Distribution Characteristics and Thermal Stability of Primary Carbide in Cast Ce-H13 Steel

Abstract

The primary carbide precipitated during the solidification process will act as the crack source to reduce the performance of H13 steel. It is necessary to obtain the nature of the primary carbide in H13 steel to reduce its detriment. Therefore, the distribution characteristics and thermal stability of the primary carbide in cast Ce-H13 steel were analyzed in this paper. There is a huge difference in the shape of the primary carbide between the 2D observation and the 3D observation. The shape of the primary carbide is a dendritic structure, and the branch is rich-V carbide and the trunk is rich-Ti-V carbide. The primary carbide size in the 3D observation increases gradually from the margin of the Ce-H13 ingot to the center. The rapid growth of the branch leads to an increase in size, and the decrease in the cooling rate is the main reason for the increase in size. When the heating temperature is 1150°C, the rich-V carbide starts to dissolve and dissolved completely at 1250°C. However, the rich-Ti-V carbide just starts to dissolve when the heating temperature is 1250°C. The number density and size of primary carbide decrease gradually with the increase of the heating temperature. Elemental Ce can effectively decrease the size of the primary carbide, but not for the number density. The calculated results are in keeping with the experimental observations. High-temperature heating can effectively reduce the primary carbide size, but cannot eliminate it.

1. Introduction

AISI H13 hot die steel is widely used in hot rolling, hot extrusion, and hot forging because of remarkable fatigue resistance performance, notable hot strength and impact toughness.¹⁾ In addition, H13 steel contains approximately 8 wt.% carbide forming elements, such as V, Mo, Cr and/or Ti, Nb. Therefore, it is easy to find the primary carbide in H13 steel.^2,3,4,5) Enrichment of the carbide forming elements at the last-to-solidify region during the solidification leads to the formation of the primary carbide.⁶⁾ It is well known that the large primary carbides will act as the crack sources to reduce the service life of H13 steel.^7,8)

Lots of work has been done to reduce the detriment of the primary carbide in H13 steel. The number density, size, amount and mean area of primary carbides decrease significantly with the increased cooling rate, but the shape of the primary carbide is insensitive to cooling rate.⁹⁾ The addition of elemental Mg in H13 steel can effectively modify Al₂O₃ into small dispersed MgAl₂O₄ or MgO, and then the dispersed oxides will act as the nucleation cores of the primary carbides to reduce their size.^10,11,12) Furthermore, the oxide nucleation core in H13 steel can be modified into Ce–O or Ce–O–S after the addition of elemental Ce. The Ce–O and Ce–O–S cannot act as the nucleation core of the primary carbide, suppressing the formation of primary carbide and further reducing its number density.¹³⁾ It has been widely confirmed in the laboratory that the appropriate addition of rare earth content can significantly improve the impact toughness and toughness of materials.^14,15,16,17)

Combined with the actual production process of H13 steel, it can be known that the H13 ingot will undergo high-temperature heating, forging and heat-treatment process. Therefore, dissolved by high temperature would be a more direct and effective way to reduce the detriment of the primary carbide. However, sometimes the large primary carbide still can be found in the final H13 steel product,¹⁸⁾ and long-term high-temperature heating leads to coarse austenite grain size. Therefore, it is very important to obtain the distribution characteristics and thermal stability of primary carbide in cast Ce-H13 steel, and few people are involved in this.

In this paper, the distribution characteristics and thermal stability of primary carbides at different heating temperatures in Ce-H13 steel were clarified in detail by comparing the morphology, composition, and size of the primary carbides in the 3D observation. It is hoped that this research work can provide a guiding role for the effective control of the primary carbides in Ce-H13 steel.

