Effect of Introducing Coke into Ore Layer on Softening-melting-dropping Characteristics of Vanadium-titanium Mixed Burden under Simulating BF Conditions

Abstract

Coke-ore mixed charging is a well-known and effective measure to strengthen BF operation and realize low reducing agent BF ironmaking. The objective of this paper is to investigate the effect of coke-ore mixing ratio on softening-melting dropping performance and permeability of vanadium-titanium mixed burden and to clarify the interaction mechanism between mixed nut coke and iron-bearing burden under BF simulating conditions. It was found that the softening-melting-dropping behaviors and permeability of mixed burden get improved obviously with a coke-ore mixing ratio of 20%. However, the generation of carbide of V and Ti would be accelerated with further increasing coke-ore mixing ratio to 50%, which would deteriorate the fluidity of slag and worsen the dropping behavior of mixed burden and lower the yield of V in liquid iron. The interaction between nut coke and vanadium-titanium mixed burden could be summarized as four parts, namely reduction strengthen, carburization promotion, permeability improvement, and the precipitation of the carbide of titanium.

1. Introduction

As a typical type of iron-bearing mineral resources rich in vanadium, titanium, chromium and so on, vanadium-titanium magnetite (VTM) is widely distributed in China, Russia, South Africa, New Zealand, and America.^1,2) In China, there is about more than 18 billion tons of VTM deposit in Panxi district and Chengde district.^3,4) Therefore, the efficient utilization of VTM is significant for the development of steel industry.

Currently, the Blast Furnace (BF) is still the predominant process for the smelting of vanadium-titanium magnetite and will remain so for the next decades in China.^5,6) However, in conventional VTM BF ironmaking practice, there are some problems in BF smelting VTM process, including high coke ratio, low gas utilization, and low yield of vanadium.^7,8) These problems of BF smelting VTM are mainly caused by the poor reducibility of VTM ore and bad gas permeability of packed bed, especially in the cohesive zone of BF. Coke mixed charging technique has been widely applied to normal BF ironmaking, aiming to improve gas permeability of packed bed and realize low reducing agent ratio (RAR) and high productivity BF operation.^{9,10,11,12,13,14,15)} Gono et al.¹⁶⁾ studied the influence of mixing low quality ore with coke, finding that the reducibility of the raw materials were improved efficiently. Mousa et al.¹⁷⁾ explored the effect of nut coke on the shaft permeability and sinter reducibility under BF simulating conditions, and demonstrated that the mixed nut coke in the sinter bed could improve the sinter reducibility and inhibit the reduction retardation phenomenon. Watakabe et al.^18,19,20) proposed high ratio coke mixed charging technique and applied it to JFE Steel’s East Japan Works NO. 6 BF, reaching an obvious improvement of gas permeability of cohesive zone. Kuzin et al.²¹⁾ demonstrated that the gas permeability of sinter layer with a mixing of 10 vol% nut coke would decrease the pressure drop in the dry shaft by 5.33%. Also, the nut coke mixing technique results in a efficient decrease of coke ratio. The statistical analysis of some European blast furnaces indicated that the coefficient replacement factor is nearly to 1.0 with introducing 5%–30% nut coke to total coke consumption.²²⁾ All these studies show that introducing coke into ore layer benefits the normal BF operation. For BF smelting VTM, the strengthen of the reduction of VTM and the improvement of gas permeability of packed bed take an important part in BF smooth operation. Therefore, we applied coke mixed charging technique to BF smelting VTM, aiming to promote reduction performance of vanadium-titanium mixed burden and improve gas permeability of packed bed and increase the yield of V.

This paper mainly focuses on the effect of introducing coke into ore layer on the performance of BF smelting VTM under laboratory conditions. Firstly, the effect of coke mixing ratio on softening-melting-dropping characteristics, and gas permeability of vanadium-titanium mixed burden with simulating BF conditions was investigated systematically. Next, the migration of valuable element V and Ti was discussed and illuminated by thermodynamics calculation and scanning electron microscopy and energy dispersive spectrometer technique (SEM-EDS). Finally, the interaction mechanism between nut coke and vanadium-titanium mixed burden was explored by interrupted softening-melting experiments.

2. Experimental

2.1. Raw Materials

The ferrous materials used in this research, including V–Ti sinter, V–Ti pellet, and lump ore, are obtained from an Iron and Steel Cooperation in China. The chemical compositions of the three kind ferrous materials are listed in Table 1. The V-bearing phases and Ti-bearing phases existed in V–Ti sinter and V–Ti pellet are Fe₂VO₄, FeTi₂O₄, and FeTiO₃. The proximate analysis and chemical compositions of the coke are listed in Table 2. The size of V–Ti sinter, V–Ti pellet and lump ore are all 10–12.5 mm, and the mixing coke size is 8–10 mm. All samples were dried at 105°C in a dry furnace to ensure the complete removal of moisture before testing softening-melting-dropping performance.

