeability, and heat exchange within the cohesive zone of the blast furnace. This paper employs the ‘Qisunny method’ to simulate the reduction droplet of blast furnace burden under the condition of different TiO₂ content in the theoretical composition of blast furnace hearth slag, and systematically researches the change rule of soft-melting and dripping performance of blast furnace burden. It systematically investigates the factors influencing the soft-melting and dripping performance of the blast furnace burden. The results indicate that, as the TiO₂ content in the theoretical composition of blast furnace final slag increases from 7.5 wt% to 25.5 wt%, several changes occur: the softening zone (ΔT_s) widens, while the melting zone (ΔT_m) narrows. Additionally, the temperature range of the cohesive zone (ΔT_c) experiences a slight widening, and the cohesive tends to shift downward. The reduction of iron oxides in the blast furnace burden occurs primarily in the softening zone, whereas the reduction of titanium oxides begins in the melting zone. Furthermore, the simulated operation line gradually deviates from the ideal operation line, necessitating an increase in airflow and coke ratio for optimal performance in the actual operation of the blast furnace.

1. Introduction

Vanadium and titanium magnetite ore is abundant in iron, vanadium, and titanium, along with trace amounts of chromium, cobalt, and other rare metals, which confer significant comprehensive mining and utilization value.^1,2,3) These ores are primarily found in countries such as Russia, Canada, South Africa, the United States, Brazil, and Finland. In China, significant deposits are located in the western regions of Panzhihua, Chengde, Ma’anshan, and Chaoyang in Liaoning province, where they are widely distributed and rich in reserves.^4,5,6) The blast furnace (BF) process predominates in the smelting of vanadium and titanium ores, being utilized by firms such as Chengsteel, Pangang, and Jianlong Group, as well as several Russian steel mills. In comparison to conventional iron ore smelting, the blast furnace smelting of vanadium and titanium ores exhibits unique interactions between coke and titanium oxides in the slag.^7,8) When coke reaches the cohesive zone (above 1000°C), it undergoes soft-melting and engages in contact infiltration, leading to a significant reaction that produces high melting point carbides TiC and nitrides TiN. Due to the large amount of solid C and nitrogen in the blast furnace, it is difficult for TiC and TiN to exist as separate phases in the slag, but rather form a solid solution Ti(C,N).⁹⁾ This phenomenon increases the viscosity of the slag, complicating the separation of slag from iron and hindering the drop of the latter. These adverse factors are closely related to the formation and characteristics of the cohesive and dropping zones.^10,11,12,13) It can be seen that the varying TiO₂ content in the blast furnace burden directly influences the formation, location, thickness, permeability, and heat exchange of the cohesive zone, as well as the formation and properties of the initial slag, thereby affecting the overall slagging system and thermal regime within the blast furnace.^14,15,16)

The researchers investigated the soft-melting and dripping performance of vanadium-titanium ore by examining different compositions, charging sequences, and burden structures. During the reduction process in the blast furnace’s lumpy zone, cohesive zone, and dropping zone, both iron and titanium oxides are reduced step by step. The presence of titanium prolongs the reduction path of iron oxides, leading to significant reductions of vanadium and chromium oxides during the soft-melt dropping process and the interaction between slag and iron.^15,16,17,18) As the TiO₂ content in the burden increases, the cohesive zone shifts downward, resulting in decreased permeability.¹⁹⁾ However, previous studies have largely focused on single burdens like sinter or pellets, neglecting the interactions of various burdens under actual blast furnace conditions. Consequently, they fail to accurately represent the dynamic migration and transformation processes of valuable metals within the composite burden in the blast furnace. Results from the anatomy of the experimental blast furnace indicate that TiC and TiN were generated in the slag in the lower part of the cohesive zone, while vanadium and titanium were already present in the metallic iron. The mass fraction of Ti(C,N) in the slag at the wind-port plane, along with the mass fractions of vanadium and titanium in the metallic iron, reached maximum values, subsequently reduced by partial oxidation while passing through the oxidizing zone of the gyratory zone.²⁰⁾ This indicates that the cohesive and dropping zones are critical areas for significant changes in slag-iron composition and temperature, as well as the generation of Ti(C,N).^21,22,23,24) However, conventional upper and lower operating methods of the blast furnace cannot directly affect this area. The impact of raw fuel quality, burden structure, and operational systems on this zone is only indirectly reflected through changes in technical and economic indicators, molten iron quality, and hearth slag performance over extended smelting practices. Therefore, it is essential to study the behavior of blast furnace burdens with varying TiO₂ contents during the soft-melt dropping process, particularly regarding the changes in titanium oxide structure and its specific influences and mechanisms of action on the ore-coke integrated burden.

This paper investigates the raw material conditions and burden structure of vanadium and titanium ores in iron and steel plants by employing the “Qisunny method” to conduct experiments on reducing molten droplets of blast furnace burdens with varying TiO₂ contents. The objective is to elucidate the reduction mechanisms of vanadium and titanium magnetite iron and titanium oxides through microstructural characterization and to analyze the smelting processes of the blast furnace in conjunction with the theory of the Rist operation line. The findings from this study offer valuable guidance for optimizing the smelting operations of vanadium and titanium ore in blast furnaces. They provide essential information for maintaining stable and efficient furnace conditions and refining operational adjustments. Moreover, the study expands the understanding of the softening and melting mechanisms of vanadium and titanium ores, as well as the coke solution loss and degradation mechanisms. It also supports the blast furnace’s capacity to increase the proportion of vanadium and titanium ores while conserving coke and reducing overall coke consumption. Additionally, this research contributes to establishing a scientific evaluation system for the burden of vanadium and titanium ore in blast furnaces.

2. Experimental Materials and Program

2.1. Experimental Raw Materials

The iron-bearing raw materials utilized in the experiments consisted of three types of sinter and pellet with varying TiO₂ contents, as detailed in Table 1. Sinter 1 and Pellet 1 were sourced from the Chengde area of Hebei Province, China, exhibiting TiO₂ contents of 1.825 wt% and 6.194 wt%, respectively. In contrast, Sinter 2 and Pellet 2 were obtained from the Panzhihua area, Sichuan Province, China, with TiO₂ contents of 5.801 wt% and 9.21 wt%, respectively. Additionally, Sinter3 is a sintered ore independently fired in the laboratory, with a titanium content of 38.88 wt% and a low iron content, which is used to increase the TiO₂ content in the integrated charge by adding this ore. Pellet 3 was sourced from the Zhanjiang area of Guangdong Province, China, and is characterized as a titanium-free pellet.

Table 1. Iron-containing raw material composition, wt%.

Compositionn	TFe	FeO	CaO	MgO	Al2O3	SiO2	TiO2
Sinter 1	63.16	13.84	10.159	3.284	2.355	5.786	1.825
Sinter 2	53.03	8.46	11.31	2.63	4.23	7.45	5.801
Sinter 3	36.42	5.07	1.043	5.601	2.036	4.629	38.882
Pellet 1	56.68	1.55	1.419	1.705	3.314	5.927	6.194
Pellet 2	52.015	2.65	0.962	4.565	5.215	6.753	9.21
Pellet 3	64.6	0.4	2.4	0.68	0.88	3.12	–

The coke used in the experiments adhered to the Chinese national standard (GB/T 1996–2017) and had a reactivity index (CRI) of 25.2%, a post-reaction strength (CSR) of 62.3%, and a fixed carbon content of 85.78%. The industrial analysis and ash composition are presented in Table 2.

Table 2. Coke industrial analysis and ash composition, wt%.

Cad	S content	Vad	Aad(14.00)						∑
			FeO	CaO	SiO2	MgO	Al2O3	Others
84.98	0.48	1.49	0.14	0.5	7.52	0.14	2.77	2.93	100.00

2.2. Experimental Equipment and Programs

To simulate the soft-melting process of blast furnace burden into slag, experiments were conducted using the “Qisunny method”. Compared with the traditional soft-melting and dripping test, this method uses coke-ore-coke layered test material, and simulates the correlation behaviors and characteristics between reduction, softening, melting, and coke dissolution loss of ore in the blast furnace through the loaded-thermogravimetric-gas-analytical linkage experimental device to carry out the method of testing and evaluating the coupled ore-coke performance. The coupled ore-coke metallurgical performance evolves sequentially in four zones, namely, indirect reduction zone, indirect and direct reduction and softening zone, molten direct reduction zone, and dropping direct reduction, and the performance evaluation indexes include the temperature and zones, indirect and direct reduction degree, and the permeability and eigenvalue of the melting zone.^25,26,27,28) The maximum control temperature of the test equipment is up to 1600°C. Mass change in the specimen was monitored in real-time using an electronic balance (ES30K-ID) with a maximum capacity of 30 kg and an accuracy of 0.1 g. An infrared gas analyzer (Gasboard-3100) was employed to measure the volume fractions of CO and CO₂ in the exhaust gases. Additionally, a displacement sensor (MIRAN KTR11) recorded changes in the height of the ore layer, while a differential pressure transmitter (F-CFEP) measured the pressure difference of the reducing gas flowing through the test material. A schematic of the device is presented in Fig. 1.

Fig. 1. The device for measuring soft-melting performance. (Online version in color.)

This experiment simulates field production conditions, with a fixed binary basicity of the blast furnace hearth slag (R=1.10), a coke ratio of 350 kg/t, a coke-butyl ratio of 20 kg/t, and a coal ratio of 160 kg/t. The TiO₂ content in the mixed burden was adjusted by adding different types of vanadium and titanium ores, allowing for a theoretical TiO₂ composition in the blast furnace hearth slag ranging from 7.5 to 25.5 wt% (with a step size of 4.5 wt%). The study aims to investigate how varying TiO₂ concentrations in the mixed burden influence the softening and dropping characteristics of the materials. The ore distribution scheme is presented in Table 3, while the theoretical chemical composition of the mixed burden can be found in Table 4.

Table 3. Experimental program.