2. Experimental Section

Three experimental cast H13 ingots with a weight of approximately 2 kg were manufactured in a vacuum induction furnace. Pure iron (99.99%) was first melted in a MgO crucible, and then Si (99.5%), Mn (99.25%), Cr (99.6%), Mo (99.99%), and V–Fe (51.84% V and 46.56% Fe) alloy were added into the MgO crucible. The temperature was kept at 1600°C for 10 min to melt the alloy and homogenize the composition of the molten steel. Subsequently, different amounts of pure elemental Ce were added into the molten steel. After keeping the temperature at 1600°C for 5 min, the molten steel was cast into a cast-iron model, and then the H13 ingot was stripped and air-cooled to room temperature. The chemical composition of the H13 steel ingots were identified at the National Analysis Center for Iron and Steel, according to the national standards of China. The results are shown in Table 1.

Table 1. Chemical composition of samples, wt.%.

Samples	C	Si	Mn	Cr	Mo	V	Ti	S	N	O	Ce
0Ce	0.38	1	0.56	4.98	1.32	0.99	0.011	0.008	0.004	0.001	0
0.0055Ce	0.4	1	0.56	4.93	1.33	1	0.011	0.005	0.005	0.0008	0.0055
0.037Ce	0.4	0.97	0.53	4.99	1.3	0.98	0.011	0.001	0.005	0.0004	0.037

The solidification microstructure of the H13 steel ingot was obtained after being corroded for 5 min by a hydrochloric acid-water solution with a volume ratio of 1:1 at 70–80°C. Nonmetallic inclusions were partially extracted from the steel sample using a non-aqueous electrolysis method, and then their 3D-morphology was observed. The electrolyte was composed of 1% tetramethylammonium chloride, 10% acetylacetone, and 89% methanol (value fraction). The 3D-morphology and chemical composition of the primary carbide were analyzed through scanning electron microscopy (SEM) equipped with an energy dispersive spectroscopy (EDS). The extracted area for the 3D observation of primary carbide was 10 mm*5 mm, and 30 fields at 1000 times were observed to count the number density and size of the primary carbide in the 3D observation.

Three samples with size at 10*10*5 mm were taken at the centerline of an anatomical H13 steel ingot and separately sealed in the Φ18 mm*88 mm silica tubes. The silica tube was filled with Ar gas to inhibit oxidation and decarburization of the sample during the heating process. The samples were austenited at 1150°C, 1200°C, and 1250°C for 1 hour and then quenched in cold water. The grain size of samples was obtained with an optical microscope after being etched with an alcohol solution containing 4% nitric acid (volume fraction). ImageJ software was used to measure the grain size with a mean linear intercept method. Finally, the Factsage 7.3 software was used to calculate the precipitation mechanism and thermal stability of the primary carbide.

3. Results

3.1. Distribution Characteristics of Primary Carbide

At present, most researchers analyze the characteristics of the primary carbide in two-dimensional (2D) observation, which is not comprehensive enough. Figure 1 shows the schematic diagram of the difference between the 2D and 3D observation. The shape of the primary carbide can be classified into trunk-carbide (TC) and branch-carbide (BC). According to our previous results,⁶⁾ the trunk-carbide is rich-Ti-V carbide and branch-carbide is rich-V carbide. The rich-Ti-V carbide precipitates first and then acts as the nucleation core of the rich-V carbide. In the 2D observation, we can only observe a part of the TC or/and BC. Therefore, on the one hand, the shape of the primary carbide is considered to be a dot or strip. On the other hand, the statistical results of the number density of primary carbide will be too large, and the statistical results of the size will be too small. There is a large deviation between the statistical results and the fact in the 2D observation.

Fig. 1.

Schematic diagram of the difference between the 2D and 3D observation. (Online version in color.)

