Effect of B₂O₃ on ESR Slag Viscosity for Remelting 9CrMoCoB Steel

Abstract

The viscosity of the slag constituted by 55%CaF₂–20%CaO–3%MgO–22%Al₂O₃–x%B₂O₃ (mass fraction: x ≤ 3) was measured by rotating cylinder method during the electroslag remelting (ESR) of 9CrMoCoB steel to analyze the effects of the B₂O₃ content on slag viscosity, break point temperature, and activation energy for viscous flow. Meanwhile, the slag’s structural characteristics were studied by Raman spectrometry. Results indicated that when T ≥ 1663 K, the viscosity changed gently with temperature, showing the value of about 0.055 Pa·s, which was rarely affected by w(B₂O₃). However, with the increase of w(B₂O₃) in the slag, the Ca₃B₂O₆ with a low melting point was formed, reducing the slag’s break point temperature from 1628 K to 1563 K. According to the NBO/T value and Raman spectra, B₂O₃, as a network generating oxide, improved the slag’s degree of polymerization, thus strengthening and complicating its network structure at high temperatures. In addition, as the w(B₂O₃) increased from 0 to 3%, the activation energy for viscous flow rose from 64 kJ·mol⁻¹ to 94 kJ·mol⁻¹ in the slag.

1. Introduction

At present, vigorously developing ultra-supercritical energy generating units is the main direction for optimizing the coal-fired power structure.^1,2,3) The 9CrMoCoB steel is a 9% Cr heat-resistant steel served under the ultra-supercritical conditions of 620°C and 32 MPa, and is a candidate material for the rotor of thermal power units.^4,5) In this instance, power generation efficiency can be increased to 47.5% in such thermal generator sets, and the CO₂ emissions can be reduced by about 10% compared with the current global average level.⁶⁾ At present, the ESR technique⁷⁾ has been used to manufacture large 9CrMoCoB ingots by JCFC and JSW in Japan, and Doosan in Korea.^8,9) However, as the alloy element B (0.008%–0.011%) can be burnt out easily during the ESR of 9CrMoCoB steel, a proper amount of B₂O₃ shall be added into the slag to address this issue.^10,11) In this process, the flow velocity in the slag pool can be affected by slag viscosity, which greatly influences the physicochemical reaction between slag and steel as well as the removal of inclusions from the steel.^12,13,14,15) Therefore, it is necessary to research the effects of B₂O₃ content on this viscosity during the ESR of 9CrMoCoB steel.

As a network generating oxide,^16,17) the B₂O₃ can affect the slag viscosity, which has already been studied by many. Xu et al.¹⁸⁾ indicated that this oxide could make the glass structure more compact in the slag. Ren et al.¹⁹⁾ analyzed the influence of B₂O₃ on the viscosity of high Ti-bearing blast furnace slag at 1773 K–1593 K and found that the slag fluidity was improved as the B₂O₃ reduced the viscosity of such slag containing the unconsumed pulverized coal (UPC). Huang et al.²⁰⁾ researched the impact of B₂O₃ on the viscosity of the ladle slag with low SiO₂ content and concluded that this slag’s viscosity, activation energy for viscous flow, and breakpoint temperature plummeted with the increase of w(B₂O₃).

However, the effects of B₂O₃ on the viscosity of CaF₂-based slag are rarely reported. Therefore, we selected the slag composed of 55%CaF₂–20%CaO–3%MgO–22%Al₂O₃–x%B₂O₃ (mass fraction: x ≤ 3) as the object of this study to research the influence of B₂O₃ content on the slag viscosity during the ESR of 9CrMoCoB steel by rotating cylinder method, seeking to optimize the slag composition and determine the proper amount of B₂O₃ in this process and to provide a theoretical basis for the formulation of slag system.

2. Experiment

2.1. Experimental Slag

The 55%CaF₂–20%CaO–3%MgO–22%Al₂O₃-based slag was added with varying amounts of B₂O₃ to measure its viscosity at high temperatures, the designed slag experiment scheme and the chemical composition of the final slag are shown in Table 1. In this experiment, we prepared the slag using analytical chemical reagents, including w(CaO) ≥ 97%, w(MgO) ≥ 98%, w(CaF₂) ≥ 98.5%, w(Al₂O₃) ≥ 98.5%, and w(B₂O₃) ≥ 98%.

