Three-Dimentional Investigations of Different Sulfide Inclusions in Steels by Using the Electrolytic Extraction Method

(1)

~ 1 ~

Three-Dimentional Investigations of

Different Sulfide Inclusions in Steels by

Using the Electrolytic Extraction Method

Rui Wang

KTH ROYAL INSTITUTE OF TECHNOLOGY

SCHOOL OF INDUSTRIAL ENGINEERING AND MANAGEMENT

DEGREE PROJECT IN MATERIALS SCIENCE AND ENGINEERING, SECOND CYCLE, 30 CREDITS

(2)

(3)

~ i ~

Three-Dimensional Investigations of Different

Sulfide Inclusions in Steels by Using Electrolytic

Extraction Method

Abstract

MnS inclusions in steel products are harmful to the mechanical properties of steel products. Thus, it is critical to take control of the formation and behavior of different MnS inclusions in various steel grades. Rare earth elements have a strong affinity to oxygen and sulfur in liquid steel, which provides a way to modify sulfide inclusions in liquid steel. However, conventional two dimensional (2D) evaluations of MnS inclusions on a polished surface of steel samples have significant disadvantages (especially by investigation of complex shape or deformed sulfides).

In the present study, heat treatments of specimens taken from an industrial ingot and a rolled 42CrMo4 steel were carried out to improve the accuracy of an evaluation of different sulfides by three dimensional (3D) investigations of inclusions after electrolytic extraction. The obtained results showed that the heat treatment at 900℃ during 5-30 minutes can significantly reduce the amount of iron carbides in steel specimens and increase the accuracy of the assessment of the inclusion characteristics after electrolytic extraction. Moreover, different types ofMnS inclusions did not visually changed their morphology during this heat treatment.

Laboratory melting experiments were also done in this study to evaluate the effect of REM additions on modifications of sulfide inclusions in aFe-0.4%Mn-0.015%S melt. The inclusions in metal samples taken from the melt after an addition of REM were observed by using 3D investigations after a completed electrolytic extraction. It was found that most of the observed REM-oxysulfide inclusions have a spherical shape and that their diameter varied in the range from 0.5 to 5.0 µm.

Key words: sulfide inclusions, rare earth metals, carbides, heat treatment, electrolytic

(4)

~ ii ~

1. Introduction

1.1 Background

As is known, non-metallic inclusions (NMI) in steel products have marked impacts on its quality and performance. Thus it is critically important to take control of inclusions in steel during the steelmaking process. Hard inclusions with sharp corners, such as TiN and Al2O3, will severely decrease the fatigue life of steel products; relatively soft

inclusions, especially the inclusions of MnS type, will greatly influence the mechanical properties of steel products. Moreover, such inclusion characteristics as the number, size, distribution in steel and other are also important factors influencing the quality and performance of steel products.

Currently, the observation method of inclusions applied broadly is a two-dimensional (2D) investigation of inclusions on a polished surface of steel samples and observation of inclusions on the section. However, this method has considerable disadvantages when it is used to observe inclusions of different shapes. This is especially true for those of an irregular shape and for elongated inclusions after deformation. The inclusion shapes observed with this method are merely the cross sectional shapes of the inclusions. The real morphology and dimension of the entire inclusions cannot be observed, which makes the inaccuracy by evaluation of inclusion characteristics. Therefore, three-dimension (3D) investigations of inclusions have gradually been adopted. This method is to observe residual non-metallic inclusions in full scale after electrolytic dissolution of steel samples, which accurately grasps the features of inclusions such as dimension and morphology. Nevertheless, when the carbon content in steel is high (usually >0.2% C), the excessive iron carbides therein may affect the accuracy of NMI observations.

The present study is focused on the steel grade 42CrMo4 having a carbon content>0.3%. The primary objective is to investigate an effect of different heat treatments on the elimination of iron carbides with saving of sulfides in medium carbon steels for the following 3D investigations of NMI after electrolytic extraction (EE). After appropriate heat treatment and EE, the carbides have no influence on the inclusion observation. In addition, a 3D investigation of inclusions is carried out to analyze the impacts of Ce in steel on the features of MnS inclusions such as the morphology, dimension and the like.

