Free Access
Issue
Metall. Res. Technol.
Volume 117, Number 1, 2020
Article Number 110
Number of page(s) 12
DOI https://doi.org/10.1051/metal/2020002
Published online 14 February 2020

© EDP Sciences, 2020

1 Introduction

High-carbon heavy rail steel has the advantages of energy saving and is widely used for transportation rails. It requires excellent fatigue resistance, high strength, high wear resistance and high corrosion resistance due to the tight contact between the rail and high-speed wheel resulting in circularly and strong alternating stress [1]. Non-metallic inclusions deteriorate the performance of rail steels. During the steelmaking and refining process, the aluminum-free of SiMn alloy is usually applied for deoxidation in order to avoid the formation of Al2O3 inclusions. However, the MnS precipitated particles in rails steels turn into the main problem. MnS particles precipitate during solidification and cooling of the steel, the most direct and effective method to eliminate MnS particles is to reduce the content of Mn and S in the steel. However, the Mn is an essential additive element and the S can hardly be reduced to a reasonable scope due to the restriction of the production rhythm and the considerations of production cost. Hence, controlling MnS particles in heavy rail steels is of very importance.

The various morphologies of MnS particles in steel were well investigated [24]. Lawrence et al. [5] investigated on rail steels with four levels of sulfur, finding that the cracks were generated at the end of pure MnS particles, and MnS was clearly separated from the base. Wakoh et al. [6] studied the effect of the S content on the MnS precipitation with various kinds of oxide nuclei and proposed that the precipitation of MnS on oxide inclusions was affected by the content of S in steel and the composition of the oxide. Yaguchi [7] reported that the MnS could nucleate in the molten steel if the S is as high as 0.3% and a lot of MnS particles still precipitated during solidification and cooling, and with the increase of holding time during heating, the size of MnS particles increased. Garbarz et al. [8] reported that for the steel containing 0.6 Mn and 0.025–0.037 S (wt.%) and being cooled at an average 250 K · s−1 cooling rate, the nano-particles in the as-cast steel specimens were 10–100 nm in size, averaging 30 nm, and the dissolution temperature for each type of the nano-MnS particles was 1573 K (1300 °C). Li et al. [9] studied an appropriate cooling rate and dissolved aluminum content for the control of MnS morphology in medium-carbon non-quenched and tempered steel, and found that the high cooling rate promoted the formation of globular MnS particles. Goto et al. [10] reported that the composition of < 10 µm oxide inclusions changed with the cooling rate during solidification of a Ti-deoxidized steel.

Baker et al. [11] observed that the perlitic shape MnS particles in steel were changed to globular shape under argon atmosphere at 1473 K (1200 °C) after a 0.5 h holding time. Matsubara [12] found that fine sulfide particles precipitated along austenite grain boundaries when the steel was cooled to 1473 K (1200 °C). Nishida et al. [13] found that average diameter of MnS particles nonlinearly depended on holding time, since the second phase was constricted and split during the heating process of the steel. Heckel et al. [14] and Nichols [15] confirmed the existence of the MnS constriction through theoretical calculation and experimental observation. Segal et al. [16] investigated the influence of particle size on the deformation behaviour of irregular MnS particles in steel during hot rolling. Gnanamuthu et al. [17] studied the shape change of different types of MnS particles and observed that all MnS particles finally changed to polyhedral shape after a prolonged holding time. Shao et al. [18] observed that the heat treatment had a significantly influence on the size and shape of large-sized elongated MnS particles in resulfurized free-cutting steels. Kim et al. [19] proposed that the MnS particles precipitated on the surface of MnO–SiO2 oxides in Si–Mn deoxidized steels and with increased cooling rate, both the size of the particle and the precipitation ratio of the MnS phase in the oxides decreased, and concluded that the majority of the MnS islands found in the specimen after isothermal holding were formed by the diffusion of Mn and S from the steel matrix to the oxide inclusions.

Although many relevant investigations have been reported, studies on the conditions suitable to control the number, size and morphology of sulfide particles in the heavy rail steel are rarely reported. In rail steels, MnS particles may cause the deterioration of transverse impact properties if the MnS particles are relatively large and elongated in the hot rolling direction. As a consequence, studies on the control of MnS particles in rail steels are necessary. In the present work, the effects of cooling rate and isothermal holding on the characteristics, i.e., the number, size and morphology of MnS particles in the heavy rail steel were investigated to elucidate the formation mechanisms of MnS by experimental studies and thermodynamic calculation. The optimal parameter of control MnS was proposed to improve the steel quality.

2 Experimental methodology

2.1 Experimental materials

Steel samples were taken in one heat of the heavy rail steel with composition in Table 1, produced by a route of “BOF steelmaking→ LF refining→ RH refining→ Continuous Casting→ rolling”. During tapping of a BOF converter, the SiMn alloy was added to the ladle for deoxidation. Initial rolling temperature of bloom was 1290 °C ± 10 °C. Steel samples from the continuous casting bloom were cut into 10 × 10 × 5 mm cubes. Approximately 40 g steel material was melted in an alumina crucible with a diameter of 12 mm, and a depth of 90 mm under a purified argon gas atmosphere.

After the steel material was melted at 1873 K (1600 °C), small samples (Φ6 × 90 mm) were obtained through the quartz tube. A 70 mm long solidified rod was use for the electrolytic separation of particles from steel as shown in Figure 1a, then the residual samples with 20 mm high cylinder were used for two-dimensional metallographic observation. Besides, several 3 mm thick plate cylinder samples were cut for heat treatment investigations as shown in Figure 1b. These plates are thin enough to avoid the interference of the sample size with the time of store energy during heating and cooling process.

