Evolution of inclusions and change of as-cast microstructure with Mg addition in high carbon and high chromium die steel

Abstract

The effects of Mg on the inclusions and the as-cast microstructure of high carbon and high chromium die steel, a grade of cold working die steel with high C content of 1.4–1.6% and high Cr content of 11.0–13.0%, were systematically investigated. It is found that inclusions vary with the route as Al₂O₃ (No Mg) to MgO·Al₂O₃ + Al₂O₃ (5 ppm Mg), and then to MgO+MgO·Al₂O₃ (11 and 15 ppm Mg). The average diameter of the inclusions decreased from 1.91 μm (no Mg) to 1.29 μm (15 ppm Mg), while the number density increased from 2.69 × 10⁴ mm^− 3 (no Mg) to ∼5.62 × 10⁴ mm^− 3 (15 ppm Mg). The changes in the size and the number density were discussed in terms of the effect of inclusions on the nucleation process and the wettability of them with steel melt. The as-cast microstructures were greatly refined with Mg addition that correlated with the evenly dispersed fine Mg containing inclusions.

Introduction

The inclusions existing, grain size and evenness have great influence on the mechanical properties of steel. Large and hard inclusions are harmful to steel performance, such as strength, toughness, cold forming ability,¹ fatigue resistance² and delayed fracture resistance.³ Grain refinement is generally accepted to be an effective way of improving strength and plasticity of steels without deteriorating toughness.⁴

Mg is very important for the metallurgical processing of several steels. Mg or Mg alloys are added for deoxygenation, desulphurisation, inclusion shape control, grain refinement and improvement in castability in continuous casting. Recently, more and more researchers have focused on the effect of trace Mg added into the steel. The dissolved Mg not only can refine grain to improve the plasticity and toughness but also can modify the inclusion to be harmless. Saxena, who have studied the effect of Mg as the refining agent on the inclusions, indicated that areatus Al₂O₃ inclusions became smaller and random distribution spines (MgO·Al₂O₃) inclusions with Mg addition.⁵ Mg has a strong affinity with O and S, so it is more easier to react with O and S to reduce the harm of MnS on the steel properties, and the oxygen in the steel could be reduced so low after deoxygenation of Mg that the inclusions were refined and distributed evenly.^6–8 In well deoxygenation steels, Mg can dramatically influence the non-metallic inclusion composition and hence the steel performance. Chang et al. ⁹ and Kim et al. ¹⁰ have researched separately the variation of inclusions and microstructures during the solidification of steels with Mg addition after deoxidised by Mn–Si–Ti. The result showed that, with the increasing of Mg, the microstructures of as-cast steel were obviously refined and the phase of inclusions were changed with the route as Ti₂O₃(core) + MnS(periphery) → MgTiO₃(core) + MnS (periphery) → MgO(core) + MnS(periphery). The particles have a size distribution that strongly affects the grain growth inhibition by pinning. It is known that Mg containing oxides have an even dispersion in steel melt, and heterogeneous nucleation sites facilitate the heterogeneous nucleation of grain and result in the finer microstructure.¹¹ The fine and uniform microstructure present a good combination of high strength and high toughness.

In the present study, the laboratory steel specimens in as-cast state with different Mg contents were prepared by adding Mg to the high carbon and high chromium die steel to investigate the evaluation of inclusions and the change of microstructure. Various traits of non-metallic particles such as composition, phase, morphology, size and number density were analysed in each steel. The microstructure of each specimen correlated with the size and number density of inclusions in order to reveal the relationship between them.

Experimental procedure

In order to clarify the effect of Mg addition on the inclusions and the microstructure in high carbon and high chromium die steel, Mg was added into a steel with the chemistry (1·54 mass%C–0·36 mass%Si–0·37 mass%Mn–11·94 mass%Cr–0·92 mass%V–0·83 mass%Mo–0·22 mass%Ni–0·016 mass%Al–0·006 mass%S–0·027 mass%P–0·002 mass%O–0·01 mass%N). Then, the characteristics of non-metallic inclusions and the final microstructure of steel matrix were studied after solidification and subsequent cooling process.