Therefore, the primary carbide size in the 3D observation from the margin of the 0.037Ce H13 ingot to the center was counted to authentically show the distribution of the primary carbide, and the results are shown in Fig. 2. The error bars indicate the standard deviation. As the distance from the margin increases, the average carbide size in the 3D observation increases from 8 μm to 26 μm. The size of the primary carbide exhibits a good symmetric distribution along the centerline. When the distances are 2.5 mm and 5 mm, the primary carbide is located in the columnar crystal region. The columnar crystal region has a faster cooling rate and a dense matrix structure, which does not provide sufficient time and space for the growth of the primary carbide. When the distances are 10 mm and 15 mm, the primary carbide is located in the equiaxed crystal region. Since the cooling rate is small and the alloying elements (Cr, Mo, V, C) are enriched in the equiaxed crystal region, it is possible to provide better growth kinetic conditions for the primary carbide.

Fig. 2.

Variation of the carbide size in 3D observation with the distance from the margin of the 0.037Ce ingot to the center. (Online version in color.)

Figure 3 shows the 3D morphology of the primary carbides at different distances from the margin to the center. The primary carbides are all located at the grain boundary, and the shapes are all dendritic structure. From the margin to the center, the rapid growth of BC leads to the significant increase of the primary carbide size. The service life of H13 steel will be significantly reduced because of the largest primary carbide size in the center of H13 ingot. The center of the H13 ingot is the most difficult part to heat in the actual production process because of its large size. Therefore, we focus on the effect of heating temperature on the primary carbide in the center of the Ce-H13 ingot.

Fig. 3.

3D morphology of primary carbide at different distances (2.5 mm, 5 mm, 10 mm, 15 mm) from the margin to the center.

3.2. Grain Size in Samples

The grain morphologies of the samples at different heating temperatures are shown in Fig. 4, and the variation of grain size with heating temperature is shown in Fig. 5. The error bars indicate the standard deviation. The grain size in the 0Ce, 0.0055Ce, and 0.037Ce sample increases gradually with the heating temperature increased from 1150°C to 1250°C. However, when the heating temperature changed from 1200°C to 1250°C, the grain size in all samples increases significantly. By comparing the grain size in the samples with different Ce content, it is well known that the grain size in the 0Ce sample is the smallest regardless of the heating temperatures. Moreover, we found some primary carbides at the grain boundary in the 0Ce sample when the heating temperatures are 1150°C and 1200°C, as shown in Fig. 1. Primary carbide hinders the growth of grain boundary during the heating process, which may be the reason for the smallest grain size in the 0Ce sample.

Fig. 4.

Grain morphology of samples at different heating temperatures. (Online version in color.)

Fig. 5.

Variation of grain size with heating temperatures in samples. (Online version in color.)

3.3. Carbide Morphology in Samples

Figure 6 shows the 3D morphology of the primary carbide in the samples at different heating temperatures. The shape of the primary carbide in the 0Ce sample is similar to a tree. The coarse primary carbide is easily aggregated, and the size reaches approximately 200 μm. However, as the Ce content in cast H13 steel increases, the shape of the primary carbide changes from dendritic to flake-like. Moreover, the primary carbide size also decreases significantly. When the heating temperature reaches 1150°C, significant dissolution occurred at the edges of the dendritic primary carbide in both 0Ce sample and 0.0055Ce sample. When the heating temperature increases to 1200°C, the rich-V carbide mostly dissolved, and the aggregation of primary carbides is interrupted, and the size remarkably reduced. When the heating temperature is 1250°C, the rich-V carbide completely dissolved. We also found that part of the rich-Ti-V carbide starts to dissolve. However, the morphology of the primary carbide in the 0.037Ce sample is unchanged when the heating temperatures are 1150°C and 1200°C. When the heating temperature reaches 1250°C, the rich-Ti-V carbide in 0.037Ce sample also starts to dissolve. Elemental Ce has little influence on the dissolution of primary carbide during the heating process, but can effectively improve the initial morphology of the primary carbide.

Fig. 6.

3D morphology of primary carbide in samples at different heating temperatures. (Online version in color.)