Table 1. Chemical compositions of ferrous materials (wt%).

Items	TFe	FeO	CaO	SiO2	MgO	Al2O3	TiO2	V2O5
V–Ti Sinter	54.66	9.40	11.32	5.33	2.12	2.14	1.92	0.21
V–Ti Pellet	59.16	0.98	1.50	5.86	3.02	1.68	2.95	0.20
Lump ore	61.57	0.19	0.15	5.66	0.19	0.66	0.13	–

Table 2. Proximate analysis and chemical compositions of coke (wt%).

Fixed Carbon	Ash	Volatiles	CaO	SiO2	Al2O3	S
85.88	12.55	1.50	0.25	6.38	3.04	0.72

According to the on-site production conditions, the mixed burden of VTM comprises 70% V–Ti sinter, 20%V-Ti pellet, and 10% lump ore. The testing schemes of ferrous materials-coke mixed charging are listed in Table 3. It can be seen that the coke mixing ratio are increased from 0% to 50% with a step size of 10%. Figure 1 shows the schematic view of the packing mode of ferrous materials-coke mixing. As shown in Fig. 1, the middle layer consists of ferrous materials mixed with coke. It should be pointed out that all tests are conducted with an equal amount of ferrous material (500 g) and coke (110 g, calculated based on a coke rate of 370 kg/tHM with simulating industrial BF), and the amount of the layer coke placing over and below the ferrous material layer is decreased gradually with increasing coke mixing ratio.

Table 3. Testing schemes of ferrous materials-coke mixed charging (wt%).

No.	Coke mixing ratio	V–Ti sinter ratio	V–Ti pellet ratio	Lump ore ratio
1#	0	70	20	10
2#	10	70	20	10
3#	20	70	20	10
4#	30	70	20	10
5#	40	70	20	10
6#	50	70	20	10

Fig. 1.

Schematic view of packing mode and various ferrous materials-coke mixing ratios.

2.2. Experiment Methods

All tests were carried out with the softening-melting-dropping apparatus, as shown in Fig. 2. The softening-melting-dropping apparatus includes electric furnace, temperature and gas control system, displacement meter system, loading device, pressure transmitter, and sample cooling vessel. An inner diameter of 75 mm graphite crucible, with some Φ8 mm dropping holes on the bottom to ensure molten materials and gas flow passing through packed bed, is used in softening-melting dropping experiment. 500 g iron bearing materials with different coke mixing ratio are charged into graphite crucible. During the softening-melting-dropping experiments, the differential pressure across the packed bed and contraction of the sample bed were continuously recorded at 10 s interval time by displacement meter and pressure transmitter, respectively. The molten slag and iron were cooled and collected in sample cooling vessel. After experiments, the chemical compositions and microstructure of dropped samples were analyzed by chemical analysis method and SEM-EDS. According to the chemical compositions of dropped slag, the slag samples synthesized with the analytical reagent oxides of CaO, SiO₂, MgO, Al₂O₃, TiO₂, and V₂O₅ were premelted in a molybdenum-lined graphite crucible using a laboratory-scale induction furnace under Ar atmosphere. Then, the pre-melted slag samples were quenched and crushed into powder for the subsequent viscosity measurements.^23,24)

Fig. 2.

Schematic view of the apparatus of softening-melting-dropping experiment.

To simulate the reducing conditions in the actual BF, the temperature profile, gas profile, and loading profile are listed in Table 4. The working temperature starts from 20°C to 1570°C with varying heating rate. During experiments, the total gas flow rate is 15 NL/min consisting of N₂, CO and CO₂ with varying gas composition at different temperature stage. To better understand the reduction, softening-melting characteristics and permeability of the mixed burden, a series of interrupted tests were also conducted. When the temperature of packed bed reached the pre-set value, the gas supply and power supply were stopped immediately, and the nitrogen gas with a flow rate of 5 NL/min was introduced to the reaction tube to cool down the mixed burden and to discharge the reduction gas as soon as possible.

Table 4. Experimental conditions for softening-melting dropping experiments.