Steel specimen preparation

In order to investigate the evolution of inclusion traits and the change of microstructure with Mg addition, four laboratory specimens of high carbon and high chromium die steels with different Mg contents were prepared by melting about 800 g of high carbon and high chromium die steel in MgO crucibles using a MoSi₂ furnace. The schematic diagram of experimental device is shown in Fig. 1. Magnesium vapour transformed from granular magnesium was injected into molten steel by using the high purity Ar, which acted as the current-carrying gas to have treatment. Sample #1 was obtained using quartz tube after constant temperature of 1600°C for 10 mins. Then, the corundum tube was inserted into the deep of molten steel to inject Ar, and subsequently, 0·2 g granular magnesium was put into corundum tube every 3 min. Samples #2, #3 and #4 were obtained.

Schematic diagram of experimental device

Analyses

The chemical compositions of the prepared steel samples were analysed by chemical analysis method. The T.O, N, C, S and Mg contents in steel were measured separately using a Leco TC 500 N₂/O₂ simultaneous analyser, an infrared C/S simultaneous analyser and an iCAP 6300 ICP-OES analyser. The chemical compositions of prepared samples are shown in Table 1.

Table 1

Chemical composition of prepared steels/mass%

	C	Si	Mn	Cr	Mo	V	Al	Mg/ppm	S	P	T.O	N
No. 1	1·53	0·36	0·38	11·92	0·80	0·92	0·016	…	0·0083	0·025	0·0025	0·0056
No. 2	1·54	0·35	0·36	11·94	0·82	0·91	0·012	5	0·0066	0·023	0·0016	0·0055
No. 3	1·52	0·36	0·37	11·90	0·83	0·92	0·015	11	0·0064	0·026	0·0015	0·0056
No. 4	1·54	0·36	0·36	11·96	0·80	0·90	0·014	15	0·0056	0·023	0·0012	0·0053

After mechanically inlaid, ground and polished, all of steel specimens were prepared as metallographic samples and then subjected to analyse using optical microscopy to obtain the statistical information on the average size and the number density of inclusions. The photos of steel samples were adopted in the view of 500 times to analyse by an image analysis technique. Then, the steel specimens for inclusions were analysed by scanning electron microscopy (SEM) equipped with energy dispersive spectrometry (EDS) and electron probe microanalysis to study the evolutions of phase, composition and morphology of complex inclusions in steels with different Mg contents.

For observation of the microstructures of the steel matrix, the samples were etched using a mixture solution of 4% nitric acid alcohol, and their photos were taken using optical microscopy under the magnification of 200 field. Then, an attempt was made to analyse the effect of Mg addition on the microstructural evolution of steel matrix with respect to traits of inclusions in order to investigate the relationship between them.

Results and discussion

Effect of Mg addition on phase and morphology evolutions of inclusions

Figure. 2 shows the morphological changes of the non-metallic inclusions with Mg addition. The typical morphology shows the inclusions comprising Al₂O₃, MgO·Al₂O₃ and MgO. However, the phase composition changes as Mg content varies.

Morphological evolution of inclusions typically found in each steel: a 0 ppm; b 5 ppm; c 11 ppm; d 15 ppm

In the No. 1 specimen containing no Mg (Fig. 2a , the typical inclusion was composed of Al₂O₃. In the No. 2 specimen containing 5 ppm Mg (Fig. 2b ), complex inclusions comprising Mg–Al oxide were dominant. The ratio of Mg to Al content in the oxide phase was measured to be unity in atomic percent. Thus, it is thought that the oxide formed as MgO·Al₂O₃ and Al₂O₃ phase and the mole percent of which was 52·56 mass% MgO·Al₂O₃+47·44 mass% Al₂O₃. In the No. 3 specimen containing 11 ppm Mg (Fig. 2c ), the constituent of the oxide was similar to that of the No. 2 specimen. However, a new MgO phase was observed and there was no pure Al₂O₃ phase. The oxide phase composition was 69·18 mass% MgO+30·82 mass% MgO·Al₂O₃. In the no. 4 specimen containing 15 ppm Mg (Fig. 2d ), the oxide phase was the same to the No. 3 specimen, but the mole percent of phase changed. The typical inclusion was measured to be the complex inclusion comprising 94·72 mass% MgO and 5·28 mass% MgO·Al₂O₃. Compared with the inclusions in No. 3 specimen, the content of MgO in the oxide phase increased greatly as the Mg content increased.