1.2 Invetigation of MnS

(6)

2

monophasic or recombine with oxides to form multiphasic oxysulfides. They are often irregularly distributed, sometimes among dendrites. Furthermore, they mostly exist in Al-free deoxidated steels with high oxygen contents. MnS in type II are aggregately distributed at the grain boundaries in chain-like or slender rod-like precipitations, which is commonly known as grain boundary sulfides. Sulfides of this type often exist in aluminum deoxidated steels without an excessive aluminum. Type IIIMnS are mostly massive with corner and irregularly distributed in steel. This type of MnS is similar to type I, but is usually monophasic and most commonly seen in excessive aluminum deoxidated steels.

Yoichi Ito et al. [2, 3] have done a lot of work in consummating Fe-Mn-S ternary phase diagrams and Fe-Mn-S-Cquaternary phase diagrams. Furthermore, they classified the shapes of sulfides in steel into four types when the sulfur content is 0.013%~0.063%: type I is spherical shape; type II was sector and chain-like shape; type III was a polyhedral shape; and type X was irregular shape. The morphologies of various inclusions are shown in Fig. 1 [4].

Fig. 1. Classification of sulfide shapes. (× 1/1): (a) type I (× 500), (b) type II (× 110), (c) type III (× 500), (d) type X (× 500).

(7)

3

sulfides, according to which the MnSinclucions precipitated after initial solidification of Fe in Fe-1%Mn-0.3%S alloy may be classified on the basis of shapes into: (1) spherical or drop-like MnS inclusions generated from the metastable monotectic reaction L1→Fe(s)+L2; (2) rod-like MnS inclusions generated from the eutectic

reaction L1→Fe(s)+MnS; (3) fishbone-like MnS inclusions generated from the

irregular eutectic reaction (seen in Figs. 2-4).

Fig. 2. Spherical monotecticMnS: (a) an optical micrograph (Fe-1%Mn-0.3%S-0.1%Ti), (b) a scanning electron micrograph

(Fe-1%Mn-0.3%S).

(8)

4

Fig. 4. Fishbone-like irregular eutectic MnS: (a) an optical micrograph (Fe-1%Mn-0.3%S-1%C), (b) a scanning electron micrograph

(Fe-1%Mn-0.3%S-5%Si).

Katsunari Oikawa also found out that in S-rich alloys (containing 1-1.3% S), theMnS inclusions were separated out as an initial crystalline phase. Because of the differences in added elements and atmosphere during smelting and solidification, the shapes of MnS inclusions mainly comprised of (1) a spherical shape, (2) a dendritic shape, and (3) a polyhedral shape (as shown in Figs. 5-7).

(9)

5

Fig. 6. Dendritic MnS: (a) an optical micrograph

(Fe-4%Mn-2%S-0.5%S-0.3%C), (b) a scanning electron micrograph (Fe-2.5%Mn-1.3%S-0.5%Al).

Fig. 7. Polyhedral MnS: (a) an optical micrograph (Fe-2.5%Mn-1.3%S-1%Al), (b) a scanning electron micrograph (Fe-2.5%Mn-1.3%S-0.5%Al-0.1%Ti).

(10)

6

Fig. 8. Typical morphology of different sulfides: (a) type I oxy-sulfide, (b) type II (+IV) eutectic MnS, (c) type III regular MnS[7-8].

To sum up, sulfides in as-cast steel tend to exhibit various shapes and distributions by reason of different generating conditions, according to which rough classifications can be done. In addition, the generating mechanism of sulfides with various shapes in different steels was discussed among researchers, but there is still controversy.

1.3 Comparison of 2D investigation and 3D investigation

In general, 2D investigations of non-metallic inclusions on a surface of polished metal samples by using a scanning electronic microscope (SEM) or an optical microscope are commonly used standard methods for the evaluation of NMI in different steel grades and alloys. However, the observation of the actual size and morphology of inclusions and clusters with this method is not valid [7-11] since only the size of a certain section of NMI can be measured on a polished metal surface. This disadvantage is especially evident with regard to observation of irregular inclusions, clusters and deformed inclusions (Fig. 9).

Fig. 9. Schematic illustrations of 2D and 3D investigations of different inclusions.

(11)

7

and, as a result, a lower accuracy of NMI observations. One of the possible ways to avoid this limitation for the EE method is to use a dissolution of iron carbides at α-γ iron transformation during heat treatment of solid steel specimens with following quenching.

1.4 MnS stability at heat treatment

McFarland et al.[13] carried out similar tests using steels with a 0.013% sulfur content. As for type II sulfide in steel, the heat treatment for different duration was applied at 925oC. The morphology changes of the sulfides is shown in Fig. 10.