Program	Iron-containing burden ratios (100%)/wt%						Theoretical composition of the hearth slag
	Sinter 1	Sinter 2	Sinter 3	Pellet 1	Pellet 2	Pellet 3	w(TiO2)/wt%	Basicity (R=CaO/SiO2)/–
1	71	–	–	13	–	16		7.5	1.1
2	63.5	16.5	–	9	11	–		12	1.1
3	50	31	3	–	16	–		16.5	1.1
4	40.8	39.2	7.5	–	12.5	–		21	1.1
5	30	49	12.5	–	8.5	–		25.5	1.1

Table 4. Theoretical chemical composition of mixed burden.

Program	TFe	FeO	CaO	MgO	Al2O3	SiO2	TiO2
1	62.55	10.09	7.78	7.78	2.24	5.38	2.1
2	59.68	10.61	8.55	8.55	3.07	6.18	3.69
3	57.43	10.12	8.77	8.77	3.38	6.42	5.35
4	55.79	9.67	8.78	8.78	3.42	6.47	7.09
5	53.91	9.15	8.80	8.80	3.48	6.54	9.03

The graphite crucible used in the experiment has an inner diameter of Φ75 mm, an inner height of 160 mm, and a bottom hole diameter of Φ10 mm. The experimental raw materials were designed to simulate the laminar distribution of blast furnace coke. Each experiment used 500 g of mixed ore with a particle size of approximately 10.0–12.5 mm, along with 175 g of coke. The coke was divided into two particle size categories: 75 g of coke with a particle size of about 13–15 mm, which had lost 25 wt% by coke gasifier dissolution, was placed on the bottom of a graphite crucible to simulate coke in the soft-melting dropping zone of a blast furnace, where the dissolution reaction occurs;²⁹⁾ 100 g of insoluble coke with a particle size of about 18–20 mm was placed in the upper layer to represent the furnace coke. The ore was situated in the middle layer.

To prevent oxidation of the raw materials during the heating process, nitrogen (N₂) was introduced for protective purposes. The heating schedule consisted of a rate of 10°C/min from 0 to 900°C, followed by a rate of 3°C/min from 900 to 1020°C, and a final rate of 5°C/min above 1020°C. The experiment was concluded when the shrinkage of the specimen ceased to change significantly. At the end of the experiment, the reducing gas was replaced with nitrogen to protect the specimens as they cooled to room temperature, after which they were removed for analysis. Following the primary slag generation experiment, the residue in the crucible was removed and separated to obtain a residual slag sample. The experimental warming and testing conditions are detailed in Table 5.

Table 5. Ore softening and melt drop experimental conditions.

Warming time/min	40	50	40	>120
Load/(kg·cm−2)	0.5	0.5	1.0
Gas composition and flow rate	N2 100% 3 L/min	N2 60% 9 L/min CO 26% 3.9 L/min CO214% 2.1 L/min	N2 70% 10.5 L/min CO 30% 4.5 L/min
Heating speed	10°C/min Up to 400°C	10°C/min Up to 900°C	3°C/min Up to 1020°C	5°C/min Dropping

Based on the molten drop experiment, which encompasses the processes of ore loading, heating, reduction, softening, melting, and dropping, various performance parameters of the burden were examined. These parameters include the softening zone (ΔT_s), melting zone (ΔT_m), cohesive zone (ΔT_c), permeability index (K), characteristic value (D), and degree of reduction (R).

In this experiment, the softening start temperature (T_10%) is defined as the temperature at which the shrinkage of the mineral layer reaches 10%. With the increase of temperature and the reduction reaction, the ore from the solid state gradually occurs softening and molten drops, the reduction from gas-solid reaction to solid-liquid-gas three-phase reaction, that is, the molten slag in contact with the coke, FeO and C occur in the molten direct reduction (MDR), or FeO-containing liquid slag drops flow through the surface of the coke to occur in the drop of the direct reduction (DDR), the reaction formula is (FeO)+C=[Fe]+CO.

When the charge enters the direct reduction zone, the molten ore layer has good contact with the coke, and the melting and reduction proceed simultaneously, which corresponds to the ΔP, ΔP/H_O,T, dRc/dt curves begin to rise steeply in parallel; as the temperature rises further, the melted slag and iron begin to drip and have better contact with the coke, and the dripping and the reduction proceeds synchronously, which corresponds to the ΔP curve falls back to a low level, ΔP/H_O,T and dRc/dt curves rise steeply again. Consequently, the softening end or melting start temperature (T_m,i) is identified by a steep rise in the differential pressure (ΔP) curve. Conversely, the melting end or dropping start temperature (T_m,f) corresponds to the temperature at which the ΔP curve declines from its peak to a lower level, followed by a significant increase in the differential pressure per unit height of the ore layer (ΔP/H_O,T) curve.²⁶⁾ The softening zone and melting zone, as well as the cohesive zone temperature interval, are calculated according to Eqs. (1), (2), and (3), respectively.

  

$$\mathrm{\Delta }{\mathrm{T}}_{\mathrm{s}}=({\mathrm{T}}_{\mathrm{m},\mathrm{i}}-{\mathrm{T}}_{10\%})$$(1)

  

$$\mathrm{\Delta }{\mathrm{T}}_{\mathrm{m}}=({\mathrm{T}}_{\mathrm{m},\mathrm{f}}-{\mathrm{T}}_{\mathrm{m},\mathrm{i}})$$(2)

  

$$\mathrm{\Delta }{\mathrm{T}}_{\mathrm{c}}=({\mathrm{T}}_{\mathrm{m},\mathrm{f}}-{\mathrm{T}}_{10\%})$$(3)

The layer height of the mixed furnace burden was recorded in real time by a displacement sensor during the experiment. The shrinkage of the material layer is calculated from the height of the ore layer as follows:

  

$$\begin{array}{c}\text{S}=\frac{{\text{H}}_{\text{600}}-{\text{H}}_{\text{T}}}{\text{H}}\times \text{100}\end{array}$$(4)

In the formula, S represents the shrinkage of the mixed burden at temperature T, %; H₆₀₀ denotes the displacement at 600°C, mm; H_T indicates the displacement at temperature T, mm; and H refers to the initial layer height of the mixed burden, mm.

The permeability index K is defined as the pressure difference of a reducing gas passing through a unit height of the material layer and is calculated as follows:

  

$$\begin{array}{c}\mathrm{K}=\mathrm{\Delta }\text{p}/{\text{H}}_{\text{o,T}}=\mathrm{\Delta }\text{p}/{\text{(H}}_{\text{600}}-{\text{H}}_{\text{T}}\text{)}=\frac{\mathrm{\Delta }\text{p}}{(\text{1}-\text{S/100})\text{H}}\end{array}$$(5)

In the formula, Δp represents the pressure difference between the inlet and outlet, corresponding to the pressure differential when the gas passes through the material layer, kPa. H_o,T denotes the height of the material layer at temperature T, mm.

The characteristic value (D) is an integral measure of the gas permeability of the burden throughout the experimental process and is utilized to characterize the overall gas permeability of the burden, as calculated in Eq. (6).

  

$$\mathrm{D}={\int }_{{\text{T}}_{\text{10\%}}}^{\text{Tm},\text{f}}\text{K}\mathrm{d}\text{T}$$(6)

The microstructure of the slag samples was examined using a Zeiss polarizing microscope (Cambridge Q500; Leica Microsystems; Wetzlar, Germany) and an S-3400N scanning electron microscope (JEOL Ltd., Tokyo, Japan). The chemical composition of both the raw materials and post-experimental samples was determined using an X-ray fluorescence spectrometer (XRF) (PANalytical Axios; Malvern Panalytical; Almelo, The Netherlands) through the X-fluorescence full-scan rapid analysis method. The testing procedure was conducted as follows: (1) grind the sample to a fine powder; (2) press the sample powder into thin slices using a tablet press; (3) place the prepared sample slices into the sample tray of the XRF spectrometer and configure the instrument parameters, including voltage, current, and duration; (4) activate the instrument to commence measurement, during which it automatically records the characteristic X-ray wavelengths and intensities of the samples; (5) the instrument then converts the elemental contents based on the intensities of the elemental spectral lines.

3. Experimental Results

3.1. Cohesive Zone

Figure 2 illustrates a schematic diagram of the softening and melting zones corresponding to the theoretical composition of blast furnace hearth slag with varying TiO₂ contents. The experimental results indicate that, as the TiO₂ content in the theoretical composition of the slag increases from 7.5 wt% to 25.5 wt%, the softening start temperature (T_10%) of the mixed burden does not exhibit a clear trend. In contrast, the softening end temperature (T_m,i) rises gradually from 1184°C to 1247°C, and the melting end (or drip start) temperature (T_m,f) increases slightly from 1407°C to 1418°C. Moreover, the softening zone (ΔT_s) temperature interval expands from 84°C to 136°C, whereas the melting zone (ΔT_m) temperature interval decreases from an initial value of 223°C to 171°C. Additionally, the total cohesive zone (ΔT_c) temperature interval broadens, indicating a downward trend in the cohesive zone.

Fig. 2. Schematic diagram of softening and melting zones for different TiO₂ contents in the theoretical composition of blast furnace hearth slag. (Online version in color.)

3.2. Air Permeability and Characteristic Value

Figure 3 illustrates the effect of permeability (K) and the characteristic values (S) associated with varying TiO₂ contents in the theoretical composition of blast furnace hearth slag. As depicted in the figure, when the TiO₂ content in the theoretical composition of the slag increases from 7.5 wt% to 25.5 wt%, the permeability index rises from 0.023 to 0.057, indicating a deterioration in the permeability of the burden. Specifically, the permeability index increases from 0.023 to 0.041 as the TiO₂ content rises from 7.5 wt% to 12 wt%, accompanied by a sharp decline in permeability. In contrast, the change in permeability is minimal when the TiO₂ content increases from 12 wt% to 16.5 wt%. Regarding the impact on the characteristic value, the results show a steep increase when the TiO₂ content rises from 7.5 wt% to 12 wt%, alongside a decline in the overall permeability of the burden. Although the characteristic value does not significantly change when TiO₂ content increases from 12 wt% to 16.5 wt%, it continues to rise up to 21 wt%, resulting in a doubling of the characteristic value and a marked deterioration in the permeability of the burden.

Fig. 3. Gas permeability and characteristic value of blast furnace hearth slag with different TiO₂ content in the theoretical composition of the burden. (Online version in color.)