The average compositions of inclusions with Mg addition are given in Fig. 3, and it shows that the Al₂O₃ content of inclusions decreases from 100 mass% to ∼22 mass%, and the MgO content increases from 0 mass% to ∼77 mass%. Finally, it can be deduced that the pure MgO inclusions will be formed when the Mg content increases sequentially.

Average compositions of inclusions with Mg addition

Thermodynamic calculations were carried out to explain the evolution of the inclusions in steels. Al₂O₃ would react with dissolved Mg to form MgO·Al₂O₃ as shown in equation (1)¹¹ when a trace of Mg is generated in liquid steel. 1 2 MgO·Al₂O₃ would sequentially react with dissolved Mg to form MgO as shown in equation (3)¹² when the Mg content is high enough. 3 4 Therefore, the stability diagram of MgO/MgO·Al₂O₃/Al₂O₃ can be calculated using equations (1)–(4)). The interaction coefficients used in thermodynamic calculations are shown in Table 2, and the activity of element i can be calculated by equation (5). Meanwhile, the activities of MgO·Al₂O₃ in equations (2) and (4) are 0.47 and 0·80 respectively, and the activities of MgO and Al₂O₃ at saturated state are 0·99 and 1 respectively. 5 The calculated phase stability diagram of MgO/MgO.Al₂O₃/Al₂O₃ is shown in Fig. 4, and the data of samples are marked in it. This figure indicates that contents of dissolved Al and Mg in steels are in the region of Al₂O₃ and MgO.Al₂O₃ formation and trend to the region of MgO formation with the increase of Mg content, which can explain the observation of inclusion evolution in steels as shown in Fig. 2. It can be thought that 15 ppm dissolved Mg in the steel is not enough high to form MgO phase. This shows decent agreement with the theoretical evolution route.

Table 2

Interaction coefficients used in thermodynamic calculation at 1873 K

j
i	C	Si	Mn	Cr	V	Mo	Al	P	S	O	N
Al	0·091¹³	0·056¹³	0·035¹⁴	0·0096¹⁵	…	…	0·045¹³	0·033¹⁶	0·03¹⁷	− 1·98¹⁸	− 0·058¹⁹
Mg	− 0·24²⁰	− 0·09²⁰	…	0·05²⁰	…	…	− 0·12²¹	…	− 1·38²²	− 460²³	…

Phase stability diagram of MgO/MgO·Al₂O₃/Al₂O₃

Effect of Mg addition on size and number density of inclusions

The relationship between two-dimensional (2-D) and three-dimensional (3-D) particle diameters derived by Fullman²⁴ is expressed as 6 where is the average 3-D particle diameter, is the average 2-D particle diameter, is the diameter of a particle, and n is the total number of measured particles. Therefore, equation (6) can be used to calculate the average 3-D particle diameter from the measured diameter of particles in a cross-section.

Meanwhile, the number of particles per unit volume, N _V, can be expressed in terms of that of per unit area, N _A ( = n/A _obs, where n is the number of particles on observed area, A _obs) and the average particle diameter, , as follows²⁴: 7 By combining equations (6) and (7), one can obtain the following relationship: 8 In this study, the average size of particles, , and the number of particles per unit area, , were measured in a cross-section and used along with equation (8)²⁴ in order to calculate the number density of inclusions, , in each specimen. The average diameter, number density of inclusions and size distribution were measured and calculated in all of specimens and are presented in Figs. 5 and 6.