Heating time/min	40	50	40	>120
Load/(kg/cm2)	0.5		1.0
Gas composition and flow rate	N2: 3 L/min	N2: 60% 9 L/min; CO: 26% 3.9 L/min; CO2: 14% 2.1 L/min	N2: 70% 10.5 L/min; CO: 30% 4.5 L/min
Heating rate/(°C/min)	10°C/min to 400°C	10°C/min to 900°C	3°C/min to 1020°C	5°C/min to 1570°C

3. Results and Discussion

3.1. Softening-melting Behavior

To evaluate the softening-melting behavior of mixed burden, some characteristic temperatures were defined in this research. T₄, the temperature at which the contraction of mixed burden reaches 4%, is defined as the softening start temperature due to the fact that the contraction of mixed burden rises obviously when the temperature reaches T₄. When the contraction of iron bearing burden reaches about 40%, the iron bearing particles are surface cohesive and the particles are gradually boned into a whole which results in the pressure drop starts rising gradually.^25,26) Therefore, T₄₀, the temperature at which the contraction of mixed burden reaches 40%, is defined as the softening end temperature.

The typical results of the softening start temperature T₄, softening end temperature T₄₀, and softening temperature interval (T₄₀-T₄) of mixed burden with different coke-ore mixing ratio are shown in Fig. 3. It can be seen that T₄ remained basically unchanged at about 1140°C with increasing coke-ore mixing ratio. T₄₀ increased firstly with increasing coke-ore mixing ratio from 0% to 20%, then only slightly changed with increasing coke-ore mixing ratio from 20% to 50%. The softening temperature interval (T₄₀-T₄) widened firstly, and then kept unchanged. A wider softening temperature interval is beneficial for BF smelting VTM, as more reduction can occur before the iron-bearing burden begin to melt and lose permeability.^27,28)

Fig. 3.

Effect of coke-ore mixing ratio on softening behavior of mixed burden.

With further increasing the temperature, the iron bearing burden starts meting and the contraction continues rising steeply. As the melting of iron bearing burden, the gas permeability of packed bed would deteriorate obviously. Therefore, the temperature accompanying the pressure drop with a substantial increasing was defined as melting start temperature T_s in this study.^23,24) The dropping temperature T_D is the temperature that the pig iron drips from the graphite crucible. The effect of coke-ore mixing ratio on T_s, T_D and the softening temperature interval (T_D-T_S) is exhibited in Fig. 4. It can be seen that T_S increased from 1257.5°C to 1269.9°C firstly with increasing coke-ore mixing ratio from 0% to 20%, then remained basically unchanged when the coke-ore mixing ratio exceeded 20%. T_D decreased obviously from 14717.5°C to 1392.1°C with increasing coke-ore mixing ratio from 0% to 40%. However, T_D performed a massive jump from 1392.1°C to 1400.8°C with increasing coke-ore mixing ratio from 40% to 50%. The softening temperature interval (T_D-T_S) narrows from 139.6°C to 126.5°C firstly when coke-ore mixing ratio ranges from 0% to 20%, then shows a negligible change with increasing coke-ore mixing ratio from 20% to 40%. However, there is a noticeable rise in (T_D-T_S) when coke-ore mixing ratio is increased to 50%. It is propitious to BF operation as the softening temperature interval progressively narrows with increasing coke-ore mixing ratio.²⁹⁾

Fig. 4.

Effect of coke-ore mixing ratio on melting behavior of mixed burden.

The softening temperature T₄₀ and melting start temperature T_S are correlated to the liquidus temperature of primary slag. It is well known that the liquidus temperature of primary slag is concerned with the slag components. Due to the consistent iron bearing burden structure and coke rate, the coke-ore mixing ratio had little influence on the Al₂O₃, TiO₂ and MgO content in the primary slag. On the contrary, with nut coke mixed in ore layer, the reduction of iron-bearing burden would be accelerated, thereby influencing the FeO content in the mixed burden. Therefore, the effect of FeO on the liquidus temperature of primary slag was investigated in this paper. According to the chemical analysis of the slag from the interrupted experiments at 1300°C and 1400°C, the FeO content of the slag in the presence (20%) of nut coke mixing is lower than that without nut coke mixing. The typical phase diagram of CaO–SiO₂–FeO_n could describe the slag forming behavior during the reduction of iron bearing burden under simulating BF conditions,^30,31) as shown in Fig. 5. It can be seen that the liquidus temperature of primary slag increased with decreasing FeO content, which results in the increasing of the softening temperature T₄₀ and melting start temperature T_S. The location and thickness of cohesive zone have a significant influence on the BF operation, especially for the gas permeability. Figure 6 shows the effect of coke-ore mixing ratio on the location and thickness of the cohesive zone of mixed burden. It can be seen that the cohesive zone gradually shifted down with increasing coke-ore mixing ratio from 0% to 20%, and the thickness of cohesive zone narrows first then widened slightly. In general, a lower location and a thinner cohesive zone are propitious to improve permeability of packed bed and to smooth BF operation. Therefore, in view of the cohesive zone, the rational coke-ore mixing ratio range is 20%–30%.

Fig. 5.

Phase diagram of CaO–SiO₂–FeO_n.

Fig. 6.