Table 1. Experimental scheme and chemical compositions of experimental slag samples.

	Initial designed compositions/wt%					Final compositions/wt%
	CaF2	CaO	Al2O3	MgO	B2O3	CaF2	CaO	Al2O3	MgO	B2O3
0#	55	20	22	3	0.0	53.23	21.46	21.97	3.12	0.00
1#	55	20	22	3	0.5	53.32	21.40	21.96	3.12	0.48
2#	55	20	22	3	1.0	53.26	21.36	21.97	3.12	0.97
3#	55	20	22	3	1.5	53.24	21.36	21.99	3.11	1.45
4#	55	20	22	3	2.0	53.28	21.34	21.99	3.11	1.94
5#	55	20	22	3	3.0	53.25	21.33	21.99	3.13	2.89

2.2. Experimental Equipment and Methods

Figure 1 illustrates the measurement of slag viscosity at high temperatures. In this experiment. The experimental equipment is BCT1700, which conforms to the American ASTMC965 and ASTMC1276 standard methods, the measuring range is 1.2 cP–5·10⁷ cP, the measuring accuracy of the inspection instrument is ±1% and the reproducibility is ±0.2%. within. The experimental equipment was calibrated with standard silicone oil at room temperature and at a rotation speed of 200 rpm. We adopted the molybdenum crucible (inner diameter: 40 mm) and rotor (in the cylindrical section, the diameter and height were 15 mm and 25 mm, respectively); and the angles at both the top and tail of the rotor were 120°. Besides, all the experimental parameters were kept constant to guarantee the reliability of the results. The distance was set at 10 mm from the bottom of the rotor to that of the crucible, and the slag’s liquid level was 3–5 mm higher than the narrower part at the tail of the rotor. Based on preliminary experiments, about 140 g slag was needed in each viscosity measurement at high temperatures.

Fig. 1.

Schematic of high temperature viscometer and its auxiliary device. (Online version in color.)

After being fed with the uniformly mixed powder, the crucible was heated automatically according to the parameters of temperature control set on the computer (5 K·min⁻¹) and then kept at 1733 K for 30 min. After that, the molybdenum rotor was inserted into the crucible slowly to be concentric with it. At this point, the bottom of the rotor was about 10 mm away from this crucible. After aligning and determining their positions, we made the rotor stir the slag evenly at the rotation speed of 200 r min⁻¹ for 20 min and then lowered the temperature at the rate of 3°C min⁻¹ to test the viscosity, which finally exhibited the value of 3.5 Pa s. Viscosity measurement is performed every 5 K interval. For each measurement, first keep the furnace chamber at the target temperature for 5 minutes to stabilize the slag temperature and viscosity. Then, the viscosity data was recorded every 10 s for a duration of 3 min. During the whole period of heating and testing, the high purity argon gas was supplied constantly at the flow rate of 0.3 NL min⁻¹ to ensure that the experiment was carried out normally.

3. Results and Analysis

3.1. Effect of B₂O₃ Content on Slag Viscosity

We tested the change of viscosity with temperature in six groups of slag according to the above steps and presented the results in Fig. 2.

Fig. 2.

Viscosity of the molten slag versus temperature. (Online version in color.)

As observed in Fig. 2, the slag viscosity decreased with the temperature increase, with each η-T curve showing a breakpoint. When the temperature was higher than the value at this point, the viscosity changed slowly, indicating favorable fluidity and thermal stability of the slag. However, when the temperature dropped below this value, the viscosity soared to 3.5 Pa·s within a small temperature range. At this point, the slag melt changed from Newtonian fluid to non-Newtonian one,²¹⁾ which undermined the slag’s fluidity and stability. When T ≥ 1663 K, the w(B₂O₃) had little impact on the viscosity, which changed gently with temperature and showed a similar value of about 0.055 Pa·s in all groups.