It is clearly seen in Fig. 5 that the average diameter of inclusions decreases, and the number density increases with the increase of Mg content in the steel matrix. As seen in Fig. 6, most inclusions in No. 1 specimen were >2 μm, the percentage of inclusions in sizes 2–5 μm was 69·21% and the inclusions < 2 μm were only 27·22. However, the inclusions were fined obviously as Mg content varies. In No. 2–4 specimens, the inclusions < 2 μm increased to 67·73%, 77·68% and 90·30%, while the sizes of 2–5 μm decreased to 30·44%, 22·32% and 9·71% respectively. The inclusions >5 μm decreased to 0% when the Mg contents was 0·0011% and 0·0015%.

Variation in average diameter and number density of inclusions with Mg addition

Variation in inclusions size distribution of samples with Mg addition

It can be clearly seen from the above results that a trace of Mg in the steel can obviously refine the inclusions. The refinement mechanism can be mainly described that the fine Mg containing inclusions can significantly affect the nucleation process of the composite inclusions in molten steel.

The critical nucleation radius of the inclusion can be expressed as follows²⁵: 9 where is the interfacial tension between the inclusion and the molten steel, S _O is the supersaturation degree of the solute element, and V _O is the molar volume of the inclusion. The parameters are shown in Table 3,²⁵ and the critical size of nuclei was calculated (S _O takes the middle value) using equation (9). The result is also listed in Table 3.

Table 3

Parameters and critical nucleation radius of inclusions

Inclusions	V _O/cm³ mol^− 1	/erg cm²	S _O	r _C/mm
MgO	11·0	1800	3·6–7·2 × 10⁴	0·23
Al₂O₃	8·6	2290	37–60	0·65

The critical nucleation radius of MgO in the molten steel is less than that of Al₂O₃, which is only less than a half of it. Thus, it is easy for Mg containing oxides to nucleate and the nucleation rate is so high that the particle is very fine. According to the heterogeneous nucleation mechanism, a large number of fine Mg containing particles existing in the steel melt will be the nucleation sites of other subsequently precipitated inclusions. So, more and finer inclusions are obtained in the steel melt. When the critical nucleation radius decreases, the nucleation rate increases and the inclusions in steel are refined. Owing to its small liquid buoyancy in the steel melt, the remaining fine inclusions increase, resulting from the fact that floating exclude power is small. At last, the number and size of inclusions change significantly.

Moreover, Kimura et al. ¹¹ indicated that the distribution of inclusions in the steel melt result from the contact angle between the inclusion and the steel melt. In the present study, Mg containing inclusions are not susceptible to growth by collision and agglomeration because they have optimum wettability for even dispersion in the steel melt, and the fact results in the changes in the size and the number density of inclusions in Mg containing specimens.¹⁰

Effect of Mg addition on microstructure of as-cast steels

The microstructural evolution of as cast steel with Mg addition is shown in Fig. 7. It is can be observed that the as-cast microstructures with Mg addition were well refined in the specimens, and the grain size decreased from 6·5 to 7·5, 7·8 and 8·5. In the no. 1 specimen containing no Mg (Fig. 7a ), the crystalline grains were coarse, and some columnar crystals were observed. With the increase of Mg content in the steels, the crystalline grains were greatly refined and the number of equiaxed grains increased, all shown in Fig. 7b–d . In the No. 4 specimen containing 15 ppm Mg (Fig. 7d ), the grain size was much finer than that of the No. 1 specimen.

Microstructural evolution of each steel with Mg addition: a 0 ppm; b 5 ppm; c 11 ppm; d 15 ppm

The chemical compositions of the specimens listed in Table 1 are virtually the same except for Mg content, and the other experimental conditions are also the same. Thus, it seems reasonable to infer that the changes of the solidification microstructure were caused by the changes in the compositions and distribution of fine inclusions but not caused by the phase transformation of the solidification process.

First, the impact of inclusions on the microstructure is achieved by pinning effect. When the crystal plane goes through inclusions, the inclusion particles will produce a drag force to the crystal plane to prevent its extension; thereby, the nucleation and growth of grain growth are restricted and the grain is finer, and that is the pinning function of second phase particles. The finer the inclusion particle is and the more the unit volume number is, the stronger the pinning effect is and the finer the grain is. The even dispersion of the Mg containing oxides can be obtained much more easily than that of Al₂O₃ inclusions, which are clustered, so that the grain growth is restricted effectively by the pinning effect of the fine inclusions on the grain boundary.