3.4. Variation of Primary Carbide Composition with Heating Temperature

We further analyzed the variation of the primary carbide composition with heating temperature to verify the conclusions in Fig. 6, and the results are shown in Fig. 7 (0.0055Ce sample) and Fig. 8 (0.037Ce sample). The results are the same for both 0.0055Ce sample and 0.037Ce sample. The outline of the primary carbide indicates its distribution area, including the TC and BC. When the heating temperature is 1150°C, we found that the Ti element is only enriched at TC, and the elemental Mo and V are enriched at both BC and TC. In other words, the Ti element is enriched at the core of the outline, and the elemental Mo and V are full of the outline of the primary carbide. However, when the heating temperature is 1250°C, the elemental Ti, V and Mo are enriched in the all outline of the primary carbide. According to above research results, the rich-V carbide dissolves first with the increasing heating temperature and then dissolves completely when the heating temperature reaches 1250°C. Therefore, we can only find the Ti element enriched at the core of the outline when the rich-V carbide starts to dissolve. The Ti element is full of the outline of the primary carbide when the rich-V carbide dissolves completely. The variation of primary carbide composition with temperature agrees well with the results in Fig. 6. The high thermal stability of rich-Ti-V carbide in cast H13 steel will significantly reduce the service life of H13 steel.

Fig. 7.

Elements mapping of primary carbide at different heating temperatures in 0.0055Ce sample. (Online version in color.)

Fig. 8.

Elements mapping of primary carbide at different heating temperatures in 0.037Ce sample. (Online version in color.)

3.5. Characterization of Primary Carbide in Samples

The number and size of the primary carbide in the 3D observation are shown in Fig. 9. Since the three samples have the same observation area, the variation of the number of the primary carbides can be used to illustrate the variation of the number density. At the same heating temperature, the number densities of the primary carbide in three samples are substantially the same. As the heating temperature increases, the number density of the primary carbide gradually decreases. The decrease in the number density is mainly due to the dissolution of small primary carbide during the heating process, which agrees well with the experimental observations.

Fig. 9.

Number and size of the primary carbide in samples. (Online version in color.)

The variation of heating temperature with the average size of primary carbide is shown in Fig. 10. The error bars indicate the standard deviation. At the same heating temperature, the average size of the primary carbide decreases with the increasing Ce content. As the heating temperature increases, the average size gradually decreases. In particular, when the heating temperature increases to 1250°C, the average size of the primary carbide decreases significantly. Therefore, the high heating temperature and the addition of Ce element can effectively decrease the size of the primary carbide. However, the excessive heating temperature would significantly deteriorate the austenite grain size and heating equipment. The heating temperature should be reasonably selected. The addition of Ce element in cast H13 steel will be a better choice in controlling the characterization of the primary carbide.

Fig. 10.

Variation of temperature with the average size of the primary carbide in samples at the 3D observation. (Online version in color.)

4. Discussion

4.1. Growth Behavior of Primary Carbide during Solidification

It is considered that the primary carbide size changes with the distance from the margin to the center, as described above. The cooling rate at the center is lower than the margin, and the cooling rate has a larger influence on the growth of primary carbide.^19,20) Therefore, the primary carbide size at different distances from the margin to the center was calculated, and the calculation method used in this model is presented in the following.

The alloy elemental concentration in the liquid phase during solidification can be speculated theoretically by the segregation equation. Ohnaka’s²¹⁾ equations are used to calculate the segregation of elements in molten steel during solidification for different cooling rates:   

$$C_L^{\text{i}}=C_0^{\text{i}}\cdot {(1-(1-\beta \cdot k_i/(1+\beta ))\cdot f_s)}^{(k_i-1)/\{1-\beta \cdot k_i/(1+\beta )\}}$$(1)

  

$$\beta =4\cdot D{}_S^i\cdot t_f/{\lambda }_2{}^2$$(2)