3.3. Reduction

Reduction, defined as the process of capturing oxygen from metal-bound ore, represents a fundamental task in the smelting process. This process employs a reducing agent, a substance with a higher affinity for oxygen, to disrupt the chemical bonds between metal ions and oxygen ions in the ore, thereby releasing the metal elements. Specifically, iron ore reduction refers to the ease with which a gaseous reducing agent can remove oxygen that is bound to iron in iron ore.

The oxygen content of iron oxides in the burden is as follows:

  

$${\text{n(O)}}_{\text{0}}=\frac{{\text{m(O)}}_{\text{0}}}{{\text{M}}_{\text{O}}}=\frac{\text{1}}{{\text{M}}_{\text{O}}}\text{(0}{\text{.429w}}_{\text{TFe}}-\text{0}{\text{.112w}}_{\text{FeO}}\text{)}\cdot {\text{m}}_{\text{ore}}$$(7)

In the formula, n(O)₀ represents the oxygen content of iron oxides in the mixed burden, mol; m(O)₀ represents the mass of oxygen in iron oxides in the mixed burden, g; M_O represents the relative atomic mass of the O atom, g/mol; m_ore denotes the mass of the mixed burden, g; w_TFe denotes the iron content of the mixed burden, %; w_FeO denotes the content of FeO in the mixed burden, %.

The reduction of the mixed burden is expressed as follows:

  

$$\begin{array}{c}{R}_{\text{O}}=\frac{n{\text{(O)}}_{\text{0}}-n{\text{(O)}}_t}{n{\text{(O)}}_{\text{0}}}\end{array}$$(8)

In the formula, Ro represents the mixed burden reduction; n(O)_t denotes the amount of oxygen in the iron oxide when the test is carried out at time t, mol.

The reduction rate of the mixed burden is given by:

  

$$\frac{\mathrm{d}{\mathrm{R}}_{\text{O}}}{\mathrm{d}\mathrm{t}}=-\frac{\text{1}}{\mathrm{n}{\text{(O)}}_{\text{0}}}\frac{\mathrm{d}\mathrm{n}\text{(O)}}{\mathrm{d}\mathrm{t}}$$(9)

In the formula, dRo/dt represents the reduction rate of the mixed burden, min⁻¹; dn(O)/dt denotes the oxygen loss rate of the mixed burden, mol·min⁻¹.

The indirect reduction speed is expressed as:

  

$$\frac{\mathrm{d}{\mathrm{R}}_{\text{O,}\mathrm{I}\mathrm{R}}}{\text{d}\mathrm{t}}=-\frac{\text{1}}{\mathrm{n}{\text{(O)}}_{\text{0}}}\frac{\mathrm{d}\mathrm{n}{\text{(O)}}_{\mathrm{I}\mathrm{R}}}{\mathrm{d}\mathrm{t}}$$(10)

In the formula, dR_O,IR/dt denotes the indirect reduction rate of the mixed burden, min⁻¹; dn(O)_IR/dt represents the oxygen loss rate for indirect reduction of the mixed burden, mol·min⁻¹.

The rate at which CO₂ produced from the reduction of the mixed burden reacts with coke to generate CO is:

  

$$\frac{{\text{dR}}_{\text{C}}}{\text{dt}}=-\frac{\text{1}}{{\text{n(O)}}_{\text{0}}}\frac{\text{dn(C)}}{\text{dt}}$$(11)

In the formula, dR_C/dt represents the dissolution reaction rate of coke, min⁻¹; dn(C)/dt represents the rate of carbon loss of coke, mol·min⁻¹.

The rate of carbon loss during direct reduction is equal to the rate of oxygen loss, which is:

  

$$\frac{{\text{dR}}_{\text{O,DR}}}{\text{dt}}=-\frac{\text{1}}{{\text{n(O)}}_{\text{0}}}\frac{{\text{dn(O)}}_{\text{DR}}}{\text{dt}}=-\frac{\text{1}}{{\text{n(O)}}_{\text{0}}}\frac{\text{dn(C)}}{\text{dt}}$$(12)

In the formula, dR_O,DR/dt denotes the direct reduction rate of the mixed burden, min⁻¹; dn(O)_DR/dt represents the direct reduction oxygen loss rate of the mixed burden, mol·min⁻¹.

The CO flow rate through the mixed burden is:

  

$$\frac{{\text{dn(CO)}}_{\text{in}}}{\text{dt}}=\frac{\text{1}}{{\text{V}}_{\text{m}}}{\text{Q}}_{\text{in}}\mathrm{\phi }{(\text{CO})}_{\text{in}}$$(13)

In the formula, dn(CO)_in/dt represents the CO flow rate through the mixed burden, mol·min⁻¹; Vm represents the molar volume of the gas in the standard state, L·mol⁻¹; Q_in denotes the gas flow rate through the mixed burden, L·min⁻¹; φ(CO)_in denotes the percentage of CO in the gas passed into the mixed furnace burden layer, %.

Excluding the effects of weight loss from graphite devices and the volatile matter of coke, the rate of weight loss can be expressed as:

  

$$\frac{\text{dm}}{\text{dt}}=\frac{{\text{M}}_{\text{O}}\text{dn(O)}}{\text{dt}}+\frac{{\text{M}}_{\text{C}}\text{dn(C)}}{\text{dt}}$$(14)

In the formula, dm/dt represents the rate of weight loss of the mixed burden, g·min⁻¹; dn(O)/dt represents the oxygen loss rate of the mixed burden being reduced, mol·min⁻¹; M_O and M_C are the relative atomic masses of O and C atoms, g/mol.

Let the gas flow rate at the outlet of the device be Q_out, where the CO and CO₂ shares are φ(CO)out and φ(CO₂)out respectively. Based on the oxygen balance of the whole process, the rate of oxygen loss from the reduction of the mixed burden is:

  

$$\frac{\text{dn(O)}}{\text{dt}}=\frac{\text{1}}{{\text{V}}_{\text{m}}}\left\{{\text{Q}}_{\text{out}}\text{[}\mathrm{\phi }{(\text{CO})}_{\text{out}}+\text{2}\mathrm{\phi }{({\text{CO}}_{\text{2}})}_{\text{out}}\text{]}-{\text{Q}}_{\text{in}}\mathrm{\phi }{\text{(CO)}}_{\text{in}}\right\}$$(15)

Based on the carbon balance associated with the reduction and coke dissolution processes during the experiments, the rate of carbon dissolution in coke can be expressed as follows:

  

$$\frac{\mathrm{d}\mathrm{n}\text{(C)}}{\mathrm{d}\mathrm{t}}=\frac{\text{1}}{{\text{V}}_{\text{m}}}\left\{{\text{Q}}_{\mathrm{o}\mathrm{u}\mathrm{t}}\text{[}\mathrm{\phi }{(\text{CO})}_{\mathrm{o}\mathrm{u}\mathrm{t}}+\mathrm{\phi }{({\text{CO}}_{\text{2}})}_{\mathrm{o}\mathrm{u}\mathrm{t}}\text{]}-{\text{Q}}_{\mathrm{i}\mathrm{n}}\mathrm{\phi }{\text{(CO)}}_{\mathrm{i}\mathrm{n}}\right\}$$(16)

Equations (14), (15), and (16) are linked to determine the gas flow rate as it exits the mixed furnace burden layer.

  

$$\begin{array}{c}{\text{Q}}_{\text{out}}=\frac{{\text{V}}_{\text{m}}\left(\frac{\text{dm}}{\text{dt}}\right)+{\text{(M}}_{\text{O}}+{\text{M}}_{\text{C}}{\text{)Q}}_{\text{in}}\mathrm{\phi }{(\text{CO})}_{\text{in}}}{{\text{(M}}_{\text{O}}+{\text{M}}_{\text{C}}\text{)}\mathrm{\phi }{(\text{CO})}_{\text{out}}+{\text{(2M}}_{\text{O}}{\text{+M}}_{\text{C}}\text{)}\mathrm{\phi }{({\text{CO}}_{\text{2}})}_{\text{out}}}\end{array}$$(17)

Equation (17) is substituted into Eq. (15) to derive the oxygen loss rate during the reduction of the mixed burden.

  

$$\begin{array}{l}{\displaystyle \frac{\text{dn(O)}}{\text{dt}}}={\displaystyle \frac{\text{1}}{{\text{V}}_{\text{m}}}}\\ \left\{{\displaystyle \frac{\left[{\text{V}}_{\text{m}}\left({\displaystyle \frac{\text{dm}}{\text{dt}}}\right)+{\text{(M}}_{\text{O}}+{\text{M}}_{\text{C}}{\text{)Q}}_{\text{in}}\mathrm{\phi }{(\text{CO})}_{\text{in}}\right]\text{[}\mathrm{\phi }{\left(\text{CO}\right)}_{\text{out}}+\text{2}\mathrm{\phi }{({\text{CO}}_{\text{2}})}_{\text{out}}\text{]}}{{\text{(M}}_{\text{O}}+{\text{M}}_{\text{C}}\text{)}\mathrm{\phi }{(\text{CO})}_{\text{out}}+{\text{(2M}}_{\text{O}}+{\text{M}}_{\text{C}}\text{)}\mathrm{\phi }{({\text{CO}}_{\text{2}})}_{\text{out}}}}-{\text{Q}}_{\text{in}}\mathrm{\phi }{\text{(CO)}}_{\text{in}}\right\}\end{array}$$(18)

Furthermore, Eq. (17) is substituted into Eq. (16) to calculate the rate of carbon loss due to the reduction of coke.