Fig. 10. Impacts of homogenization time on sulfide morphology: (a), (b) and (c) are longitudinal samples, with isothermal time of 0, 5, 10h, respectively; (d),(e)

and (f) are corresponding lateral samples. Sulfur content was 0.013% and holding temperature was 925oC).

Murtyet al.[14] investigated the morphology changes of type III sulfides during heat treatment after rolling in AISI 3030 steel with an increased sulfur content up to a value of 0.1%. The heat treatment temperature was 1583K (1310oC), and the holding

(12)

8

times were 0, 10, 20, 50, 100 h. The sulfide morphology change is shown in Fig. 11.

Fig. 11. Impacts of homogenization time on sulfide morphology: (a)-(e) are longitudinal samples, with isothermal time of 0, 10, 20, 50, 100 h respectively; (f)-(j) are corresponding lateral samples. Sulfur content was 0.1% and holding

temperature was 1583K)[14].

(13)

9

increasing of holding time when the holding time was above 10h. The shapes changed from a tabular shape after rolling to a cylindrical or rod-like shape, and then fractured to a cotton-ball-like shape.Finally, the shape became spherical.

The results show that the time type III sulfides need for fracturing and spheroidizing was longer and the isothermal homogenization temperature was higher compared to type II sulfides. After the deformed samples were isothermally homogenized, MnS inclusions were spheroidized. This led to a significant enhancement of the steel ductility, reduction of area and impact value, especially the lateral performances. Thus, the steel products can get favorable comprehensive mechanical properties.

Moreover, Shao et al. [15] reported that aspheroidization of elongated MnS inclusions in resulfurized free-cutting steels (0.086% C, 0.01% Si, 1.02% Mn and 0.32% S) started at 800oC, as clearly shown in Fig. 12. It can be seen that the aspect ratio of MnS inclusions observed on polished steel surfaces (2D investigation) decreases by almost 2-3 times at 800oC in comparison to the results for steel samples before heat treatment.

Fig. 12. Aspect ratio of MnS inclusions on the surface of steel samples at different temperatures during heat treatment [15].

The effects of heat treatments using different holding temperatures and times are different with respect to various sulfides present in different steels. Therefore, it should be further studied how to choose proper holding temperatures and times according to the target steel grade, in order to ensure that the morphology change of MnS type inclusions can be deminished.

1.5 Modification of MnS by Ce addition

(14)

10

additions of elements having a higher affinity with sulfur compared to manganese. For instance, rare earth elements (such as La, Ce, Pr, etc.) have a strong affinity to harmful elements such as oxygen and sulfur in liquid steel. They partially form stable compounds in liquid steel after an addition. Their purification functions are also reflected in reducing the harmful effects of N, P and low-melting-point elements. The deoxidizing capacity of rare earth elements is superior to Mg, Al and Ti, and it is equivalent to the capacity of Ca; their desulfuring capacity is superior to Mg but inferior to Ca[16]. Acording to thermodynamic calculations, products of rare earth and O, S depend on the thermodynamic equilibrium results of each element. The sequence is: RexOy, RexOyS, ReSy, RexNy and RexCy. As for Al deoxidized steels, rare

earth elements like Ce can react with the deoxidation product Al2O3 according to the

following reaction (1).

2Ce+Al2O3=Ce2O3+2Al (1)

It is one of the functions of rare earth elements in steel to control the morphology of inclusions. Irregular flocculent Al2O3-SiO2inclusions will be generated after an Al

deoxidation of liquid steel. These can be transformed into spherical rare earth complex inclusions, SiO2-ReAlO3, after the addition of rare earth. Thereafter, the

aggregated Al2O3 will fade away; S in primary liquid steel may segregate with Mn at

the grain boundary during solidification at the decreased temperature, and form MnS inclusions[16]. After adding an appropriate amount of rare earth elements, Re2O2S

can be formed, which may completely or partially substitute MnS inclusions. As MnS inclusions have a relatively low melting point, they may easily be deformed when being hot worked. This causesvariations in the steel products properties and especially the high temperature properties. The Re2O2S inclusions formed by an addition of

REM has a high melting point. Also, since its large particles have emerged from liquid steel and the small particles dispersed in liquid steel as spheres or points, the properties won’t be instable during hot working because of the imparity of the thermal expansion coefficients [17].

1.6 Purpose of study

The goals of this study can be formulated as follows:

1. FeC3 in steel has a strong impact on the observation of inclusions when using the

3D investigation method. Hence it is desirable that the FeC3 in steel would be reduced

by heat treatment so that the impact on the observationscould be decreased.