Effect of coke-ore mixing ratio on location and thickness of cohesive zone of mixed burden.

3.2. Shaft Permeability

The gas permeability of packed bed can be evaluated efficiently by maximum pressure drop and permeability index (S value). The influence of mixed nut coke on the maximum pressure drop of mixed burden during softening-melting experiment is given in Fig. 7. It can be seen that the maximum pressure drop decreased perceptibly from 26247.4 Pa to 6718.7 Pa with increasing coke-ore mixing ratio from 0% to 50%. Generally, the maximum pressure drop generates at the cohesive zone of packed bed. Therefore, the notably decreasing of maximum pressure drop indicates that the permeability of cohesive zone improved obviously with nut coke mixing.

Fig. 7.

Effect of coke-ore mixing ratio on ΔPmax of mixed burden.

Permeability index (S value) was introduced to quantify the permeability of cohesive zone of mixed burden with different coke-ore mixing ratio. The S value can be calculated by the following Eq. (1).   

$$S={\int }_{T_S}^{T_D}(P_m-\mathrm{\Delta }{P}_S)\cdot dT$$(1)

Where T_S and T_D are the softening start temperature and dropping temperature, respectively. P_m and ΔP_S are the pressure drop at certain temperature and the pressure drop at T_S, respectively.

Generally, a smaller S value indicates a greater permeability of mixed burden. Figure 8 presents the effect of coke-ore mixing ratio on S value. It can be seen that the S value decreased from 3423.6 kPa·°C to 425.3 kPa·°C with increasing coke-ore mixing ratio from 0% to 50%, which means the permeability of vanadium-titanium mixed burden could be improved obviously by nut coke mixed charging technique.

Fig. 8.

Effect of coke-ore mixing ratio on S-value of mixed burden.

Permeability depends on porosity and micro and macro structure of sample bed. To clarify the mechanism of the improvement of permeability in cohesive zone, the interrupted softening-melting experiments of mixed burden with a coke-ore mixing ratio of 20% and without coke-ore mixing were carried out. The pre-set value of interrupted temperature was 1200°C, 1300°C and 1400°C. After the sample bed cooling down to room temperature, the liquid resin was injected in the graphite crucible to fix the sample bed and to remain the structure of slag and mixed nut coke. After solidification, the samples bed were cut into two pieces and the cross sections were scanned. The contraction curves of packed bed during softening-melting experiment, combining macro structures of sample bed interrupted at 1200°C, 1300°C and 1400°C, are shown in Fig. 9. It can be seen that the contraction degree of packed bed with 20% coke-ore mixing is obviously smaller than the packed bed without coke-ore mixing. At the same temperature, a smaller contraction degree indicates a higher porosity of sample bed. This phenomenon could be illustrated by the macro structure of sample bed shown in Fig. 9. The mixed nut coke acted as skeleton in the sample bed which guarantees the voids and channels existed and improves the permeability of mixed burden in cohesive zone. On the contrary, the sinter and pellet gradually collapsed and stuck to each other without nut coke mixed, which results in the porosity of packed bed is decreased notably and the gas passing through is blocked.

Fig. 9.

Contraction curves and macrostructures of mixed burden in the presence and absence (20%) of nut coke.

3.3. Dropping Behavior

The dropping behavior of VTM mixed burden includes dropped ratio of slag and iron, dropping temperature T_D, and yield of V in the dropped iron. The slag dropped ratio (DR_slag), iron dropped ratio (DR_iron) and yield of V (Y_v) could be calculated by the following equations.   

$$D{R}_{\text{slag}}=\frac{m_{\text{slag}}}{(w_{\text{CaO}}+w_{{\text{SiO}}_{\text{2}}}+w_{\text{MgO}}+w_{{\text{Al}}_{\text{2}}{\text{O}}_3}+w_{{\text{TiO}}_{\text{2}}})\times m_T}\times 100\%$$(2)

  

$$D{R}_{\text{iron}}=\frac{m_{\text{iron}}}{w_{\text{TFe}}\times m_T}\times 100\%$$(3)

  

$$Y_{\text{v}}=\frac{m_{\text{iron}}\times w_{v^{\prime }}}{w_{\text{V}}\times m_T}\times 100\%$$(4)

Where m_slag, m_iron and m_T are the mass of dropped slag, mass of dropped iron and total iron mass in iron-bearing materials, respectively. w_CaO, w_SiO2 w_MgO, w_Al2O3, w_TiO2, w_TFe and w_V are the concentration of CaO, SiO₂, MgO, Al₂O₃, TiO₂, Fe and V in the iron-bearing materials. w_V’ is the concentration of V in the dropped iron.