  

$$\begin{array}{l}{\displaystyle \frac{\text{dn(C)}}{\text{dt}}}={\displaystyle \frac{\text{1}}{{\text{V}}_{\text{m}}}}\\ \left\{{\displaystyle \frac{\left[{\text{V}}_{\text{m}}\left({\displaystyle \frac{\text{dm}}{\text{dt}}}\right)+{\text{(M}}_{\text{O}}{\text{+M}}_{\text{C}}{\text{)Q}}_{\text{in}}\mathrm{\phi }{(\text{CO})}_{\text{in}}\right]\text{[}\mathrm{\phi }{(\text{CO})}_{\text{out}}+\mathrm{\phi }{({\text{CO}}_{\text{2}})}_{\text{out}}\text{]}}{{\text{(M}}_{\text{O}}+{\text{M}}_{\text{C}}\text{)}\mathrm{\phi }{(\text{CO})}_{\text{out}}+{\text{(2M}}_{\text{O}}+{\text{M}}_{\text{C}}\text{)}\mathrm{\phi }{({\text{CO}}_{\text{2}})}_{\text{out}}}}-{\text{Q}}_{\text{in}}\mathrm{\phi }{\text{(CO)}}_{\text{in}}\right\}\end{array}$$(19)

Additionally, Eqs. (7) and (18) are integrated into Eq. (9) to yield the calculated rate of reduction for the mixed burden:

  

$$\begin{array}{l}{\displaystyle \frac{d{R}_O}{dt}}=-{\displaystyle \frac{\text{1}}{n{\text{(}O\text{)}}_{\text{0}}}}{\displaystyle \frac{dn\text{(}O\text{)}}{dt}}=\\ -{\displaystyle \frac{{\displaystyle \frac{\text{1}}{{\text{V}}_{\text{m}}}}\left\{{\displaystyle \frac{\left[{\text{V}}_{\text{m}}\left({\displaystyle \frac{dm}{dt}}\right)+{\text{(M}}_{\text{O}}+{\text{M}}_{\text{C}}\text{)}{Q}_{in}\phi {(\text{CO})}_{in}\right]\text{[}\phi {(\text{CO})}_{out}+\text{2}\phi {({\text{CO}}_{\text{2}})}_{out}\text{]}}{{\text{(M}}_{\text{O}}+{\text{M}}_{\text{C}}\text{)}\phi {(\text{CO})}_{out}+{\text{(2M}}_{\text{O}}+{\text{M}}_{\text{C}}\text{)}\phi {({\text{CO}}_{\text{2}})}_{out}}}-{\text{Q}}_{\text{in}}\mathrm{\varphi }{\text{(CO)}}_{\text{in}}\right\}}{{\displaystyle \frac{\text{1}}{{\text{M}}_{\text{O}}}}\text{(0}\text{.429}{w}_{TFe}-\text{0}\text{.112}{w}_{FeO}\text{)}\cdot {\text{m}}_{\text{ore}}}}\end{array}$$(20)

Similarly, Eqs. (7) and (19) are applied to Eq. (11) or Eq. (12) to ascertain the coke dissolution reaction rate, which concurrently represents the direct reduction rate of the mixed burden.

  

$$\begin{array}{l}{\displaystyle \frac{d{R}_C}{dt}}={\displaystyle \frac{d{R}_{\mathrm{O}\text{,}DR}}{dt}}=-{\displaystyle \frac{\text{1}}{n{\text{(}\mathrm{O}\text{)}}_{\text{0}}}}{\displaystyle \frac{dn\text{(C)}}{dt}}=\\ -{\displaystyle \frac{{\displaystyle \frac{\text{1}}{{\text{V}}_{\text{m}}}}\left\{{\displaystyle \frac{\left[{\text{V}}_{\text{m}}\left({\displaystyle \frac{dm}{dt}}\right)+{\text{(M}}_{\text{O}}+{\text{M}}_{\text{C}}\text{)}{Q}_{in}\phi {(\text{CO})}_{in}\right]\text{[}\phi {(\text{CO})}_{out}+\phi {({\text{CO}}_{\text{2}})}_{out}\text{]}}{{\text{(M}}_{\text{O}}+{\text{M}}_{\text{C}}\text{)}\phi {(CO)}_{out}+{\text{(2M}}_{\text{O}}+{\text{M}}_{\text{C}}\text{)}\phi {({\text{CO}}_{\text{2}})}_{out}}}-Q_{in}\phi {\text{(CO)}}_{in}\right\}}{{\displaystyle \frac{\text{1}}{{\text{M}}_{\text{O}}}}\text{(0}\text{.429}{w}_{TFe}-\text{0}\text{.112}{w}_{FeO}\text{)}\cdot {\text{m}}_{\text{ore}}}}\end{array}$$(21)

Summing the above Eq. (20) yields the formula for total reduction as:

  

$$\begin{array}{l}{\text{R}}_{\text{O}}={\sum }_{\text{t}}{\displaystyle \frac{{\text{dR}}_{\text{O}}}{\text{dt}}}=\\ -{\displaystyle \frac{{\sum }_{\text{t}}{\displaystyle \frac{\text{1}}{{\text{V}}_{\text{m}}}}\left\{{\displaystyle \frac{\left[{\text{V}}_{\text{m}}\left({\displaystyle \frac{\text{dm}}{\text{dt}}}\right)+{\text{(M}}_{\text{O}}+{\text{M}}_{\text{C}}{\text{)Q}}_{\text{in}}\mathrm{\phi }{(\text{CO})}_{\text{in}}\right]\text{[}\mathrm{\phi }{(\text{CO})}_{\text{out}}+\text{2}\mathrm{\phi }{({\text{CO}}_{\text{2}})}_{\text{out}}\text{]}}{{\text{(M}}_{\text{O}}+{\text{M}}_{\text{C}}\text{)}\mathrm{\phi }{(\text{CO})}_{\text{out}}+{\text{(2M}}_{\text{O}}+{\text{M}}_{\text{C}}\text{)}\mathrm{\phi }{({\text{CO}}_{\text{2}})}_{\text{out}}}}-{\text{Q}}_{\text{in}}\mathrm{\phi }{\text{(CO)}}_{\text{in}}\right\}}{{\displaystyle \frac{\text{1}}{{\text{M}}_{\text{O}}}}\text{(0}{\text{.429w}}_{\text{TFe}}-\text{0}{\text{.112w}}_{\text{FeO}}\text{)}\cdot {\text{m}}_{\text{ore}}}}\end{array}$$(22)

Summing the results derived from Eq. (21) yields the formula for the degree of direct reduction as follows:

  

$$\begin{array}{l}{\text{R}}_{\text{O,DR}}=\\ -{\displaystyle \frac{{\sum }_{\text{t}}{\displaystyle \frac{\text{1}}{{\text{V}}_{\text{m}}}}\left\{{\displaystyle \frac{\left[{\text{V}}_{\text{m}}\left({\displaystyle \frac{\text{dm}}{\text{dt}}}\right)+({\text{M}}_{\text{O}}+{\text{M}}_{\text{C}}){\text{Q}}_{\text{in}}\mathrm{\phi }{(\text{CO})}_{\text{in}}\right][\mathrm{\phi }{(\text{CO})}_{\text{out}}+\mathrm{\phi }{({\text{CO}}_{\text{2}})}_{\text{out}}]}{({\text{M}}_{\text{O}}+{\text{M}}_{\text{C}})\mathrm{\phi }{(\text{CO})}_{\text{out}}+({\text{2M}}_{\text{O}}+{\text{M}}_{\text{C}})\mathrm{\phi }{({\text{CO}}_{\text{2}})}_{\text{out}}}}-{\text{Q}}_{\text{in}}\mathrm{\phi }{(\text{CO})}_{\text{in}}\right\}}{{\displaystyle \frac{\text{1}}{{\text{M}}_{\text{O}}}}(\text{0}{\text{.429w}}_{\text{TFe}}-\text{0}{\text{.112w}}_{\text{FeO}})\cdot {\text{m}}_{\text{ore}}}}\end{array}$$(23)

Indirect reduction can be determined by subtracting the rate of direct reduction from the total reduction as follows:

  

$$\begin{array}{c}{\text{R}}_{\text{O,IR}}={\text{R}}_{\text{O}}-{\text{R}}_{\text{O,DR}}\end{array}$$(24)

The reduction of ore material occurs primarily in the following order,indirect reduction (IR) in the lumpy zone, coexistence of indirect and direct reduction (IR+DR) in the softening zone, melt direct reduction (MDR) in the melting zone, and drop reduction (DDR) in the dropping zone. Figure 4 illustrates the degree of reduction of the burden with varying TiO₂ content across different regions within the theoretical composition of the blast furnace hearth slag. As the TiO₂ content in the hearth slag increased from 7.5 wt% to 25.5 wt%, the indirect reduction (R_i(IR)) in the block zone decreased from 48.80% to 46.10%. Meanwhile, both indirect reduction (R_i(IR+DR)) and direct reduction (R_d(IR+DR)) in the softening zone exhibited a gradual increase, rising from 1.30% and 3.09% to 1.99% and 6.48%, respectively. Conversely, in the melting zone, both indirect reduction (R_i(MDR)) and direct reduction (R_d(MDR)) showed a decreasing trend, declining from 2.45% and 24.4% to 1.11% and 22.3%, respectively.

Fig. 4. Theoretical compositions of blast furnace hearth slag with different TiO₂ contents in different regions of the reduction of the burden. (Online version in color.)

3.4. Microscopic Composition

The mineral composition and structure of the burden are the primary factors influencing its pyrometallurgical properties. After the experiment, undripped material was sampled and ground to a particle size of less than 200 mesh. The results of the X-ray diffractometer analysis are presented in Fig. 5, while the results of the scanning electron microscopy (SEM) and energy-dispersive X-ray spectroscopy (EDS) inspections are shown in Figs. 6, 7, 8, 9, 10.

Fig. 5. XRD patterns of soft-melting undrops from experiments with different TiO₂ contents in the theoretical composition of blast furnace hearth slag.

Fig. 6. EDS pattern and elemental content analysis of soft-melting undripped material of 7.5% titanium content test of hearth slag. (Online version in color.)

Fig. 7. EDS pattern and elemental content analysis of soft-melting undripped material of 12% titanium content test of hearth slag. (Online version in color.)

Fig. 8. EDS pattern and elemental content analysis of soft-melting undripped material of hearth slag titanium content 16.5% test. (Online version in color.)

Fig. 9. EDS pattern and elemental content analysis of soft-melting undripped material of hearth slag titanium content 21% test. (Online version in color.)

Fig. 10. EDS pattern and elemental content analysis of soft-melting undripped material of 25.5% titanium content test of hearth slag. (Online version in color.)