(15)

11

2. Experimental

2.1 Samples from an industrial heat

The steel grade used in the present study was a42CrMo4 grade, of which the major compositions are shown in Table 1. Ingot and rolled samples were taken from industrial heat to be used in the heat treatment and melting experiments.

Table 1. Composition of tested 42CrMo4 steel

C (wt%) Si (wt%) Mn (wt%) Cr (wt%) S (wt%) O (wt%) 0.42 0.20% 0.80% 1.0% 0.025% <10ppm

2.2 Heat treatment of steel specimens

The ingot samples and rolled samples were sliced into specimens (~15x10 mm) with a thickness of 4~5mm. Then, the specimens were set erectly in the furnace at 900oC, as can be seen in Fig. 13.

Fig. 13. Schematic illustration of heat treatment setup.

The holding times of these specimens at this temperature were 5, 10, 15 and 30 min. The temperature variation, sampling time and the corresponding label are shown in

Table 2. After a completed holding, each specimen was quenched in water

(16)

12

Table 2. Sampling time and the label of specimens in the heat treatment experiments

Holding time in furnace at 900oC Ingot sample Rolled sample 5 min IM-900-1 IMA-900-1 10 min IM-900-2 IMA-900-2 15 min IM-900-3 IMA-900-3 30 min IM-900-4 IMA-900-4

2.3 Melting experiments with modification of sulfides by REM addition

A vertical induction furnace was used in the present melting experiments. The schematic illustration of the furnace is shown in Fig. 14. The melting temperature was 1600oC. High purity Al2O3 crucibles were used in two melting experiments. During

the experiments,Ar gas with a pressure of 0.5mbar was injected into the furnace volume to protect the liquid steel from oxidation by oxygen in the air atmosphere. A thermocouple was placed between the crucible and the furnace wall to monitor the temperature.

Fig. 14. Schematic illustration of the melting experiments.

(17)

13

During the experiments, samples were taken from the melt by suction using quartz tube (inside Ø 6 mm) 1 min before REM addition (sample QT0) and after 1, 3, 5 and 10 min of REM addition (samples QT1, QT3, QT5 and QT10). After sampling of liquid metal, the quartz tube samples were quenched in the water immediately. Then, the melt in the crucible was cooled in the furnace at a rate of ~2°C/min to 1400°C (sample IS).

In order to investigatethe precipitated non-metallic inclusions (NMI), a central section of each QT sample (Ø6x20 mm) was prepared for an electrolytic extraction study.

2.4 Electrolytic extractions and investigation of inclusions

In the present study, the Electrolytic Extraction (EE) method was used for 3D investigations of NMI in steel samples. In this method, the steel samples are firstly grinded with abrasive paper, placed in electrolyte and extracted with a total electric charge of 500C. The schematic illustration of the EE setup is shown in Fig. 15. During electrolytic extraction process, the metal matrix of specimen was dissolved while the non-metallic inclusions, which are more stable, did not dissolve in the electrolyte. 0.1g of metal was filtratedfrom the electrolyte after a completed dissolution. Then, the NMI on a surface of the film filter with an open pore size of 0.4 µm were observed by using the scanning electron microscope (SEM) combined with an electron probe microanalysis (EPMA).Specifically, the features of the inclusions such as size, composition and morphology were determined.

The electrolyte used for dissolution of specimens from the heat treatment experiments was 10% AA (10% acetyl acetone-1% tetramethylammonium chloride-methanol); the electrolyte used for the Ce-modified samples from the melting experiments was 2%TEA (2 v/v% triethanol amine – 1 w/v% tetramethylammonium chloride - methanol).In addition, the following electric parameters were used for EE process: current – 40–60 mA and voltage – 3.6–3.8 V.

(18)

14

3. Results and discussion

3.1 3D investigations of inclusions in steel samples before and after heat treatment

Typical non-metallic inclusions, extracted from the steel samples before and after heat treatment, were investigated in 3D by using SEM. Typical SEM images of non-metallic inclusions on film filter before (a) and after heat treatment of the steel specimens for 5 min (b) and 30 min (c) are shown in Fig. 16. As can be seen in Fig. 16a, more than 95% of the filter area were covered by carbides, which hindered valid observations of most inclusions. After 5~30 min of holding a sample at 900oC during heat treatment, the area covered by iron carbides decreased by upto ~5-10% (Figs. 16b and 16c). As a result, most of the inclusions could be observed and evaluated on the film filters.