The effect of coke-ore mixing ratio on DR_slag and DR_iron is shown in Fig. 10. Both the DR_slag and DR_iron increased firstly then decreased with increasing coke-ore mixing ratio, and reached the maximum value when coke-ore mixing ratio ranges from 20%–30%. It is well known that the DR_slag is primarily concerned with the slag viscosity. Based on the composition of dropped slag, the slag samples for viscosity measurements were synthesized with the analytical reagent oxides of CaO, SiO₂, MgO, Al₂O₃, TiO₂, and V₂O₅. In order to improve the accuracy of viscosity measurements, the analytical reagent oxides roasted and mixed adequately in a certain proportion were pre-melted in a molybdenum-lined graphite crucible using a laboratory-scale induction furnace under Ar atmosphere. Then, the pre-melted slag samples were quenched and crushed into powder for the subsequent viscosity measurements. The viscosity of dropped slag at 1400°C is shown in Fig. 10. It can be seen that the viscosity of dropped slag decreased with increasing coke-ore mixing ratio from 0% to 30% and reached the minimal value at the coke-ore mixing ratio of 30%. Further increasing coke-ore mixing ratio from 30% to 50%, the viscosity performed a massive jump. Generally, a higher viscosity would thicken the slag and deteriorate its fluidity. The thickening and deterioration of the fluidity of slag would decrease the gas and liquidus permeability and finally worsen the dropping behavior, which results in the DR_slag decreased obviously.

Fig. 10.

Effect of coke-ore mixing ratio on the slag and iron dropped ratio.

The iron dropped ratio is mainly concerned with the melting point of iron. As shown in Fig. 11, the carbon content of dropped iron increased with increasing coke-ore mixing ratio, which results in the theoretical melting point of iron decreased from 1312°C to 1237°C. The increasing of carbon content of dropped iron is mainly caused by the promotion of carburization process with nut coke mixing. However, it should be pointed out that the iron and slag dropped ratio decreased sharply when coke-ore mixing ratio is higher than 30%. This phenomenon is mainly caused by the formation of carbide of V and Ti.

Fig. 11.

Effect of coke-ore mixing ratio on carbon content and theoretical melting point of dropped iron.

To investigate the formation of the carbide of V and Ti, the thermodynamics of the reduction of V-bearing phases and Ti-bearing phases are calculated by Factsage 7.0, as shown in Fig. 12. In the experimental temperature range, the V-bearing phases could be reduced to the oxide of low valence states of V (VO, V₂O₃) and carbide of V (VC), and the Ti-bearing phases could be reduced to the oxide of low valence states of Ti (TiO₂, Ti₃O₅, Ti₂O₃) and carbide of Ti (TiC). Based on the thermodynamics analysis, it can be deduced that the increasing of coke-ore mixing ratio, accompanying the strengthen of reduction potential, would promote the generation of carbide of V and Ti.

Fig. 12.

Thermodynamics calculation of reduction process (a) V-bearing phases, (b) Ti-bearing phases. (Online version in color.)

Figure 13 shows the BSE image and EDS mapping of interface between non-dropped iron and slag with 50% coke-ore mixing ratio. It can be seen that the non-dropped slag and iron contains three type phases, namely white phase, grey phase and dark phase. According to the EDS mapping analysis, it can be concluded that the white phase is metallic iron, and the dark phase, containing Ca, Si, Mg, Al, and O, mainly comprises silicate, and the grey phase, rich in V, Ti and C, is the carbide of V and Ti. Noticeably, the carbide of V and Ti is widely distributed across the interface between the molten iron and slag.

Fig. 13.

BSE image and EDS mapping of interface between non-dropped iron and slag with 50% coke-ore mixing. (Online version in color.)

The influence of the carbide of V and Ti on the dropping behavior of slag and iron can be illustrated by the BSE images and EDS analysis of non-dropped slag with different coke-ore mixing ratio, as shown in Fig. 14. The non-dropped slag mainly comprises perovskite, melilite, spinel, and non-dropped iron. With increasing coke-ore mixing ratio from 0% to 20%, the residual non-dropped iron particles in slag gradually decreased due to the decreasing of viscosity of slag and the improvement of permeability. However, with further increasing coke-ore mixing ratio to 50%, the carbide of V and Ti precipitated at the interface across the molten iron and slag. The carbide of V and Ti phase performs a liquidus temperature higher than 2500°C. Previous literatures illustrated that the carbide of V and Ti solid particles would increase the viscosity of slag sharply and thickening the slag once they entered into the molten phases (slag and metal).³²⁾ The worsening of the slag fluidity deteriorates the separation of iron and slag, as shown in Fig. 14(c), which finally results in the decreasing of the dropping behavior.

Fig. 14.

BSE images and EDS analysis of non-dropped slag (a) without coke mixing (b) 20% coke mixing (c) 50% coke mixing (d) EDS of point 1 (e) EDS of point 2 (f) EDS of point 3 (g) EDS of point 4.