From Figs. 5 and 6, 7, 8, 9, 10, it is evident that titanium (Ti) and vanadium (V) are primarily contained within perovskite, while most calcium (Ca), silicon (Si), magnesium (Mg), and aluminum (Al) are found in silicate phases, such as yellow feldspar and pyroxene. Notably, pyroxene exhibits a high mass fraction of magnesium and a low mass fraction of aluminum compared to yellow feldspar. When the theoretical composition of the blast furnace hearth slag has a TiO₂ content of 7.5%, the main physical phases consist of perovskite (CaTiO₃), pyroxene (CaMgSiO₃), and a small amount of titanium carbide (TiC) and titanium nitride (TiN).

As the TiO₂ content increases, the physical phases continue to include perovskite (CaTiO₃), TiC, TiN, and pyroxene (CaMgSiO₃), with an increase in diffraction peaks corresponding to perovskite, indicating a rise in the purity of its components. Additionally, a small amount of Al_0.5Ca₂Mg_0.75O₇Si_1.75 was detected in the samples. When the TiO₂ content in the theoretical composition of the blast furnace hearth slag ranges from 16.5% to 25%, a gradual decrease in the diffraction peak of pyroxene is observed, alongside an increase in the intensity of the feldspar peak. This transition indicates that pyroxene (CaMgSiO₃) is being converted to feldspar (CaMgSi₂O₆), which has a higher melting point. Concurrently, the concentrations of high-melting-point compounds, such as calcium titanium oxide (CaTiO₃), TiC, and TiN, continue to rise, while the content of silicates decreases. Notably, spinel (Al₂MgO₄) begins to form as the TiO₂ content increases.

At a TiO₂ content of 25.5%, pyroxene (CaMgSiO₃) is no longer observed in the silicate phase, having been replaced by feldspar (CaMgSi₂O₆). At this stage, the primary physical phases are perovskite (CaTiO₃), TiC, TiN, feldspar (CaMgSi₂O₆), and spinel (Al₂MgO₄). Overall, as the TiO₂ content increases, the abundance of high-melting-point compounds such as perovskite, TiC, and TiN in the slag also increases, while low-melting-point silicate compounds transition towards higher melting-point alternatives. The variation in composition is summarized in Table 6.

Table 6. Phase composition of undripped matter in different experiments, %.

TiO2	Silicate, titanium-containing compound product composition
25.5	CaTiO3(↑), TiC(↑), TiN(↑), CaMgSi2O6(↑), Al2MgO4(↑), Fe2.5Ti0.5O4(↑), FexO(↑)
21	CaTiO3(↑), TiC(↑), TiN(↑), CaMgSiO3(↓), CaMgSi2O6(↑), Al2MgO4, Fe2.5Ti0.5O4(↑), FexO(↑)
16.5	CaTiO3(↑), TiC(↑), TiN(↑), CaMgSiO3(↓), Fe2.5Ti0.5O4(↑), FexO(↑)
12	CaTiO3(↑), TiC(↑), TiN(↑), CaMgSiO3(↓), Al0.5Ca2Mg0.75O7Si1.75, Fe2.5Ti0.5O4(↑), FexO(↑)
7.5	CaTiO3, TiC, TiN, CaMgSiO3, Fe2.5Ti0.5O4, FexO

Note: Arrows in parentheses indicate the change in content compared to the lower experimental content. XRD is only available for components greater than 3% in the physical phase.

3.5. Rist Operation Line

Based on the composition of the incoming iron ore and the resulting slag iron composition obtained from the experiment, the simulated blast furnace operating line was constructed using the calculation method relevant to actual blast furnace operations.^30,31,32,33) The iron composition and slag sulfur content for the different experimental conditions are presented in Table 7.

Table 7. Iron composition and slag sulfur content for different experiments, wt%.

Program	Iron composition					Slag content (S)
	[Si]	[Mn]	[P]	[Ti]	[V]
1	0.897	0.12	0.269	0.479	0.09	0.134
2	1.952	0.189	0.082	0.562	0.104	0.106
3	1.239	0.126	0.075	0.416	0.099	0.142
4	1.536	0.121	0.128	0.605	0.122	0.125
5	1.22	0.13	0.105	0.482	0.122	0.092

The simulation of the experiment requires two points for determining the Rist operation line, which can be plotted using any combination of two parameters among points A, B, D, and E. In this study, the line connecting points A and B is selected as the simulated Rist operation line. The line between points U and V intersects line AB at point P, while line PW serves as the ideal operating line of Rist. As illustrated in Fig. 11(a), the detailed steps for this process are as follows:

Fig. 11. Simulated Rist operating line for different TiO₂ contents in the theoretical composition of blast furnace hearth sla. (Online version in color.)

First, point A is utilized to represent the degree of oxidation of the incoming ore and the oxidation of carbon in the blast furnace top gas. In this paper, the horizontal coordinates of point A are obtained by calculating the composition of the roof gas from the blast furnace produced on-site at a domestic iron and steel plant, while the vertical coordinates of point A are obtained by calculating the degree of oxidization of the iron ore fed into the furnace in the experiment. Equations (25) and (26) provide the calculations needed for this purpose.

  

$$\frac{\text{O}}{\text{C}}=\frac{{\text{2CO}}_{\text{2}}+\text{CO}}{{\text{CO}}_{\text{2}}+\text{CO}}=\text{1}+\frac{{\text{CO}}_{\text{2}}}{{\text{CO}}_{\text{2}}\text{+CO}}={\text{x}}_{\text{A}}$$(25)

  

$$\frac{\text{O}}{\text{Fe}}=\frac{\left({\text{Fe}}_{\text{2}}{\text{O}}_{\text{3}}\cdot \frac{\text{48}}{\text{160}}+\text{FeO}\cdot \frac{\text{16}}{\text{72}}\right)\text{/16}}{\text{TFe/56}}={\text{y}}_{\text{A}}$$(26)

The horizontal coordinate of point B is set at 1, while the vertical coordinate is associated with the direct reduction of iron in the ore. This value can be calculated based on the direct reduction data obtained from experimental processing. The vertical coordinates of point B are respectively B₁=0.72; B₂=0.66; B₃=0.622; B₄=0.611; B₅=0.585.

The composition of each component was determined through X-ray fluorescence (XRF) analysis of the post-experimental slag iron. Additionally, the oxygen consumed for the reduction and desulfurization of trace elements was calculated using Eqs. (27), (28), (29), (30), (31), (32), (33).

  

$$\frac{\text{10}\cdot \text{2}\cdot \mathrm{\omega }\text{[si]\%/28}}{\mathrm{\omega }{\text{(Fe)}}_{\text{Reduction}}\text{/56}}=\frac{\text{40}\mathrm{\omega }\text{[si]\%}}{\mathrm{\omega }{\text{(Fe)}}_{\text{Reduction}}}={\text{y}}_{\text{Si}}$$(27)

  

$$\frac{\text{10}\cdot \mathrm{\omega }\text{[Mn]\%/55}}{\mathrm{\omega }{\text{(Fe)}}_{\text{Reduction}}\text{/56}}=\frac{\text{10}\text{.182}\mathrm{\omega }\text{[Mn]\%}}{\mathrm{\omega }{\text{(Fe)}}_{\text{Reduction}}}={\text{y}}_{\text{Mn}}$$(28)

  

$$\frac{\text{10}\cdot \text{2}\text{.5}\mathrm{\omega }\text{[P]\%/31}}{\mathrm{\omega }{\text{(Fe)}}_{\text{Reduction}}\text{/56}}=\frac{\text{45}\text{.161}\mathrm{\omega }\text{[P]\%}}{\mathrm{\omega }{\text{(Fe)}}_{\text{Reduction}}}={\text{y}}_{\text{p}}$$(29)

  

$$\frac{\text{10}\cdot \text{2}\cdot \mathrm{\omega }\text{[Ti]\%/48}}{\mathrm{\omega }{\text{(Fe)}}_{\text{Reduction}}\text{/56}}=\frac{\text{23}\text{.333}\mathrm{\omega }\text{[Ti]\%}}{\mathrm{\omega }{\text{(Fe)}}_{\text{Reduction}}}={\text{y}}_{\text{Ti}}$$(30)

  

$$\frac{\text{10}\cdot \text{2}\text{.5}\cdot \mathrm{\omega }\text{[V]\%/51}}{\mathrm{\omega }{\text{(Fe)}}_{\text{Reduction}}\text{/56}}=\frac{\text{27}\text{.451}\mathrm{\omega }\text{[V]\%}}{\mathrm{\omega }{\text{(Fe)}}_{\text{Reduction}}}={\text{y}}_{\text{V}}$$(31)

  

$$\frac{\mathrm{\mu }\cdot \mathrm{\omega }\text{[Mn]\%/32}}{\mathrm{\omega }{\text{(Fe)}}_{\text{Reduction}}\text{/56}}=\frac{\text{1}\text{.75}\mathrm{\mu }\mathrm{\omega }{\text{(S)}}_{\text{\%}}}{\mathrm{\omega }{\text{(Fe)}}_{\text{Reduction}}}={\text{y}}_{\text{S}}$$(32)

  

$${\mathrm{y}}_{\mathrm{f}}={\mathrm{y}}_{\mathrm{s}\mathrm{i}}+{\mathrm{y}}_{\mathrm{M}\mathrm{n}}+{\mathrm{y}}_{\mathrm{P}}+{\mathrm{y}}_{\mathrm{T}\mathrm{i}}+{\mathrm{y}}_{\mathrm{V}\mathrm{O}}+{\mathrm{y}}_{\mathrm{S}}$$(33)

In the formula, ω[si]%, ω[Mn]%, ω[P]%, ω[Ti]%, ω[V]% is the percentage of each component in the pig iron; ω(S)_% is the percentage of S in the slag; μ represent the amount of slag.

The total heat consumption (Q) for the reduction of oxides other than iron oxides was calculated using Eq. (34).