Second, Turnbull and Vonnegut²⁶ proposed the heterogeneous nucleation theory that the disregistry between some low index plans of inclusion and metal phase can be very small. The interfacial energy required for the transformation is much lower when the disregistry is much smaller, so the nucleation of inclusion phase takes place much more easily and the grain size becomes smaller.

Conclusions

The typical morphology of the inclusions found in the steels was that of inclusions comprising Al₂O₃, MgO.Al₂O₃ and MgO. The phase of the inclusions changed sequentially from Al₂O₃ (no Mg) to MgO.Al₂O₃ + Al₂O₃ (5 ppm Mg), and then to MgO+MgO.Al₂O₃ (11 and 15 ppm Mg). The content of dissolved Mg in the steel is not enough high to form MgO phase.

With the increase of Mg content in the steels, the average diameter of the inclusions decreased from 1·91 μm (no Mg) to 1·29 μm (15 ppm Mg), while the number density increased from 2·69 × 10⁴ mm^− 3 (no Mg) to ∼5·62 × 10⁴ mm^− 3 (15 ppm Mg). These changes in size and number density are thought to result from the fact that the fine Mg containing inclusions significantly affect the nucleation process of the composite inclusions in molten steel that the inclusions in steel are refined when the critical nucleation radius decreases and the nucleation rate increases and from the optimum wettability of Mg containing oxides with steel melt for their even dispersion in it.

The final microstructure of the as cast steel was greatly refined with Mg addition, and the grain size decreased from 6·5 to 7·5, 7·8 and 8·5. It is thought that, first, the evenly dispersed fine Mg containing inclusions had a pinning effect on the microstructure. The finer the inclusion particle is and the more the unit volume number is, the stronger the pinning effect is and the finer the grain is. Second, they took place in the heterogeneous nucleation easily and resulted in the finer microstructure in Mg containing steels.

Footnotes

Acknowledgements

This work was financially supported by the National Natural Science Foundation of China (grant no. 51174050), the Fundamental Research Project of the Ministry of Education of China (grant No. N110402010) and the National High Technology Research and Development Program of China (grant No. 2012AA03A503).

References

Deng

Z. Y.

Zhu

M. Y.

: ‘Evolution mechanism of non-metallic inclusions in Al-killed alloyed steel during secondary refining process’, ISIJ Int., 2013, 53, 450–458. doi: 10.2355/isijinternational.53.450

Murakami

Takada

Toriyama

: ‘Super-long life tension-compression fatigue properties of quenched and tempered 0·46% carbon steel’, Int. J. Fatigue, 1998, 20, 661–667. doi: 10.1016/S0142-1123(98)00028-0

Takai

Seki

Sakita

Takayama

: ‘Effect of Si and Ca addition on delayed fracture of high-strength steels’, Tetsu-to-Hagane, 1993, 79, 685–691.

Jones

S. J.

Bhadeshia

H. K. D. H.

: ‘Competitive formation of inter- and intragranularly nucleated ferrite’, Metall. Mater. Trans. A, 28A, 1997, 2005–2013. doi: 10.1007/s11661-997-0157-8

Saxena

S. K.

: ‘Production of ultra-clean steels with better mechanical properties with magnesium treatment’, in ‘1996 Steelmaking Conference Proceeding, Warrendale, Iron&Steel Soc of AIME’, 89; 1996.

Saxena

S. K.

: ‘Refining reaction of magnesium in steel at steelmaking temperature’, CIM Bull, 1982, 75, 130.

Wang

Zhi

: ‘Study on the effect of magnesium-containing alloys in refining treatment of molten steel’, J. Iron Steel Res., 1989, 1, 21–30.

Mapelli

: ‘Control and engineering of non-metallic inclusions belonging to xSiO2-yCaO-zAl2O3 system in Ca-treated Al-killed and Al-Si-killed steel’, Steel Res. Int., 2006, 77, 462–471.