Except the deterioration of the dropping behavior of slag and iron with a coke-ore mixing ratio of 50%, the yield of V in dropped iron also performed a similar tendency. Figure 15 shows the effect of coke-ore mixing ratio on yield of V in the dropped iron. It can be seen that the Y_V increased firstly then decreased with increasing coke-ore mixing ratio, and reached the maximum value in the range of 20% to 40%. Especially, the Y_V lowered substantially when coke-ore mixing ratio increased to 50%. The decrease of Y_V could be attributed to the following reasons: (1) The reduction of V-bearing oxides to the carbide of V would decreased the amount of V existed in the dropped iron. (2) The generation of the carbide of V and Ti would deteriorate the fluidity of slag and restrict the mass diffusion of V from slag to iron.

Fig. 15.

Effect of coke-ore mixing ratio on yield of V in the dropped iron.

3.4. Interaction between Iron-bearing Burden and Mixed Nut Coke

The influence of mixed nut coke on softening-melting-dropping behavior could be explained by the interaction between iron-bearing burden and mixed nut coke. Figure 16 shows the reduction degree of mixed burden in the presence and absence (20%) of nut coke at different interrupted temperature. The reduction degree of mixed burden was calculated by following Eq. (5).   

$$\begin{array}{l}\text{Reduction}\mathrm{\hspace{0.17em} }\text{degree}=\frac{T_{\text{O}}-(16n_{\text{FeO}}/72+48n_{{\text{Fe}}_{\text{2}}{\text{O}}_{\text{3}}}/160)\times m^{\prime }}{T_{\text{O}}}\\ \mathrm{\hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em}\hspace{0.5em}}\times 100\%\end{array}$$(5)

Where T_O is the total removable oxygen of mixed burden before reduction, which can be calculated by the chemical compositions of the raw materials listed in Table 1. The n_FeO and n_Fe2O3 are the FeO and Fe₂O₃ concentration of the mixed burden after serious interrupted experiments. The m′ is the total mass of mixed burden after different interrupted experiments.

Fig. 16.

Reduction degree of mixed burden in the presence and absence (20%) of nut coke at different temperature.

As shown in Fig. 16, the reduction degree of mixed burden with 20% nut coke mixed was higher than that of the mixed burden without coke-ore mixing, especially at 1300°C. This phenomenon could be attributed to the following reasons. Firstly, the mixed nut coke acts as skeleton in the mixed burden during softening and melting, which guarantees the formation of voids and gas channels and promotes the reduction gas diffusion.

Secondly, without nut coke mixing, the amount of fayalite (Fe₂SiO₄) increased rapidly due to the existence of large amount of wustite, as shown in Fig. 17. It is well known that the fayalite is hardly reducible compared to iron oxides, and starts softening at about 1150°C and melts completely at about 1200°C. Therefore, the mixed burden tended to stick and collapse and finally formed a structure that represented a resistance for the gas diffusion and lowering the reduction rate.

Fig. 17.

BSE images and EDS analysis of mixed burden in the presence and absence (20%) of nut coke at 1200°C.

Figure 18 shows the BSE images and EDS mapping of mixed burden in the presence and absence (20%) of nut coke at 1300°C. As shown in Fig. 18(a), the mixed burden with nut coke mixed could be divided into three parts, namely nut coke zone, metallic iron zone and iron oxides zone. The metallic iron zone is located near the interface between nut coke and iron-bearing burden, which indicates that the mixed nut coke could strengthen the reduction efficiently. Without coke-ore mixing, as shown in Fig. 18(b), a relative narrower metallic iron zone formed only at the outer side of iron-bearing burden, and the metallic iron decreased sharply towards the core of iron-bearing burden. Besides, the iron-bearing burden without coke-ore mixing performed a microstructure with a high density, which represented a resistance for the reduction gas diffusion.

Fig. 18.

BSE images and EDS mapping of mixed burden in the presence and absence (20%) of nut coke at 1300°C (a) with 20% coke-ore mixing (b) without coke-ore mixing (c) EDS mapping of interface between metallic iron and iron oxides zone. (Online version in color.)

The carburization process between iron and coke has a significant effect on the dropping behavior of mixed burden. To explore carburization process between iron and mixed nut coke, the mixed burden samples obtained from the interrupted softening-melting tests at 1400°C were analyzed by SEM-EDS, as shown in Fig. 19. It can be seen that the bright phase, dark phase, and grey phase are metallic iron shell, nut coke, and slag, respectively. A carbon concentration gradient at the border of metallic iron can be observed from the line scan across the interface between metallic iron and nut coke, which indicates that the carburization process in the metallic iron could be promoted efficiently due to the closely located nut coke in the mixed burden.