  

$$\begin{array}{c}{\text{q}}_{\text{Si}}^{\prime }=\text{22}\mathrm{\hspace{0.17em} }\text{667(1}\mathrm{\hspace{0.17em} }\text{000}+{\text{W}}_{\text{Recycled}\mathrm{\hspace{0.17em} }\text{iron}}\text{)}{\mathrm{\omega }}_{\text{[Si]\%}}\\ {\text{q}}_{\text{Mn}}^{\prime }=\text{5}\mathrm{\hspace{0.17em} }\text{222(1}\mathrm{\hspace{0.17em} }\text{000}+{\text{W}}_{\text{Recycled}\mathrm{\hspace{0.17em} }\text{iron}}\text{)}{\mathrm{\omega }}_{\text{[Mn]\%}}\\ {\text{q}}_{\text{P}}^{\prime }=\text{26}\mathrm{\hspace{0.17em} }\text{258(1}\mathrm{\hspace{0.17em} }\text{000}+{\text{W}}_{\text{Recycled}\mathrm{\hspace{0.17em} }\text{iron}}\text{)}{\mathrm{\omega }}_{\text{[P]\%}}\\ {\text{q}}_{\text{Ti}}^{\prime }=\text{14}\mathrm{\hspace{0.17em} }\text{784(1}\mathrm{\hspace{0.17em} }\text{000}+{\text{W}}_{\text{Recycled}\mathrm{\hspace{0.17em} }\text{iron}}\text{)}{\mathrm{\omega }}_{\text{[Ti]\%}}\\ \text{Q}={\text{q}}_{\text{Si}}^{\prime }+{\text{q}}_{\text{Mn}}^{\prime }+{\text{q}}_{\text{P}}^{\prime }+{\text{q}}_{\text{Ti}}^{\prime }\end{array}$$(34)

In the formula: W_{Recycled iron} is about 10 kg/t.

Connect points A and B and intersect the y-axis at point E. Use the coordinates of point E (0, −(y_f+y_b)) to find the value of y_b. Find q_b by substituting the obtained Q into y_b·q_b=y_d·q_d+Q.

In the formula, q_d represent the heat dissipation for direct reduction of FeO, which is fixed to be about 152197 kJ/kg; y_d represent the number of atoms burning C in front of the BF tuyere when smelting 1 kg of atomic Fe, y_b=y_d=r_d; q_b is the effective heat released from the combustion of carbon in front of the BF tutere, kJ/kg.

Utilization of the formula X_p=$\frac{q_d}{q_d+q_b}$, Y_P= Y_U+ X_P (Y_V−Y_U) to find the coordinates of point P. It is now known that the coordinates of point U are (0, −y_f) and the coordinates of point V are (1, Q/q_d), and the connection UV intersects point P with AE. AE is the simulated Rist operating line for this mineral allocation condition. It is known that the coordinates of point W are ((1.32–1.29), 1.05), and PW is the Rist ideal operating line for this allocation condition.

The vertical coordinate of point B represents the direct reduction during the smelting process. As the TiO₂ content in the theoretical composition of the blast furnace hearth slag increases, point B shifts downward (Fig. 11(b)), indicating a continuous decrease in the direct reduction of the burden. Simultaneously, the downward shift of point E reflects a change in the amount of reactive air, which decreases as the TiO₂ content in the theoretical slag composition increases (Fig. 11(c)). Consequently, the amount of carbon monoxide (CO) generated by the blowing process increases.

Point P, which is the intersection of the AE line and the UV line, is influenced by the heat balance of the blast furnace. As the TiO₂ content in the theoretical composition of the blast furnace hearth slag rises, point P gradually shifts downward (Fig. 11(d)). In this region, most of the substances generated during the reduction process are further reduced, leading to an increased heat requirement in the high-temperature area. Therefore, the downward shift of point P indicates a higher heat demand for the reduction of substances in this region.

Furthermore, as the TiO₂ content in the theoretical composition of the blast furnace hearth slag increases, the Rist operating line exhibits significant changes. The AE line shifts downward (Fig. 11(a)), with an increasing slope, resulting in a higher coke ratio required for blast furnace smelting.

4. Analysis and Discussion

Vanadium and titanium magnetite smelting in a blast furnace involves the processing of a vanadium- and titanium-containing iron burden, which includes a mixture of vanadium and titanium sinter, pellets, lump ore, and coke. This mixed burden is loaded from the top of the blast furnace. As the burden descends through the furnace throat and into the belly, it encounters an ever-increasing temperature gradient and the action of rising gas flow. This process facilitates reduction, softening, and melting, ultimately leading to the formation of slag and molten iron, which then drips down.^1,2) In this paper, we comparatively analyze and discuss the experimental results concerning the softening, melting, dripping, and permeability of the materials involved.

The mineral composition of vanadium and titanium sinter ores typically includes titanium hematite, titanium magnetite, perovskite, and silicate-bearing minerals, along with small quantities of calcium ferrite, brookite, and residual ilmenite. In contrast, pelletss predominantly consist of hematite (Fe₂O₃). Within the lump zone of the furnace, indirect reduction mainly occurs, with the reduction efficiency decreasing from 48.80% to 46.10% as the TiO₂ content increases (Fig. 4). This may be due to the fact that as the proportion of sintered ore in the incoming raw material increases, the content of difficult-to-reduce titanium magnetite increases and the easily reducible titanium hematite and calcium ferrate decreases, along with an increase in the high melting-point gangue-like mineral phases, which results in a slightly decreasing trend in the indirect reducibility of the charge. In this process, the iron oxides within the mixture are reduced stepwise by carbon monoxide (CO) to form wüstite.³⁴⁾ During this reaction, titanium hematite, ferrite phases, and brookite in the sinter are converted to titanomagnetite through the removal of oxygen. Magnetite begins to reduce and decompose, resulting in the precipitation of fine wüstite, while perovskite remains unchanged.³⁵⁾ Meanwhile, in the upper blast furnace block belt, a phase transition occurs during the reduction of titanium hematite to titanium magnetite, during which the volume of the sintered ore expands, the internal stress increases, and cracking and chalking intensify. The primary chemical reactions involved in these processes are as follows:

  

$$\begin{array}{l}m\mathrm{F}{\mathrm{e}}_2{\mathrm{O}}_3-n\mathrm{F}\mathrm{e}\mathrm{O}\cdot \mathrm{T}\mathrm{i}{\mathrm{O}}_2+\mathrm{C}\mathrm{O}\to \\ m\mathrm{F}\mathrm{e}\mathrm{O}\cdot \mathrm{F}{\mathrm{e}}_2{\mathrm{O}}_3-n(2\mathrm{F}\mathrm{e}\mathrm{O}\cdot \mathrm{T}\mathrm{i}{\mathrm{O}}_2)+\mathrm{F}\mathrm{e}\mathrm{O}+\mathrm{C}{\mathrm{O}}_2\end{array}$$(35)

  

$$\begin{array}{l}\mathrm{C}\mathrm{a}\mathrm{O}\cdot \mathrm{F}{\mathrm{e}}_2{\mathrm{O}}_3(2\mathrm{C}\mathrm{a}\mathrm{O}\cdot \mathrm{F}{\mathrm{e}}_2{\mathrm{O}}_3)+\mathrm{C}\mathrm{O}\to \\ \mathrm{m}\mathrm{F}\mathrm{e}\mathrm{O}\cdot \mathrm{T}\mathrm{i}{\mathrm{O}}_2-\mathrm{n}(2\mathrm{F}\mathrm{e}\mathrm{O}\cdot \mathrm{T}\mathrm{i}{\mathrm{O}}_2)+\mathrm{C}\mathrm{a}\mathrm{O}+\mathrm{C}{\mathrm{O}}_2\end{array}$$(36)

  

$$\begin{array}{l}m\mathrm{F}\mathrm{e}\mathrm{O}\cdot \mathrm{F}{\mathrm{e}}_2{\mathrm{O}}_3-n(2\mathrm{F}\mathrm{e}\mathrm{O}\cdot \mathrm{T}\mathrm{i}{\mathrm{O}}_2)+3m\mathrm{C}\mathrm{O}\to \\ 3m\mathrm{C}{\mathrm{O}}_2+3m\mathrm{F}\mathrm{e}\mathrm{O}+n(2\mathrm{F}\mathrm{e}\mathrm{O}\cdot \mathrm{T}\mathrm{i}{\mathrm{O}}_2)\end{array}$$(37)

  

$$\begin{array}{l}m(\mathrm{F}\mathrm{e}\mathrm{O}\cdot \mathrm{T}\mathrm{i}{\mathrm{O}}_2)-\mathrm{n}(2\mathrm{F}\mathrm{e}\mathrm{O}\cdot \mathrm{T}\mathrm{i}{\mathrm{O}}_2)+\mathrm{C}\mathrm{O}\to \\ \mathrm{F}\mathrm{e}\mathrm{O}+2\mathrm{F}\mathrm{e}\mathrm{O}\cdot \mathrm{T}\mathrm{i}{\mathrm{O}}_2+\mathrm{C}{\mathrm{O}}_2\end{array}$$(38)

  

$$\mathrm{F}\mathrm{e}\mathrm{O}+\mathrm{C}\mathrm{O}\to \mathrm{F}\mathrm{e}+\mathrm{C}{\mathrm{O}}_2$$(39)

As the temperature continues to rise and reaches the threshold for the coke gasification reaction, the carbon dioxide (CO₂) produced by high-temperature indirect reduction reacts with coke, generating carbon monoxide (CO) (Eq. (40)). This process initiates the coke melt-loss reaction. A significant quantity of wüstite and veinstone components combine to form a low melting point liquid slag, promoting the contraction and deformation of the burden. Softening begins at temperatures between 1100°C and 1122°C, during which titanomagnetite completely disappears, and ulvospinel and ilmenite are reduced to wüstite and metallic iron, resulting in an increase in the perovskite content within the material.³⁶⁾

In the softening zone, the softening end temperature (T_m,i) gradually increases with the rising TiO₂ content, while the softening zone (ΔT_s) also expands (Fig. 2). The degree of indirect reduction varies from 1.30% to 3.09%, and the degree of direct reduction ranges from 1.99% to 6.48% (Fig. 4). The growing presence of unreduced FeO, which recombines with ilmenite in the vanadium-titanium ore to form ulvospinel (Eq. (44)). This interaction results in a coupled reduction pathway for iron and titanium during the reduction process. Consequently, the FeO bound to titanide is more challenging to reduce compared to free FeO, complicating the reduction of high-temperature vanadium-titanium ore and the separation of slag and iron.³⁷⁾