Based on the effects of B₂O₃ content on the slag viscosity at high temperatures in Fig. 2, we obtained Fig. 3: the viscosity increased gradually as this content rose from 0 to 3%. Thereinto, the viscosity rose slightly within the range from 0 to 2% but increased significantly from 2% to 3%. For example, the viscosity increased by 0.009 Pa·s at 1688 K as the w(B₂O₃) rose from 0 to 2%, while it increased by 0.010 Pa·s with the content rising from 2% to 3%.

Fig. 3.

Effect of w(B₂O₃) on the slag viscosity at high temperature. (Online version in color.)

The network generating oxide B₂O₃ can break down into trihedral ([BO₃]³⁻) and tetrahedral ([BO₄]⁵⁻) structures, which, as new units, will react with aluminate to enhance the degree of polymerization in the network structure of the slag, thus increasing the viscosity at high temperatures. The slag structure under different B₂O₃ contents will be elaborated on in later chapters.

3.2. Effect of B₂O₃ Content on Turning Point Temperature of Molten Slag

The break point temperature (T_Br) of the slag viscosity refers to the value at the tangent point between the 45° line and η-T curve.²²⁾ When the temperature was lower than this value, the viscosity increased remarkably. In a sense, the T_Br, which represents the boundary between good and poor slag fluidity, is an important feature indicating the viscosity.

According to the slag’s break point temperature in Table 2, the T_Br reduced gradually as the w(B₂O₃) got higher, showing the maximum value of 1628 K at the w(B₂O₃) of 0 and the minimum one of 1563 K at the w(B₂O₃) of 3%. Thereinto, when the w(B₂O₃) increased from 0 to 1%, the T_Br reduced by 25 K, from 1628 K to 1603 K; within the range from 1% to 2%, the T_Br reduced by 30 K, from 1603 K to 1573 K; and with the w(B₂O₃) rising from 2% to 3%, the T_Br decreased by 10 K, from 1573 K to 1563 K.

Table 2. The break point temperature of the slag.

	0#	1#	2#	3#	4#	5#
TBr/K	1628	1618	1603	1593	1573	1563

To analyze the impact of w(B₂O₃) on the meltability of the slag, we first cooled the high-temperature slag in the air and then dried it in an oven for 48 hours at 110°C to drain the water. After that, the phase of all six slag samples was analyzed by X-ray diffraction (XRD), with the results shown in Fig. 4.

Fig. 4.

XRD patterns of the slags quenched at the desired temperatures. (Online version in color.)

As indicated in Fig. 4, all six groups of slag showed the crystallization phase CaF₂ at the 1 peak. But at the 2 peak, the phase was Ca₁₂Al₁₄O₃₂F₂ and MgO in groups 0#–2#, Ca₁₂Al₁₄O₃₂F₂, MgO and MgAl₂O₄ in group 3#, and Ca₁₂Al₁₄O₃₂F₂ and MgAl₂O₄ in groups 4# and 5#. Besides, the crystallization phase Ca₃B₂O₆ was observed in group 5# but not detected in groups 1#–4# at the 3 peak. Perhaps this is because the Ca₃B₂O₆ can only be precipitated at lower temperatures, which is prejudicial to the formation and growth of crystal nuclei. Meanwhile, the amount of Ca₃B₂O₆ precipitated from the slag in groups 1#–4# was too small to be detected. However, as the w(B₂O₃) reached 3% in group 5#, this phase was precipitated at such an increasing amount that it could be observed. This indicated that the increase of w(B₂O₃) contributed to the formation of phase Ca₃B₂O₆ with a low melting point, which lowered the slag’s melting temperature and improved its meltability, thus reducing the T_Br.

The slag viscosity of all groups was suitable for the ESR of 9CrMoCoB steel at high temperatures. The lower the break point temperature of the slag, the smaller the change of viscosity with temperature, which was conducive to the stability of ESR. However, this temperature only decreased by 10 K when the w(B₂O₃) rose from 2% to 3%. Therefore, it was a better choice to add 2% B₂O₃ into the slag constituted by 55%CaF₂–20%CaO–3%MgO–22%Al₂O₃–x%B₂O₃ (x ≤ 3).