Where, $C_L^{\text{i}}$ is the concentration (mass percentage) of the solute i in the liquid phase, $C_0^{\text{i}}$ is the initial concentration (mass percentage) of the solute i, k_i is the partition coefficient of solute i, f_s is the solid phase fraction. $D_S^i$ is the solute diffusivity of solute i in the solid phase, m²/s, λ₂ is the secondary dendrite arm spacing, m; t_f is the local solidification time, s, which can be calculated by Eq. (3).   

$$t_f=\frac{T_L-T_S}{R_C}$$(3)

Where T_L and T_S is the liquidus temperature and solid temperature, K, obtained from the calculation results of Factsage 7.2 software. R_C is the local cooling rate, K/s. According to Mao’s results,¹⁰⁾ the variation of R_C with λ₂ can be obtained by Eq. (4).   

$${\lambda }_2=175.4\cdot R_C{}^{-0.322}$$(4)

According to the calculated results by Factsage 7.2, as shown in section 4.2, the rich-Ti-V carbide (the main elements are Ti and N) precipitates first. Therefore, the following equilibrium²²⁾ was assumed to exist on the interface of the precipitated primary carbide TiN and the molten steel.   

$${(\mathrm{T}\mathrm{i}\mathrm{N})}_{\mathrm{s}}=[\mathrm{T}\mathrm{i}]+[\mathrm{N}]$$(5)

  

$$l\text{o}\mathrm{g}{f}_{[\mathrm{T}\mathrm{i}]}\cdot [\mathrm{T}\mathrm{i}\%]\cdot f_{[\mathrm{N}]}\cdot [\mathrm{N}\%]/a_{(\mathrm{T}\mathrm{i}\mathrm{N})}=-16\mathrm{\hspace{0.17em} }580/T+5.9$$(6)

Where, a_i is the activity, and f_i is the activity coefficient, which can be calculated by Wagner’s equations. The temperature of the solidification interface, T, can be calculated as follows:²³⁾   

$$T=T_0-{\displaystyle \frac{T_0-T_L}{1-f_s\cdot {\displaystyle \frac{T_L-T_S}{T_0-T_S}}}}$$(7)

Since the primary carbide precipitates at the end of the solidification, the interfacial reaction is comparatively fast. The Ti content at the interface is relatively higher than N content in the present conditions. Therefore, the mass transport of N from the molten steel to the reaction interface is assumed as the controlling step during the growth of the primary carbide. The growth of primary carbide is expressed by Eq. (8).   

$$r\cdot \frac{dr}{dt}=\frac{M_{TiN}}{100\cdot M_{Fe}}\cdot \frac{{\rho }_{Fe}}{{\rho }_{TiN}}\cdot D_L^N\cdot (C_L^N-C_{eq}^N)$$(8)

Where, r is the primary carbide radius, m, $D_L^N$ is the diffusion coefficient of solute N in molten steel, m²/s, M is the molecular weight, ρ is the density, kg/m³, $C_{eq}^N$ is the equilibrium content of N calculated by Eq. (6). Extended data needed for this model were obtained from Meng’s work,²⁴⁾ the extended and experimental data are shown in Table 2.

Table 2. Key parameters of the growth model.

distance	λ2		ki	$D_S^i$
5 mm	25.4 μm	C	0.34	0.0761* exp (−32160/R*T )
10 mm	32.8 μm	N	0.48	0.91 exp (−40270/R*T)
15 mm	39.2 μm	Cr	0.86	0.0012* exp (−52340/R*T)
20 mm	45.0 μm	Mo	0.585	0.068* exp (−59000/R*T)
		V	0.63	0.284* exp (−61900/R*T)
$D_L^N$	3.25*10^(−7)* exp (−11500/(R*T))	Ti	0.33	0.15* exp (−59980/R*T)

Note: R is the gas content of 8.314 J*(mol*K)⁻¹, and T is the temperature in Kelvin

Figure 11 shows the variation of the calculated radius of the primary carbide with the solid fraction at different distances. By comparing the experimental data in Fig. 2 with the calculated results in Fig. 11, the calculated primary carbide size is approximate twice the statistical average size, which is generally consistent with the statistical maximum size. On the one hand, the consumption of elemental Ti and N during the growth of TiN is not taken into consideration in the calculation model. On the other hand, since the primary carbide precipitates at the end of the solidification, the growth of which may be restricted by the Ce element. Hence, the calculated primary carbide size is slightly larger than the experimental primary carbide size. However, the calculation results can still effectively prove the experimental results in Fig. 2. The decrease in the cooling rate is the main reason for the gradual increase in the size of the primary carbide from the edge to the center.