  

$$\mathrm{C}+\mathrm{C}{\mathrm{O}}_2=2\mathrm{C}\mathrm{O}$$(40)

  

$$2\mathrm{F}\mathrm{e}\mathrm{O}\cdot \mathrm{T}\mathrm{i}{\mathrm{O}}_2+\mathrm{C}\mathrm{O}=\mathrm{F}\mathrm{e}+\mathrm{F}\mathrm{e}\mathrm{O}\cdot \mathrm{T}\mathrm{i}{\mathrm{O}}_2+\mathrm{C}{\mathrm{O}}_2$$(41)

  

$$\mathrm{F}\mathrm{e}\mathrm{O}\cdot \mathrm{T}\mathrm{i}{\mathrm{O}}_2+\mathrm{C}\mathrm{O}=\mathrm{F}\mathrm{e}+\mathrm{T}\mathrm{i}{\mathrm{O}}_2+\mathrm{C}{\mathrm{O}}_2$$(42)

  

$$\mathrm{F}\mathrm{e}\mathrm{O}+\mathrm{C}\mathrm{O}=\mathrm{F}\mathrm{e}+\mathrm{C}{\mathrm{O}}_2$$(43)

  

$$\mathrm{F}\mathrm{e}\mathrm{O}+\mathrm{F}\mathrm{e}\mathrm{O}\cdot \mathrm{T}\mathrm{i}{\mathrm{O}}_2=2\mathrm{F}\mathrm{e}\mathrm{O}\cdot \mathrm{T}\mathrm{i}{\mathrm{O}}_2$$(44)

In the melting zone of the blast furnace, the initial slag, which contains a substantial amount of FeO, begins to melt at temperatures exceeding 1184°C to 1247°C (Fig. 2). However, at this stage, the slag viscosity remains high, causing the incompletely melted initial slag to obstruct the gas flow channels. This blockage results in a rapid increase in gas differential pressure. Concurrently, the significant amount of FeO from the vanadium and titanium ores combines tightly with ilmenite, ulvospinel, and other titanium-bearing minerals,³⁸⁾ making it challenging to separate the molten slag from the iron and hindering the reduction of FeO. The volume of the initial slag in the blast furnace further exacerbates this issue, causing the differential pressure during the soft-melting process of the vanadium and titanium ore to continue rising, thereby deteriorating permeability (Fig. 3).

As the TiO₂ content in the theoretical composition of the blast furnace hearth slag increases, the melting start temperature (T_m,i) gradually rises, while the melting end temperature (T_m,f) exhibits only a slight increase. Consequently, the melting zone (ΔT_m) narrows (Fig. 2). Additionally, the degrees of indirect and direct reduction in the melting zone show a decreasing trend, falling from 2.45% and 24.4% to 1.11% and 22.3%, respectively (Fig. 4).

The primary slag flows downward along the coke voids, maintaining good contact with the surface of the coke. At elevated temperatures, this facilitates a faster direct reduction. Moreover, the reduction of ilmenite in contact with coke generates titanium carbonitride Ti(C,N), which further worsens the burden’s permeability. Only when the temperature rises sufficiently to form iron droplets can some of these issues be alleviated. However, a significant amount of titanium-containing primary slag can lead to “flooding” and “foam slag” phenomena in the high-temperature gas flow.³⁹⁾

  

$$(\mathrm{F}\mathrm{e}\mathrm{O})+\mathrm{C}=\mathrm{F}\mathrm{e}+\mathrm{C}\mathrm{O}$$(45)

  

$$\mathrm{T}\mathrm{i}{\mathrm{O}}_2+3\mathrm{C}=\mathrm{T}\mathrm{i}\mathrm{C}+2\mathrm{C}{\mathrm{O}}_2$$(46)

  

$$\mathrm{T}\mathrm{i}{\mathrm{O}}_2+2\mathrm{C}+1/2{\mathrm{N}}_2=\mathrm{T}\mathrm{i}\mathrm{N}+2\mathrm{C}\mathrm{O}$$(47)

Analyzing the results presented in Figs. 5 and 6, 7, 8, 9, 10 indicates that the concentration of high melting point compounds, such as perovskite, in the slag increases with rising TiO₂ content in the burden. The coupled reduction pathway for iron and titanium in the reduction process of vanadium and titanium ores is illustrated in Fig. 12, wherein iron and titanium oxides in the cohesive zone are reduced according to the sequence: Fe₂O₃·TiO₂→Fe₂TiO₄→FeTiO₃→FeTi₂O₅.

Fig. 12. Reduction mechanism of vanadium-titanium magnetite. (Online version in color.)

According to the pathway of intermediate material generation, perovskite is decomposed from ulvospinel in the melting zone, while ulvospinel is formed from the reaction of magnetite with TiO₂ in the softening zone and gradually accumulates. As the TiO₂ content increases, the mass fraction of ulvospinel generated in the softening zone also rises. However, these ulvospinel particles become surrounded by iron-containing compounds, which restrict their further reduction. This limitation results in a lower FeO content during the presoftening phase, complicating the formation of a liquid phase and leading to an increase in the softening end temperature.^40,41) Considering that the softening characteristics of the ore during the temperature-raising reduction process also depend on the amount of high and low melting point minerals produced by the ore in the process, the softening end temperature of the ore is mainly influenced by the high melting point minerals. As the ratio of sintered ores in the raw materials entering the furnace is increasing, especially the proportion of high titanium sintered ores increases from 3% in Scheme 3 to 12.5% in Scheme 5, which makes the content of high melting point compounds such as perovskite, calcium silicate, and TiC and TiN produced due to the reduction of TiO₂ increase in the primary slag of the blast furnace, which makes the primary slag of the blast furnace viscous and difficult to flow, and which is the main reason for the end of the furnace charge softening in the course of the experiments (Tm, i) and the end of melting temperature (Tm, f) increased and deteriorated gas permeability during the experiment.^42,43,44) This observation is consistent with findings by Jiao et al.

In blast furnace operations, within a certain range, the indirect reduction effect is enhanced due to exothermic reactions. When the softening zone narrows and the lumpy belt widens, the development of indirect reduction is facilitated, reducing the coke ratio while also decreasing the gas flow resistance in the cohesive zone, thus enhancing the efficiency of the blast furnace smelting. In this experiment, as the TiO₂ content in the burden rises, the extent of indirect reduction increases; however, the softening end temperature and melting end temperature also rise concomitantly with the titanium content. This is accompanied by a downward shift of the B-point (Fig. 11(b)), while the degree of direct reduction of the burden continues to decline, leading to increased reactions between titanium compounds (Table 6).

As the TiO₂ content in the burden increases, the E-point of the Rist operating line shifts downward (Fig. 11(c)). Consequently, the amount of CO generated by the blast furnace increases, driven partly by oxygen obtained from non-iron oxide reductions and desulfurization processes, which further intensifies non-iron oxide reduction. This results in an increased sulfur content in the iron water, making it increasingly challenging to remove sulfur from the slag. Additionally, another portion of the oxygen is introduced via the blast, where it combines with carbon to generate CO as a reducing gas during the ascent in the furnace. As the TiO₂ content rises, the blast furnace requires more oxygen to provide reducing gas for the reduction of the burden.

The P-point of the Rist operating line gradually shifts downward as the TiO₂ content in the theoretical slag composition of the blast furnace hearth increases (Fig. 11(d)). Most of the substances produced in the reduction process are reduced in this region, which contributes to an increased demand for heat in the high-temperature area. The downward shift of the P-point can, therefore, be interpreted as indicating a higher heat requirement for the reduction of substances in this region.

These changes cause the simulated Rist operating line to gradually diverge from the ideal operating line, necessitating increased airflow and coke ratio in actual blast furnace operations.

5. Conclusion

(1) As the TiO₂ content in the theoretical composition of the blast furnace hearth slag increased from 7.5 wt% to 25.5 wt%, the softening end temperature (T_m,i), melting end temperature (T_m,f), permeability (K), and eigenvalue (S) of the burden exhibited an upward trend. Indirect reduction (R_i(IR)) in the massive zone, along with indirect reduction (R_i(IR+DR)) and direct reduction (R_d(IR+DR)) in the soft zone, showed a gradual increase. Conversely, direct reduction (R_d(MDR)) and indirect reduction (R_i(MDR)) in the melting zone exhibited a gradual decline. The softening zone (ΔT_s) widened, while the melting zone (ΔT_m) narrowed. In addition, the temperature interval of the cohesive zone (ΔT_c) slightly increased, and the cohesive zone tended to shift downward.

(2) In the softening zone, the blast furnace burden primarily experienced the reduction of iron oxides, while the reduction of titanium oxides commenced in the melting zone. As the TiO₂ content in the theoretical composition of the blast furnace hearth slag increases, the abundance of high melting point substances such as perovskite, TiC, and TiN in the soft-melting undripping slag also increases. Concurrently, the pyroxene phase in the silicate transforms into the feldspar phase, which exhibits a higher melting point. When the TiO₂ content exceeds 16.5 wt%, spinel is generated in the melting zone. In the cohesive zone, iron and titanium oxides are reduced according to the pathway: Fe₂O₃·TiO₂→Fe₂TiO₄→FeTiO₃→FeTi₂O₅.

(3) As the TiO₂ content in the theoretical composition of the blast furnace hearth slag increases, the B-point of the Rist operating line shifts downward, indicating a decrease in direct reduction. The E-point also shifts downward, resulting in an increased amount of CO generated by the blast. Additionally, the P-point gradually moves downward, signifying a rising heat requirement for the reduction of substances in the high-temperature region. Consequently, the simulated Rist operating line gradually deviates from the ideal operating line, necessitating an increase in both the blast rate and coke ratio during blast furnace operation.

Conflict of Interest Statement

On behalf of all authors, the corresponding author states that there is no conflict of interest.

Acknowledgement

The support of the National Natural Science Foundation of China (No. 52174319) is greatly acknowledged.