3.3. Effect of B₂O₃ Content on High Temperature Structure of Molten Slag

The slag’s high-temperature structure directly impacts its viscosity, and different oxides play various roles when the slag forms its network structure at high temperatures. For example, the network generating oxide B₂O₃, which is a major component of this structure, can generate a bridging oxygen bond at high temperatures, thus improving the degree of polymerization and viscosity of the slag.²³⁾ By contrast, Al₂O₃ is an amphoteric oxide, serving as network modifying or generating oxide in the structure.²⁴⁾

Mills et al.²⁵⁾ measured the slag’s degree of polymerization based on the NBO/T value (the ratio of the amount of non-bridging oxygen to that of tetrahedral polymers in the slag, namely the amount of such oxygen in each tetrahedron). The larger the NBO/T value, the lower the degree of polymerization. Conversely, a smaller value implies a higher degree.²⁶⁾ This degree is characterized by NBO/T in the slag’s high-temperature structure,^27,28) as presented in Eq. (1):   

$$\text{NBO}/T=\frac{\text{2}{x}_{\text{CaO}}+\text{2}{x}_{{\text{CaF}}_{\text{2}}}-\text{2}{x}_{{\text{Al}}_{\text{2}}{\text{O}}_{\text{3}}}+\text{2}{x}_{\text{MgO}}}{\text{2}{x}_{{\text{Al}}_{\text{2}}{\text{O}}_{\text{3}}}+\text{2}{x}_{{\text{B}}_{\text{2}}{\text{O}}_{\text{3}}}}$$(1)

Where, NBO: Unbridged oxygen number; T: Cation number of quartic coordination; x: Molar fraction of component.

Figure 5 can be obtained by substituting the data into Eq. (1).

Fig. 5.

The value of NBO/T versus the B₂O₃ additional amount. (Online version in color.)

In this figure, as the w(B₂O₃) increased in the slag, the NBO/T value gradually reduced, causing the degree of polymerization to present a rising trend. This is because there are more cations of quartic coordination due to the addition of the network generating oxide B₂O₃, which compacts the network structure and improves the degree of polymerization. That is, the slag structure becomes more complex at high temperatures.

To determine this structure, we heated the slag of six groups to 1723 K according to the parameters in Table 1 for water quenching. Then, the slag was ground into the size below 100 μm and analyzed by Raman spectrometry. The corresponding spectra for these groups at room temperature were illustrated in Fig. 6, which indicated that the peak intensity weakened at 523 cm⁻¹; while the peak was formed from nothing and gradually intensified near 930 cm⁻¹, with its value shifting from 920 cm⁻¹ to 940 cm⁻¹. Fan et al.²⁹⁾ proved that the peak at 523 cm⁻¹ was formed by the Al–O stretching vibration in [AlO4]⁵⁻, while that near 930 cm⁻¹ was formed by the Al_IV–O–B_III bending vibration.^30,31)

Fig. 6.

Raman spectra of six groups of slag. (Online version in color.)

I It is generally recognized that the slag’s degree of polymerization is affected by changes in the relative content and conversion amount of different structural units.³²⁾ Thereinto, this relative content can be deduced from the area of corresponding Raman spectral band. Figures 7 and 8 describe the Raman spectra fitted by peak splitting in groups 0#–5#.

Fig. 7.

Al–O Peak Fitting of Raman Spectrum. (Online version in color.)

Fig. 8.

Al_IV–O–B_III Peak Fitting of Raman Spectrum. (Online version in color.)

As shown in Fig. 7, the red curve is the fitting of Al–O peaks. It can be seen that as w(B₂O₃) in the slag increases, the height of the Al–O peak gradually decreases. That is, the Al–O peak shows a weakening trend.

As shown in Fig. 8, the blue curve is the peak fitting situation of Al_IV–O–B_III. It can be seen that with the increase of w(B₂O₃) in the slag, the peaks of Al_IV–O–B_III appear from scratch, and the width and height of the peaks are gradually increasing. That is, the peak of Al_IV–O–B_III shows a tendency to become stronger.