Fig. 19.

BSE image and EDS analysis of the interface between iron and nut coke in the mixed burden samples with a coke-ore mixing ratio of 20% at 1400°C. (Online version in color.)

In conclusion, the interaction between iron-bearing burden and mixed nut coke could be summarized in four parts, as shown in Fig. 20. Firstly, the mixed nut coke could strengthen the reduction of mixed burden efficiently. In case of the ore layer without nut coke mixing, the ore was reduced by CO and only R1 occurred. However, in case of ore layer with nut coke mixing, the carbon gasification also occurred and a higher local CO pressure could be provided, which would accelerate reduction of mixed burden. Secondly, the mixed nut coke closed to the iron-bearing burden could promote the carburization process efficiently. Obviously, the promotion of carburization of iron would improve the dropping behavior of mixed burden. Thirdly, the permeability of packed bed would be improved significantly. In case of ore layer with nut coke mixing, the gas could pass through the voids, channels and mixed coke matrix, and the cohesive zone could be dispersed effectively, thereby improving the permeability of packed bed sufficiently. By contrast, the packed bed tends to collapse and stick without nut coke mixing, which would block the gas flow and deteriorate the permeability of packed bed. However, the coke mixing ratio is strictly limited due to the fact that an excessive coke mixing ratio would results in the formation of carbide of Ti. The excessive mixing coke would promote the reactions R3–R7, as shown in Fig. 20. The TiC mainly precipitated at the interface of the metal-slag and coke-slag. It has been reported that the effect of TiC particles on the viscosity of slag is much greater than that expected from Einstein–Roscoe theory, which means the viscosity of slag would increase steeply with the formation of TiC.³⁰⁾ Therefore, the formation of TiC would thicken the slag, which would obviously worsen the gas and liquidus permeability and finally deteriorate the dropping behavior of the molten slag and iron.

Fig. 20.

Schematic illustrations of the interaction mechanism between mixed nut coke and iron-bearing burden. (Online version in color.)

4. Conclusion

In this paper, the effect of coke-ore mixing ratio on softening-melting-dropping performance and permeability of packed bed was investigated under BF conditions firstly. Then, the interaction mechanism between mixed nut coke and iron-bearing burden was illustrated. The following conclusions can be obtained from this study.

(1) The softening-melting-dropping behavior of vanadium-titanium mixed burden could be improved efficiently when coke-ore mixing ratio is 20%.

(2) With further increasing coke-ore mixing ratio to 50%, the strengthen of reduction potential would accelerate the formation of carbide of V and Ti, which would deteriorate the fluidity of slag and worsen the dropping behavior of mixed burden and lower the yield of V in liquid iron.

(3) The mixed nut coke acts as skeleton in packed bed, which guarantee the voids and gas channels existed and improve the permeability of mixed burden efficiently.

(4) The reducibility of mixed burden could be improved by the mixed nut coke, which was mainly attributed to the decreasing of the amount fayalite and the promotion of reduction gas diffusion.

(5) The mixed nut coke closely locted to the iron-bearing burden could promote the carburization process efficiently, which would decrease the dropping temperature and improve the dropping behavior of mixed burden.

Acknowledgements

The authors are especially thankful to National Natural Science Foundation of China (51574067), China Postdoctoral Science Foundation (2016M601321) and and Fundamental Research Funds for the Central Universities (N172503016).