(a) (b) (c)

Fig. 16. Typical SEM images of non-metallic inclusions on film filters before (a) and after heat treatment of steel specimens for 5 min (b) and 30 min (c). 3.2 MnS inclusions in 3D investigation

(19)

15

Table 3. Features of different typical MnS inclusions found in the 3D investigation on samples taken after hetat treatment of steel specimens from

ingot.

Type of MnS inclusions Size range (μm) Composition

5-10 MnS+ (Al,Mg,Ca)O 8-15 MnS 10-25 MnS, MnS+ (Al,Mg,Ca)O

Table 4. Features of different typical MnS inclusions found in the 3D investigation on samples taken after hetat treatment of rolled steel specimens.

Type of MnS inclusions Size range (μm) Composition

(20)

16

80-250 MnS+ (Al,Ca)O

3.3 Modification of sulfide inclusions by Ce additions

Some typical SEM images of non-metallic inclusions on a surface of film filter (3D investigation) are shown in Fig. 17. As can be seen, the formed non-metallic inclusions have a globular shape. The size (diameter) of most observed inclusions varies in the range from 0.5 to 5.0 µm.

Fig. 17. Typical SEM images of non-metallic inclusions on a surface of film filter after electrolytic extraction of the QT1 sample (Exp. 1).

The main expected reactions in the melt after addition of REM are the following: 2REM +3O → REM 2O3 (2)

xREM + yS→ REMxSy (3)

2REM + 2O + S → REM 2O2S (4)

However, it was found that most of the spherical inclusions observed in 3D after an electrolytic extraction of the metal samples of the experimental melts (Fe-0.4%Mn-0.015%S) also contained Si, Mn, Al and Fe elements, as shown in Table

5. Therefore, the REM added in the melt can react with primary oxide inclusions

(21)

17

Table 5. Distribution of main elements in observed non-metallic inclusions (QT1 sample, Exp. 1)

Compositions of investigated non-metallic inclusions were analyzed quantitative by using EPMA. Typical inclusions on film filter and typical EPMA Spectrum obtained by determination of their composition are shown in Fig. 18. In addition, the composition analysis of the investigated inclusions after electrolytic extraction is given in Table 6. 1 2 3 4 5 6 7 8 9 10 keV 0 5 10 15 20 25 30 35 40 45 50 cps/eV C O F Al Si Mn Mn La La La Ce Ce Ce S

(22)

18

Table 6. Contents of main elements in observed non-metallic inclusions (in mass%)

O Si Ce La Al Mn S

Average* 42.7±5.6 12.0±1.7 13.7±4.5 8.0±3.4 1.0±1.1 22.0±4.2 0.6±0.3 Min/Max 35.9/54.1 7.7/14.1 3.5/19.4 2.0/16.1 0.0/4.1 13.1/28.5 0.2/1.3 *: average value ± standard deviation of results.

A comparison of contents of main elements (such as O, S, Al, Si, Mn, Ce and La) in non-metallic inclusions having different sizes (QT1 sample, Exp. 1) is shown in Fig.

19.

It can be seen that the formed complex REM oxy-sulfides contain some amount of S (0.2-1.3%). It means that a smallest amount of S will segregate and precipitate as large size MnSinclusions in the central zone of the ingots.

(23)

19 0 5 10 15 20 0 1 2 3 4 5 La, mass% Ce, mass% Conten t of e lemen t (ma ss %) Size of inclusion (µm) Ce La RW1-QT1-1 0 10 20 30 0 0,5 1 1,5 0 1 2 3 4 5 Mn, mass% S, mass% RW1-QT1-1 Conten ts of M n (ma ss%) Size of inclusion (µm) Conten ts of S (ma ss%) S Mn

(24)

20

4. Conclusions

Based on obtained results in this study, the following conclusions can be made: 1. Heat treatment and electrolytic extraction of specimens of 42CrMo4 steel grade:

 Heat treatment of metal samples at 900o

C during 5-30 minutes helps to dissolve the iron carbides in the 42CrMo4 steel. This promotes the application of the electrolytic extraction technique for three dimensional investigations of different non-metallic inclusions in the given steel grade.  A 5 minutes holding time of the steel specimens at 900o

C is enough for getting rid of most FeC3 in samples taken from ingots and from rolled steel

products.

 Different types ofMnS inclusions did not dissolve visibly within the heat treatment conditions (5-30 min, 900°C) used in this study.

 The inclusion characteristics (such as morphology, composition, size range) in steel can be well determined by using a 3D analysis of inclusions after an electrolytic extraction of the heat-treated specimens of a 42CrMo4 steel (0.42% C).