Based on the fitting results of Raman spectra by peak splitting in all groups, we analyzed the change in the area ratio of these spectra corresponding to Al–O and Al_IV–O–B_III, as shown in Fig. 9.

Fig. 9.

The area ratio of the corresponding Raman band with the increase of w(B₂O₃). (Online version in color.)

With the network generating oxide B₂O₃ added into the slag, there was an increasing number of network-forming cation B³⁺, which improved the degree of polymerization. It was observed in Fig. 9 that with the increase of w(B₂O₃), the proportion of Al–O decreased, while that of Al_IV–O–B_III increased. In other words, due to the addition of B₂O₃, a network structure composed of [AlO4] tetrahedral network and [BO3] triangular chain was formed, which enhanced the degree of polymerization and complicated the high-temperature structure of the slag. Under such circumstances, the viscous flow faced larger frictional resistance in this slag, thus increasing the viscosity at high temperatures.

3.4. Effect of B₂O₃ Content on Activation Energy of Slag Viscous Flow

As another important feature indicating the viscosity, the activation energy for viscous flow (E_η), which is taken as the basic physical property data, reflects not only the viscosity of the slag with different components, but also the sensitivity of such viscosity to various temperatures in the slag with a certain composition. What’s more, the variation of slag structure can be predicted to a certain extent by that of this activation energy because the E_η value will be constant if there is no change in major viscous components in this structure.³³⁾

According to Boltzmann distribution law, the relationship between the slag viscosity η and temperature T can be expressed by the Arrhenius equation, as shown in Eq. (2):   

$$ln\mathrm{\hspace{0.17em} }\eta =ln\mathrm{\hspace{0.17em} }\mathrm{A}+({\mathrm{E}}_{\eta }/\text{R})\times 1/\mathrm{T}$$(2)

where, η: viscosity, Pa·s; T: Kelvin temperature, K; E_η: viscous flow activation energy, J⸱mol⁻¹; A: frequency factor; R: standard molar gas constant, 8.314 J⸱mol⁻¹·K⁻¹.

Based on the above analysis, the lnη and 1/T were fitted in Fig. 10.

Fig. 10.

Calculation of the activation energy for viscous flow of the six slags. (Online version in color.)

According to Fig. 10, there was a good linear relationship between lnη and T⁻¹ in six groups of slag, showing the correlation coefficient R of about 0.99, which indicated that the viscosity was closely correlated with the Arrhenius behavior. In this case, the activation energy for viscous flow was regarded as a constant and calculated by fitting analysis in these groups, as presented in Fig. 11.

Fig. 11.

The activation energy of the six slags versus the B₂O₃ additional amount. (Online version in color.)

We observed from Fig. 11 that as the w(B₂O₃) changed from 0 to 3%, the E_η value rose from 64 kJ·mol⁻¹ to 94 kJ·mol⁻¹, indicating the increase of the energy barrier for viscous flow and the formation of some complex structural components in the slag, which was consistent with the change of viscosity and slag structure at high temperatures in previous chapters.

4. Conclusion

(1) When the w(B₂O₃) rose from 0 to 3% at 1663 K or higher temperatures, the slag viscosity increased with the continuous addition of B₂O₃.

(2) With the increase of w(B₂O₃), the Ca₃B₂O₆ with a low melting point was formed in the slag, which improved this slag’s meltability and reduced its break point temperature from 1628 K to 1563 K. Hence, it was more suitable to control the w(B₂O₃) at about 2%.

(3) Based on the NBO/T value and Raman spectra, we summarized that the increase of w(B₂O₃) helped improve the slag’s degree of polymerization and complicate its high-temperature structure, which enhanced the frictional resistance of the viscous flow and damaged the slag fluidity.

(4) When the w(B₂O₃) increased from 0 to 3% at high temperatures, the slag viscosity rose slightly, and at the same time, the E_η value increased from 64 kJ·mol⁻¹ to 94 kJ·mol⁻¹.