Fig. 11.

Variation of the solid fraction with primary carbide at different distances. (Online version in color.)

4.2. Thermodynamic Calculation of Thermal Stability of Primary Carbide

Taking the chemical composition of the 0Ce sample in Table 1 as an example, the equilibrium solidification was calculated with Factsage 7.3 thermodynamic software with the FToxid and FSstel databases. The total system was set as 100 g. The calculated result is shown in Fig. 12. The Al–Ti–O phase precipitates before the solidification, and the MnS phase precipitates at the end of solidification. The Al–Ti–O and MnS inclusion were found in the experimental observations of the 0Ce sample. However, they were disappeared completely in the 0.037Ce sample because of the decrease in O and S content, as shown in Table 1.

Fig. 12.

Equilibrium calculation of 0Ce sample. (Online version in color.)

The FCC#2 phase is the carbide. The FCC#2 phase precipitates during the solidification, and the precipitation temperature is approximately 1365°C. Therefore, it was very difficult to completely dissolve primary carbide at 1250°C. Figure 13 shows the mass fraction of the main elements in the FCC#2 phase. The Ce content has little influence on the precipitation temperature and precipitation sequence of primary carbides in cast H13 steel. The rich-Ti-V carbide precipitates first and then followed by the rich-V carbide. As shown in Figs. 7 and 8, when the heating temperature is 1150°C, the rich-V carbide is just beginning to dissolve. When the heating temperature is 1250°C, the rich-V carbide completely dissolved, and the rich-Ti-V carbide is still stable. The calculated results are in keeping with the experimental observations.

Fig. 13.

Mass fraction of main elements in FCC#2 phase in 0Ce sample (a), 0.0055Ce sample (b), and 0.037Ce sample (c). (Online version in color.)

The variation of morphology and composition of the primary carbide with heating temperature is shown in Fig. 14. As the heating temperature increases, the shape of primary carbide in H13 steel changes from dendritic to flake, and the size and number density decreases significantly. The rich-V carbide can be dissolved into the matrix during the heating process, but not for rich-Ti-V carbide.

Fig. 14.

Variation of morphology and composition of primary carbide with heating temperature. (Online version in color.)

5. Conclusions

The distribution characteristics and thermal stability of primary carbides in Ce-H13 steel are obtained by comparing the morphology, composition, and size of primary carbides at different heating temperatures and different locations from the margin to the center. The study’s conclusions are the following.

(1) There is a large difference in the primary carbide morphology between the 2D observation and the 3D observation. The primary carbide size in 3D observation increases gradually from the margin of the Ce-H13 steel to the center.

(2) As the heating temperature increases, the primary carbide in cast H13 steel can dissolve into the matrix. The rich-V carbide at BC starts to dissolve at 1150°C and disappears completely at 1250°C. When the heating temperature reaches 1250°C, the rich-Ti-V carbide in TC is just beginning to dissolve.

(3) The number density and size of primary carbide in the 3D observation decrease gradually with the heating temperature increases. The Ce element can effectively decrease the size of primary carbide but has little influence on the number density. The calculated results agree well with the experimental observations.

Acknowledgements

The authors are grateful for support from the National Natural Science Foundation of China (NO. 51874034).