References

• 1)   H. G.  Du: Principle of Smelting Vanadium Titanium Magnetite in Blast Furnace, Science Press, Beijing, (1996), 40 (in Chinese). ISBN7-03-005118-1
• 2)   X. X.  Xue,  S. T.  Yang and  Y.  Zhang: Blast Furnace Process Smelting of Chromium-bearing Vanadium-titanium Magnetite: Theory and Practice, Science Press, Beijing, (2020), 249 (in Chinese). ISBN978-7-03-064485-5
• 3)   J.  Tang,  M.  Chu,  C.  Feng,  F.  Li,  Y.  Tang and  Z.  Liu: ISIJ Int., 57 (2017), 1156. https://doi.org/10.2355/isijinternational.ISIJINT-2016-714
• 4)   C. G.  Bai: Ph.D. Thesis, Chongqing University, (2003), 1 (in Chinese). https://doi.org/10.7666/d.y794946
• 5)   K.  Hu,  X.  Lv,  S.  Li,  W.  Lv,  B.  Song and  K.  Han: Metall. Trans. B, 49 (2018), 1963. https://doi.org/10.1007/s11663-018-1284-x
• 6)   K.  Zhou,  J.  Song,  Z.  You,  H.  Xie and  X.  Lv: ISIJ Int., 60 (2020), 1409. https://doi.org/10.2355/isijinternational.ISIJINT-2019-559
• 7)   N. J.  Bristow and  C. E.  Loo: ISIJ Int., 32 (1992), 819. https://doi.org/10.2355/isijinternational.32.819
• 8)   A.  Dehghan-Manshadi,  J.  Manuel,  S.  Hapugoda and  N.  Ware: ISIJ Int., 54 (2014), 2189. https://doi.org/10.2355/isijinternational.54.2189
• 9)   X. W.  Lv,  Z.  Lun,  J.  Yin and  C.  Bai: ISIJ Int., 53 (2013), 1115. https://doi.org/10.2355/isijinternational.53.1115
• 10)   Z. X.  Gao,  G. J.  Cheng,  X. X.  Xue,  H.  Yang and  P. N.  Duan: Steel Res. Int., 89 (2018), 1700543. https://doi.org/10.1002/srin.201700543
• 11)   K. X.  Jiao,  J. L.  Zhang,  Z. Y.  Wang,  Y. X.  Liu and  R. Z.  Xu: High Temp. Mater. Process., 37 (2018), 149. https://doi.org/10.1515/htmp-2016-0148
• 12)   T.  Nishimura,  K.  Higuchi,  M.  Naito and  K.  Kunitomo: ISIJ Int., 51 (2011), 1316. https://doi.org/10.2355/isijinternational.51.1316
• 13)   Z. M.  Yan,  X. W.  Lv,  W. C.  He and  J.  Xu: ISIJ Int., 57 (2017), 31. https://doi.org/10.2355/isijinternational.ISIJINT-2016-420
• 14)   X. L.  Liu,  S. L.  Wu,  W.  Huang,  K. F.  Zhang and  K. P.  Du: ISIJ Int., 54 (2014), 2089. https://doi.org/10.2355/isijinternational.54.2089
• 15)   Z.  Ning,  X. Y.  Wang and  S. T.  Yang: Minerals, 14 (2024), 293. https://doi.org/10.3390/min14030293
• 16)   H. T.  Wang,  W.  Zhao,  M. S.  Chu,  R.  Wang,  Z. G.  Liu and  X. X.  Xue: J. Cent. South Univ., 24 (2017), 39. https://doi.org/10.1007/s11771-017-3406-z
• 17)   W. D.  Tang,  S. T.  Yang and  X. X.  Xue: ISIJ Int., 59 (2019), 988. https://doi.org/10.2355/isijinternational.ISIJINT-2018-681
• 18)   Y.  Ujisawa,  K.  Nakano,  Y.  Matsukura,  K.  Sunahara,  S.  Komatsu and  T.  Yamamoto: ISIJ Int., 45 (2005), 1379. https://doi.org/10.2355/isijinternational.45.1379
• 19)   Y. Z.  Wang,  J. L.  Zhang,  Z. J.  Liu,  Y. P.  Zhang,  D. H.  Liu and  C. B.  Du: Iron Steel, 52 (2017), 20 (in Chinese). https://doi.org/10.13228/j.boyuan.issn0449-749x.20160520
• 20)   X.  Zhan: Iron Steel Vanadium Titan., 5 (1984), 3. CNKI:SUN:GTFT.0.1984-02-001
• 21)   J.  Li,  Z. T.  Zhang,  L.  Liu,  W. L.  Wang and  X. D.  Wang: ISIJ Int., 53 (2013), 1696. https://doi.org/10.2355/isijinternational.53.1696
• 22)   Y. X.  Liu,  J. L.  Zhang,  G. H.  Zhang,  K. X.  Jiao and  K. C.  Chou: Ironmak. Steelmak., 44 (2016), 609. https://doi.org/10.1080/03019233.2016.1223907
• 23)   A.  Shankar,  M.  Görnerup,  A. K.  Lahiri and  S.  Seetharaman: Metall. Trans. B, 38 (2007), 911. https://doi.org/10.1007/s11663-007-9087-5
• 24)   Z.  Yan,  X.  Lv,  J.  Zhang,  Y.  Qin and  C.  Bai: Can. Metall. Q., 55 (2016), 186. https://doi.org/10.1080/00084433.2015.1126903
• 25)   Y. Q.  Jiang,  M. X.  Wu,  T. L.  Li,  Q.  Wang,  S.  Song,  H. L.  Wu,  J. C.  Huang,  S. T.  Yang,  C. Y.  Sun and  S. Z.  Wang: Fuel, 371 (2024), 132055. https://doi.org/10.1016/j.fuel.2024.132055
• 26)   Q.  Wang,  Y. S.  Song,  T. L.  Li,  Z. X.  Li and  E.  Tim: Iron Steel, 59 (2024), 44 (in Chinese). https://doi.org/10.13228/j.boyuan.issn0449-749x.20230324
• 27)   M. X.  Wu,  Z. X.  Li,  J. C.  Huang,  Q.  Wang,  T. L.  Li,  S. T.  Yang,  H. M.  He and  Y. Q.  Jiang: Fuel, 357 (2024), 129991. https://doi.org/10.1016/j.fuel.2023.129991
• 28)   G. L.  Tan,  Q.  Wang,  H. X.  Zheng,  W.  Zhao,  S.  Zhang and  Z. S.  Liu: J. Phys. Chem. A, 115 (2011), 5517. https://doi.org/10.1021/jp203580r
• 29)   Q.  Wang,  R.  Guo,  X.  Zhao,  J.  Sun,  S.  Zhang and  W.  Liu: Fuel, 182 (2016), 879. https://doi.org/10.1016/j.fuel.2016.05.101
• 30)   D. M.  Kundrat: Metall. Trans. B, 17 (1986), 705. https://doi.org/10.1007/bf02657133
• 31)   A. V.  Pavlov,  O. P.  Onorin,  N. A.  Spirin,  A. A.  Polinov,  V. V.  Lavrov and  I. A.  Gurin: Mater. Sci. Eng., 411 (2018), 012057. https://doi.org/10.1088/1757-899x/411/1/012057
• 32)   A.  Rist and  N.  Meysson: Jom, 19 (2017), 50. https://doi.org/10.1007/bf03378564
• 33)   A.  Spanlang,  W.  Wukovits and  B.  Weiss: Berg Huettenmaenn Monatsh, 165 (2020), 258. https://doi.org/10.1007/s00501-020-00979-y
• 34)   S. T.  Yang,  M.  Zhou,  X. X.  Xue,  T.  Jiang and  C. Y.  Sun: Jom, 71 (2019), 2812. https://doi.org/10.1007/s11837-019-03533-5
• 35)   S. T.  Yang,  M.  Zhou,  T.  Jiang and  X. X.  Xue: Int. J. Miner. Metall. Mater., 25 (2018), 145. https://doi.org/10.1007/s12613-018-1557-z
• 36)   Y. N.  Qie,  Q.  Lyu,  X. J.  Liu,  J. P.  Li,  C. C.  Lan,  S. H.  Zhang and  C. J.  Yan: Metall. Trans. B, 49 (2018), 2622. https://doi.org/10.1007/s11663-018-1299-3
• 37)   N. J.  Welham and  J. S.  Williams: Metall. Trans. B, 30 (1999), 1075. https://doi.org/10.1007/s11663-999-0113-7
• 38)   Y. X.  Liu,  J. L.  Zhang,  Z. Y.  Wang,  K. X.  Jiao,  G. H.  Zhang and  K. C.  Chou: Int. J. Miner. Metall. Mater., 24 (2017), 130. https://doi.org/10.1007/s12613-017-1387-4
• 39)   J. Y.  Xiang,  J.  Wang,  Q. J.  Li,  C.  Shan,  G. B.  Qiu,  W. Z.  Yu and  X. W.  Lv: Can. Metall. Q., 59 (2020), 151. https://doi.org/10.1080/00084433.2020.1715696
• 40)   G. J.  Cheng,  X. X.  Xue,  T.  Jiang and  P. N.  Duan: Metall. Trans. B, 47 (2016), 1713. https://doi.org/10.1007/s11663-016-0628-7
• 41)   J. H.  Liu,  Q.  Wang,  Z. G.  Guan,  Z. J.  He and  Z.  Zhuang: Metalurgija, 59 (2020), 7. https://hrcak.srce.hr/224739
• 42)   B.  Chen,  J.  Wen,  T.  Jiang,  L.  Li,  T.  Yu,  W.  Tang and  P.  Hu: ISIJ Int., 63 (2023), 1607. https://doi.org/10.2355/isijinternational.ISIJINT-2023-171
• 43)   K.  Jiao,  J.  Zhang,  Z.  Liu,  S.  Kuang and  Y.  Liu: ISIJ Int., 57 (2017), 48. https://doi.org/10.2355/isijinternational.ISIJINT-2016-419
• 44)   K.-i.  Ohno,  Y.  Kitamura,  S.  Sukenaga,  S.  Natsui,  T.  Maeda and  K.  Kunitomo: Tetsu-to-Hagané, 108 (2022), 703 (in Japanese). https://doi.org/10.2355/tetsutohagane.TETSU-2022-025