Chang

C. H.

Jung

I. H.

Park

S. C.

: ‘Effect of Mg on the evolution of non-metallic inclusions in Mn-Si-Ti deoxidised steel during solidification: experiments and thermodynamic calculations’, Ironmaking Steelmaking, 2005, 32, 251–257. doi: 10.1179/174328105X23932

10.

Kim

H. S.

Chang

C. H.

Lee

H. G.

: ‘(Evolution of inclusions and resultant microstructural change with Mg addition in Mn/Si/Ti deoxidized steels’, Scr. Mater., 2005, 53, 1253–1258. doi: 10.1016/j.scriptamat.2005.08.001

11.

Kimura

S. E. I.

Nakajima

Mizouguchi

: ‘Behavior of alumina-magnesia complex inclusions and magnesia inclusions on the surface of molten low-carbon steels’, Metall. Mater. Trans. B, 2001, 32, 79–85. doi: 10.1007/s11663-001-0010-1

12.

Todoroki

Mizuno

: ‘Variation of inclusion composition in 304 stainless steel deoxidized with aluminum’, Iron Steelmaker, 2003, 30, 60–67.

13.

Sigworth

G. K.

Elliott

J. F.

: ‘The thermodynamics of liquid dilute iron alloys’, Metall. Sci., 1974, 8, 298–310. doi: 10.1179/msc.1974.8.1.298

14.

Z. T.

: ‘Thermodynamic description for concentrated metallic solutions using interaction parameters’, Metall. Mater. Trans. B, 2001, 32, 87–103. doi: 10.1007/s11663-001-0011-0

15.

Ohta

Suito

: ‘Thermodynamics of aluminum and manganese deoxidation equilibria in Fe-Ni and Fe-Cr alloys’, ISIJ Int., 2003, 43, 1301–1308. doi: 10.2355/isijinternational.43.1301

16.

‘Steelmaking data sourcebook’ , 280; 1988, New York, Gordon and Breach Science Publishers.

17.

Cho

S. W.

Suito

: ‘Assessment of aluminum-oxygen equilibrium in liquid iron and activities in Cao-Al₂O₃-SiO₂ slags’, ISIJ Int., 1994, 34, 177–185. doi: 10.2355/isijinternational.34.177

18.

Higuchi

Numata

Fukagawa

Shinme

: ‘Effect of method of Ca treatment on composition and shape of non-metallic inclusions’, Tetsu-to-Hagane, 1996, 82, 671–676.

19.

Elliott

J. F.

, Chapter 21, ‘Electric furnace steelmaking’; 1985, Chelsea, MI, Book Crafters Inc.

20.

Nadif

Gatellier

: ‘Influence of calcium and magnesium addition on the solubility of oxygen and sulfur In liquid steel’, Rev. Metall. Cah. Inf. Tech., 1986, 83, 377–394.

21.

Todoroki

Mizuno

: ‘Effect of silica in slag on inclusion compositions in 304 stainless steel deoxidized with aluminum’, ISIJ Int., 2004, 44, 1350–1357. doi: 10.2355/isijinternational.44.1350

22.

Itoh

Hino

Banya

: ‘Deoxidation equilibrium of magnesium in liquid iron’, Tetsu-to-Hagane, 1997, 83, 623–628.

23.

Ohta

Sutio

: ‘Calcium and magnesium deoxidation in Fe-Ni and Fe-Cr alloys equilibrated with CaO-Al₂O₃ and CaO-Al₂O₃-MgO slags’, ISIJ Int., 2003, 43, 1293–1300. doi: 10.2355/isijinternational.43.1293

24.

Ohta

Suito H

: ‘Characteristics of particle size distribution of deoxidation products with Mg, Zr, Al, Ca, Si/Mn and Mg/Al in F-10mass%Ni alloy’, ISIJ Int., 2006, 46, 14–21. doi: 10.2355/isijinternational.46.14

25.

Suito

Ohta

: ‘Characteristics of particle size distribution in early stage of deoxidation’, ISIJ Int., 2006, 46, 33–41. doi: 10.2355/isijinternational.46.33

26.

Turnbull

Vonnegut

: ‘Nucleation catalysis’, Ind. Eng. Chem., 1952, 44, 1292–1298. doi: 10.1021/ie50510a031