References

• 1)   S. H.  Chang,  T. P.  Tang and  K. T.  Huang: ISIJ Int., 50 (2010), 569.
• 2)   Y.  Xie,  G. G.  Cheng,  L.  Chen,  Y. D.  Zhang and  Q. Z.  Yan: ISIJ Int., 56 (2016), 995.
• 3)   C. B.  Shi,  X. C.  Chen,  H. J.  Guo,  Z. J.  Zhu and  X. L.  Sun: Metall. Mater. Trans. B, 44 (2013), 378.
• 4)   M.  Wang,  D. S.  Ma,  Z. T.  Liu,  J.  Zhou,  H. X.  Chi and  J. Q.  Dai: Acta Metall. Sin., 50 (2014), 285.
• 5)   A.  Dadoo,  S. M. A.  Boutorabi and  S.  Kheirandish: Opt. Laser Technol., 112 (2019), 236.
• 6)   Y.  Xie,  G. G.  Cheng,  X. L.  Meng,  L.  Chen and  Y. D.  Zhang: ISIJ Int., 56 (2016), 1996.
• 7)   Y.  Huang,  G. G.  Cheng,  S. J.  Li and  W. X.  Dai: Steel Res. Int., 90 (2019), 1900035.
• 8)   Q. C.  Jiang,  H. L.  Sui and  Q. F.  Guan: ISIJ Int., 44 (2004), 1103.
• 9)   J. T.  Tchuindjang and  J.  Lecomte-Beckers: Int. J. Fatigue, 29 (2007), 713.
• 10)   M. T.  Mao,  H. J.  Guo,  F.  Wang and  X. L.  Sun: ISIJ Int., 59 (2019), 848.
• 11)   J.  Li,  J.  Li,  C. B.  Shi,  L. L.  Wang,  Z.  Wu and  H.  Wang: Can. Metall. Q., 55 (2016), 321.
• 12)   J.  Li,  J.  Li,  L. L.  Wang and  Q. T.  Zhu: Met. Sci. Heat Treat., 58 (2016), 330.
• 13)   H.  Wang,  J.  Li,  C. B.  Shi and  J.  Li: Ironmaking Steelmaking, 44 (2017), 128.
• 14)   Y.  Huang,  G. G.  Cheng,  S. J.  Li and  W. X.  Dai: Steel Res. Int., 89 (2018), 1800371.
• 15)   J.  Hufenbach,  A.  Helth,  M. H.  Lee,  H.  Wendrock,  L.  Giebeler,  C. Y.  Choe,  K. H.  Kim,  U.  Kühn,  T. S.  Kim and  J.  Eckert: Mater. Sci. Eng. A, 674 (2016), 366.
• 16)   Y.  Li,  G.  Yang,  Z. H.  Jiang,  C. Y.  Chen,  S.  Sun and  P. F.  Du: Ironmaking Steelmaking. https://doi.org/10.1080/03019233.2018.1538183
• 17)   K. P.  Liu,  X. L.  Dun,  J. P.  Lai and  H. S.  Liu: Mater. Sci. Eng. A, 528 (2011), 8263.
• 18)   H. L.  Liu,  C. J.  Liu and  M. F.  Jiang: Mater. Des., 33 (2012), 306.
• 19)   Y.  Xie,  G. G.  Cheng,  L.  Chen,  Y. D.  Zhang and  Q. Z.  Yan: Metall. Res. Technol., 113 (2016), 206.
• 20)   Y. N.  Wang,  J.  Yang,  X. L.  Xin,  R. Z.  Wang and  L. Y.  Xu: Metall. Mater. Trans. B, 47 (2016), 1378.
• 21)   I.  Ohnaka: Trans. Iron Steel Inst. Jpn., 26 (1986), 1045.
• 22)   J. X.  Chen: Common Charts and Databook for Steelmaking, The Metallurgical Industry Press, Beijing, (2010), 718.
• 23)   Z. T.  Ma and  D.  Janke: ISIJ Int., 38 (1998), 46.
• 24)   Y.  Meng and  B. G.  Thomas: Metall. Mater. Trans. B, 34 (2003), 685.