(5) To sum up, the w(B₂O₃) shall be set at about 2% in the slag constituted by 55%CaF₂-20%CaO-3%MgO-22%Al₂O₃-x%B₂O₃ (x ≤ 3) because, under this condition, the viscosity will be kept steady within a larger temperature range. Thus, the ESR can be conducted steadily to form a thin and uniform slag shell.

Acknowledgments

This project was supported by the National Nature Science Foundations of China (grant No. 51974076 and 2016YFB0300203). Also, this project supported by The “Innovation & Entrepreneurship” Talents Introduction Plan of Jiangsu Province in 2018.

References

• 1)   N.  Zhou,  D.  Fridley,  N. Z.  Khanna,  J.  Ke,  M.  McNeil and  M.  Levine: Energy Policy, 53 (2013), 51. https://doi.org/10.1016/j.enpol.2012.09.065
• 2)   L. B.  Mei,  H. W.  Shen,  S. Y.  Wang,  L. T.  Zhang and  F.  Sun: Therm. Turbine, 41 (2012), 183. https://doi.org/10.13707/j.cnki.31-1922/th.2012.03.004
• 3)   Z. D.  Liu,  S. C.  Cheng,  G. B.  Tang,  H. S.  Bao,  G.  Yang and  Y.  Gan: Iron Steel, 46 (2011), 1. https://doi.org/10.13228/j.boyuan.issn0449-749x.2011.03.001
• 4)   J. T.  Kim and  B. G.  Kong: Mater. Sci. Forum, 654–656 (2010), 398. https://doi.org/10.1179/174892408x382879
• 5)   N.  Blaes,  B.  Donth and  D.  Bokelmann: Energy Mater., 2 (2007), 207. https://doi.org/10.1179/174892408x382879
• 6)   S. C.  Chetal,  T.  Jayakumar and  A. K.  Bhaduri: Proc. Indian Natl. Sci. Acad., 81 (2015), 739. https://doi.org/10.16943/ptinsa/2015/v81i4/48294
• 7)   D. S.  Kim,  G. J.  Lee,  M. B.  Lee,  J. I.  Hur and  J. W.  Lee: IOP Conf. Ser. Mater. Sci. Eng., 143 (2016), 012002. https://doi.org/10.1088/1757-899X/143/1/012002
• 8)   R.  Viswanathan and  W.  Bakker: J. Mater. Eng. Perform., 10 (2001), 96. https://doi.org/10.1361/105994901770345394
• 9)   V.  Weber,  A.  Jardy,  B.  Dussoubs,  D.  Ablitzer,  S.  Rybéron,  V.  Schmitt,  S.  Hans and  H.  Poisson: Metall. Mater. Trans. B, 40 (2009), 271. https://doi.org/10.1007/s11663-008-9208-9
• 10)   D.  Hou,  Z. H.  Jiang,  T. P.  Qu,  D. Y.  Wang,  F. B.  Liu and  H. B.  Li: J. Iron Steel Res. Int., 26 (2019), 20. https://doi.org/10.1007/s42243-018-0107-2
• 11)   T. J.  Wen,  H. J.  Zhang,  X. J.  Li,  L.  Yu,  Y.  Ren,  H.  Liu and  L. F.  Zhang: JOM, 70 (2018), 2157. https://doi.org/10.1007/s11837-018-3049-z
• 12)   X. M.  Li,  L.  Liu,  J. X.  Zhao,  X. M.  Li,  W. Z.  Ding,  F.  Chen and  J.  Chen: Met. World, 5 (2015), 56. https://doi.org/10.3969/j.issn.1000-6826.2015.05.15
• 13)   Z. B.  Li: Foundry, 53 (2004), 855. https://doi.org/10.3321/j.issn:1001-4977.2004.11.001
• 14)   Z. B.  Li and  J.  Fu: Spec. Steel, 20 (1999), 7. https://doi.org/10.3969/j.issn.1003-8620.1999.02.002
• 15)   K.  Ogino: Bull. Jpn. Inst. Met., 18 (1979), 684. https://doi.org/10.2320/materia1962.18.684