2. Remelting of Fe-0.4%Mn-0.015%S with addition of REM for modification of sulfides:

 REM addition in the melt helps to reduce the dissolved S content in the liquid steel due to formation complex REM-oxysulfides.

 Most of the observed REM-oxysulfide inclusions have a spherical shape and the inclusion diameters vary from 0.5 to 5.0 µm.

5. Future work

 A quantitative investigation of the MnS: particle size distribution, aspect ratio, etc.

(25)

21

6. Acknowledgement

(26)

22

7. References

[1] Sims C E. The non-metallic constituents of steel[J]. Transactions of the Metallurgical Society ofAIME, 1959,215(3): 367-393.

[2] Ito Y, Yonezawa N, Matsubara K. The composition of eutectic conjugation in Fe-Mn-S system[J]. Transactions of the Iron and Steel Institute of Japan, 1980, 20(1): 19-25.

[3] Ito Y, Yonezawa N, Matsubara K. Effect of carbon on the composition of eutectic conjugation in the Fe—Mn~S system and equilibrium composition of sulphide in solid steel[J]. Transactions of the Iron and Steel Institute of Japan, 1980, 20(5); 301-308.

[4] Ito Y, Masumitsu N, Matsubara K. Formation of manganese sulfide in steel[J]. Transactions of the Iron and Steel Institute of Japan, 1981,21(7):477-484.

[5] Oikawa K, Ohtani H, Ishida K, Nishizawa T. The control of the morphology of MnS inclusions in steel during solidification[J]. ISIJ International, 1995, 35(4): 402-408.

[6] Anmark N, Karasev A, Jönsson P G. The effect of different non-metallic inclusions on the machinability of steels[J]. Materials, 2015, 8(2): 751-783.

[7] Karasev A, Inoue R,Tilliander A,Jönsson, PG. Application of electrolytic extraction for three-dimensional investigation of inclusion characteristics in the steelmaking area. In ISIJ-VDEh-Jernkontoret Joint Symposium—14th VDEh-ISIJ Seminar on Science and Technology of Process Metallurgy, Proceedings of the 8th Japan-Nordic Countries Joint Symposium on Science and Technology of Process Metallurgy, Osaka, Japan, 15–16 April 2013; pp. 1–5.

[8] Karasev A, Bi Y, Jönsson P G. Three-dimensional investigation of large-size inclusions and clusters in steels by using the electrolytic extraction technique[C]//AISTech 2013 Iron and Steel Technology Conference; Pittsburgh, PA; United States; 6 May 2013 through 9 May 2013. 2013: 1139-1145.

[9] Kanbe Y, Karasev A, Todoroki H, Jönsson P G. Application of Extreme Value Analysis for Two- and Three-Dimensional Determinations of the Largest Inclusion in Metal Samples[J]. ISIJ International, 2011, 51(4):593-602.

(27)

23

[11] Kanbe Y, Karasev A, Todoroki H, Jönsson P G. Analysis of Largest Sulfide Inclusions in Low Carbon Steel by Using Statistics of Extreme Values[J]. Steel Research International, 2011, 82(4): 313-322.

[12] Janis D, Inoue R, Karasev A, Jönsson, P G. Application of different extraction methods for investigation of nonmetallic inclusions and clusters in steels and alloys[J]. Advances in Materials Science and Engineering, 2014, 2014:1-7.

[13] McFarland W H, Cronn J T. Spheroidization of type II manganese sulfides by heat treatment[J]. Metallurgical and Materials Transactions A, 1981, 12(5): 915-917. [14] Murty Y V,Morral J E, Kattamis T Z, MehrabianR. Initial coarsening of manganese sulfide inclusions in rolled steel during homogenization[J]. Metallurgical and Materials Transactions A, 1975, 6(11): 2031-2035.

[15]Shao X, Wang X, Jiang M, Wang W. Effect of heat treatment conditions on shape control of large-sized elongated MnS inclusions in resulfurized free-cutting steels[J]. ISIJ international, 2011, 51(12): 1995-2001.

[16] Gmpp J C, Matway R J. Residual elemenls in slainless steels [C]//Electric Furnace Proceedings. Dallags, 1996, 497-451.

[17] Nabeshima S, Ando K, Nakato H. Effect of Al and Re metal concentration on the composition of inclusions in Si-Mn killed steel[C]//Sino-Janp Symposium on Science & Tech. Iron & Steel. Xian, 2001, 59.