References

• 1)   R.  Paunova: Metall. Mater. Trans. B, 33B (2002), 633.
• 2)   J.  Tang,  M. S.  Chu,  C.  Feng,  F.  Li,  Y. T.  Tang and  Z. G.  Liu: ISIJ Int., 57 (2017), 1156.
• 3)   Z. G.  Liu,  M. S.  Chu,  H. T.  Wang,  W.  Zhao and  X. X.  Xue: Int. J. Miner. Metall. Mater., 23 (2016), 25.
• 4)   W.  Zhao,  H. T.  Wang,  Z. G.  Liu,  M. S.  Chu,  Z. W.  Ying and  J.  Tang: Steel Res. Int., 88 (2017), 1.
• 5)   A. D.  Manshadi,  J.  Manuel,  S.  Hapugoda and  N.  Ware: ISIJ Int., 54 (2014), 2189.
• 6)   W.  Zhao,  H. T.  Wang,  Z. G.  Liu,  M. S.  Chu,  Z. W.  Ying and  J.  Tang: JOM, 69 (2017), 1737.
• 7)   H. G.  Du: Principle of Blast Furnace Smelting Vanadium-Titanium Magnetite, Science Press, Beijing, (1996), 1.
• 8)   S.  Samanta,  S.  Mukherjee and  R.  Dey: JOM, 67 (2015), 467.
• 9)   Y.  Kashihara,  Y.  Morikawa,  T.  Sato,  N.  Ishiwata and  M.  Sato: ISIJ Int., 55 (2015), 1165.
• 10)   S.  Matsuhashi,  H.  Kurosawa,  S.  Natsui,  T.  Kon,  S.  Ueda,  R.  Inoue and  T.  Ariyama: ISIJ Int., 52 (2012), 1990.
• 11)   Z. J.  Liu,  J. L.  Zhang and  T. J.  Yang: ISIJ Int., 55 (2015), 1146.
• 12)   K.  Terui,  Y.  Kashihara,  T.  Hirosawa and  T.  Nouchi: ISIJ Int., 57 (2017), 1804.
• 13)   K.  Terui,  Y.  Kashihara,  T.  Hirosawa and  T.  Nouchi: Tetsu-to-Hagané, 103 (2017), 86.
• 14)   Y.  Kashihara,  Y.  Sawa and  M.  Sato: Tetsu-to-Hagané, 100 (2014), 312.
• 15)   Y.  Kashihara,  Y.  Iwai,  N.  Ishiwata,  N.  Oyama,  H.  Matsuno,  H.  Horikoshi,  K.  Yamamoto and  M.  Kuwabara: Tetsu-to-Hagané, 102 (2016), 9.
• 16)   M.  Gono,  K.  Iwatsuki,  M.  Takasaki,  K.  Nojima and  T.  Miwa: Tetsu-to-Hagané, 68 (1982), S709.
• 17)   E. A.  Mousa,  A.  Babich and  D.  Senk: ISIJ Int., 51 (2011), 350.
• 18)   S.  Watakabe,  K.  Takeda,  H.  Nishimura,  S.  Goto,  N.  Nishimura,  T.  Uchida and  M.  Kiguchi: ISIJ Int., 46 (2006), 513.
• 19)   S.  Watakabe,  K.  Takeda,  H.  Nishimura,  S.  Goto,  N.  Nishimura,  T.  Uchida and  M.  Kiguchi: Tetsu-to-Hagané, 92 (2006), 901.
• 20)   A.  Murao,  Y.  Kashihara,  N.  Oyama,  M.  Sato,  S.  Watakabe,  K.  Yamamoto and  Y.  Fukumoto: Tetsu-to-Hagané, 102 (2016), 614.
• 21)   A. V.  Kuzin,  S. L.  Yaroshevskiy and  N. S.  Chlaponin: Proc. Int. Conf. on Theory and Practice of Hot Metal Production, KGGMK Krivorozhstal, Krivoy Rog, (2004), 369.
• 22)   S. L.  Yaroshevskii,  V. A.  Nozdrachev,  A. P.  Chebotarev,  V. A.  Rudenko,  S. A.  Feshchenko,  A. M.  Kuznetsov,  V. P.  Padalka,  N. S.  Khlaponin and  A. V.  Kuzin: Metallurgist, 44 (2000), 598.
• 23)   R. Z.  Xu,  J. L.  Zhang,  Z. Y.  Wang and  K. X.  Jiao: Steel Res. Int., 88 (2017), 1.
• 24)   C.  Feng,  M. S.  Chu,  J.  Tang,  Y. T.  Tang and  Z. G.  Liu: Steel Res. Int., 87 (2016), 1274.
• 25)   S. L.  Wu,  B. Y.  Tuo,  L. H.  Zhang,  K. P.  Du and  Y.  Sun: Steel Res. Int., 85 (2014), 233.
• 26)   C. E.  Loo,  L. T.  Matthews and  D. P.  O’dea: ISIJ Int., 51 (2011), 930.
• 27)   W.  Zhao,  M. S.  Chu,  H. T.  Wang,  Z. G.  Liu and  Y. T.  Tang: Int. J. Miner. Metall. Mater., 23 (2016), 501.
• 28)   G. J.  Cheng,  X. X.  Xue,  T.  Jiang and  P. N.  Duan: Metall. Mater. Trans. B, 47B (2016), 1713.
• 29)   M. S.  Chu: Raw Fuels and Auxiliary Materials in Ferrous Metallurgy, 1st ed., Metallurgical Industry Press, Beijing, (2010), 158.
• 30)  Verein Deutscher Eisenhüttenleute: Slag Atlas, 2nd ed., Verlag Stahleisen GmbH, Düsseldorf, (1995), 126.
• 31)   M.  Matsumura,  M.  Hoshi and  T.  Kawaguchi: ISIJ Int., 45 (2005), 594.
• 32)   Y. X.  Liu,  J. L.  Zhang,  G. H.  Zhang,  K. X.  Jiao and  K. C.  Chou: Ironmaking Steelmaking, 44 (2017), 609.