• 16)   J. W.  Cheng,  F. F.  Chen,  S. X.  Dai,  Y. S.  Xu and  R. Q.  Tan: J. Chin. Ceram. Soc., 41 (2013), 475 (in Chinese).
• 17)   A. X.  Lu and  J. W.  Huang: J. Cent. South Univ. Technol., 32 (2001), 398. https://doi.org/10.3969/j.issn.1672-7207.2001.04.018
• 18)   X. Y.  Xu,  S. X.  Gu,  S. Q.  Xia and  A. N.  Gu: J. Ceram., 36 (2015), 594. https://doi.org/10.13957/j.cnki.tcxb.2015.06.005
• 19)   S.  Ren,  J. L.  Zhang,  L. S.  Wu,  W. J.  Liu,  Y. N.  Bai,  X. D.  Xing,  B. X.  Su and  D. W.  Kong: ISIJ Int., 52 (2012), 984. https://doi.org/10.2355/isijinternational.52.984
• 20)   X. H.  Huang,  J. L.  Liao,  K.  Zheng,  H. H.  Hu,  F. M.  Wang and  Z. T.  Zhang: Ironmaking Steelmaking, 41 (2014), 67. https://doi.org/10.1179/1743281213Y.0000000107
• 21)   Y. M.  Gao,  S. B.  Wang,  C.  Hong,  X. J.  Ma and  F.  Yang: Int. J. Miner. Metall. Mater., 21 (2014), 353. https://doi.org/10.1007/s12613-014-0916-7
• 22)   T.  Iida,  H.  Sakai,  Y.  Kita and  K.  Murakami: High Temp. Mater. Process., 19 (2000), 153. https://doi.org/10.1515/HTMP.2000.19.3-4.153
• 23)   Y.  Guo: Master’s thesis, Jiangsu University, (2016), 1 (in Chinese), https://d.wanfangdata.com.cn/thesis/D01001308, (accessed 2022-05-03).
• 24)   Z.  Wang,  Q. F.  Shu and  K. C.  Chou: Steel Res. Int., 84 (2013), 766. https://doi.org/10.1002/srin.201200256
• 25)   K. C.  Mills and  S.  Sridhar: Ironmaking Steelmaking, 26 (1999), 262. https://doi.org/10.1179/030192399677121
• 26)   X. L.  Tang,  M.  Guo,  X. D.  Wang,  Z. T.  Zhang and  M.  Zhang: J. Univ. Sci. Technol. Beijing, 32 (2010), 1542 (in Chinese). http://doi.org/10.13374/j.issn1001-053x.2010.12.005
• 27)   K. C.  Mills,  A. B.  Fox,  Z.  Li and  R. P.  Thackray: Ironmaking Steelmaking, 32 (2005), 26. https://doi.org/10.1179/174328105x15788
• 28)   C. B.  Shi,  J.  Li,  J. W.  Cho,  F.  Jiang and  I. H.  Jung: Metall. Mater. Trans. B, 46 (2015), 2110. https://doi.org/10.1007/s11663-015-0402-2
• 29)   Q. X.  Fan,  Q. J.  Yu and  J. X.  Wei: Bull. Chin. Ceram. Soc., 26 (2007), 905. https://doi.org/10.3969/j.issn.1001-1625.2007.05.013
• 30)   H. Y.  Fan,  G. N.  Wang and  L. L.  Hu: Solid State Sci., 11 (2009), 2065. https://doi.org/10.1016/j.solidstatesciences.2009.09.007
• 31)   H.  Li,  P.  Hrma,  J. D.  Vienna,  M. X.  Qian,  Y. L.  Su and  D. E.  Smith: J. Non-Cryst. Solids, 331 (2003), 202. https://doi.org/10.1016/j.jnoncrysol.2003.08.082
• 32)   Q. H.  Li,  J. T.  Gao,  Y. L.  Zhang, Z. Q. An and Z. C. Guo: Metall. Mater. Trans. B, 48 (2017), 346. https://doi.org/10.1007/s11663-016-0858-8
• 33)   J.  Qi,  C. J.  Liu and  M. F.  Jiang: J. Non-Cryst. Solids, 475 (2017), 101. https://doi.org/10.1016/j.jnoncrysol.2017.09.014