Table 1

Chemical composition of the rail steel material.

thumbnail Fig. 1

Schematic of sample preparation (I: electrolytic sample; II: metallographic specimen; III: isothermal holding samples).

2.2 Experimental procedures

The heating experiments were carried out using a vertical MoSi2 electric-resistance furnace under argon atmosphere equipped with a thermocouple and series automatic temperature controller with an accuracy of ± 2 K. Since the melting point of MnS is 1883 K (1610 °C) the MnS-free state were reached after the stabilization of the melt at 1903 K (1630 °C) holding over 60 min. After holding at 1903 K (1630 °C) for 60 min, three cooling modes were employed: (1) water cooling: a portion of the melt was sucked into a silica tube and immediately quenched in water to form rod-shaped specimens, resulting in an average 80.4 K · s−1 cooling rate; (2) air cooling: the melt in crucible was taken out by the quartz tube and cooled to room temperature in air; (3) furnace cooling: the crucible with the remaining melt was continuously cooled down to room temperature in the furnace by turning off the power. The cooling rate was measured as follows: the quartz tube and the melt were then quenched into the bucket, at the same time, a thermocouple was placed in the center of the quartz tube. Ultimately, the cooling rate was characterized by the thermocouple showing the change in temperature. The cooling rates of the air cooling and furnace cooling were 3.8 K · s−1 and 1.8 K · s−1, respectively. During heating experiments, just the quenching samples were heated at 1473 K (1200 °C), 1573 K (1300 °C) and 1673 K (1400 °C), with different holding time of 10, 30, 60 and 120 min, respectively. After holding for a predetermined period of time, the sample was quickly dropped into the water. The diagrammatic sketch of heat treatment process is showed in Figure 2.

thumbnail Fig. 2

Diagrammatic sketch for the process of cooling method and isothermal holding.

2.3 Analysis of samples

The resulted samples were mounted and well-polished. An automated analysis tool named ASPEX (Application Specific Products employing Electron Beam and X-ray technology) Explorer was used to quantify the size, composition, morphology and amount of inclusions. The characteristics of MnS particles on 15 mm2 cross section area of each sample were analyzed. The types of inclusions were categorized according to the weight percentage of elements detected in them. Thus, the weight percentage of Mn > 30% and S > 20% was defined as a MnS particle. The detailed rule file is showed in literature [20].

In order to reveal the three-dimensional (3D) morphology of inclusions and particles, the electrolytic extraction using non-aqueous electrolyte [21] was used. During this extraction process, the steel sample served as the anode, the stainless steel worked as the cathode, and the composition of the electrolytic solution was 5 pct triethanolamine+ 1 pct 4-methyl ammonium chloride+ 5 pct glycerin+ 89 pct methyl alcohol. The electrolytic extraction were performed under 0.04 A · cm−2 electric current, 0–5 °C temperature, and 8 h electrolytic time. The main steps for electrolytic extraction were sample electrolysis→ elutriation→ magnetic separation→ ultrasonic cleaning→ washing and drying→ weighing→ characterization of extracted particles. Thus, 3D morphologies of inclusions were observed by scanning electron microscopy (SEM) equipped with energy dispersive spectroscopy (EDS).

3 Effect of cooling rates

3.1 Effect of cooling rates on morphology of MnS particles

The cooling rate has a major effect on the size, morphology and the number density of MnS during cooling. Typical two-dimensional (2D) morphologies of inclusions in rail steels cooled at three cooling modes are shown in Figure 3. The dark region of the inclusions is MgO–Al2O3–CaO–SiO2 (TiOx) oxide whereas the bright area is MnS phase.

Figure 4 shows 3D morphology of inclusions at three cooling modes. With water cooling, there were less number of MnS particles, mainly spherical and spindle phases stemming from the metastable monotectic reaction. (Figs. 4a1 and 4a2). With air cooling, a large number of MnS particles were observed, mainly spherical or in short irregular rods (Figs. 4b1 and 4b2). With furnace cooling, MnS particles were rod-like (Fig. 4c1) resulted from either eutectic reaction or monotectic reaction, and some of them were as long as over 100 μm. The size of MnS particles increased with the decrease of the cooling rate, which has been reported elsewhere [22]. Hence, their size at furnace cooling was larger than other cooling modes. It was found that many oxide inclusions were partially or fully covered with MnS particles, as example in Figures 4b3, 4c2 and 4c3.

Generally, MnS precipitated on the surface of the oxide inclusions as heterogeneous nucleation during cooling of the continuous casting bloom. With water cooling mode, due to the short cooling time, there were little precipitated MnS phase on the surface of the oxide, as shown in Figure 4a3. With air cooling, due to the long cooling time, there were a little MnS phase was precipitated on the surface of some oxide inclusions, as shown in Figure 4b3. With the further decrease of the cooling rate, the MnS was precipitated on the surface of oxide inclusions, and the cover percentage of MnS on the oxide was far larger than that of water cooling and air cooling, as shown in Figure 4c3. For furnace cooling, over 90% of the oxides co-existed with MnS phase. The influence of the cooling rate on the size of MnS particles is reflected by two important parameters of temperature and time. If the cooling starts from a high temperature, longer time is needed for the formation of MnS by the diffusion of Mn and S in the steel. In the current study, isothermal holding experiments were established in order to determine the effects of temperature and time on the precipitation rate of MnS, which will be discussed in a later section.

thumbnail Fig. 3

Typical 2D morphology of inclusions at three cooling modes.

thumbnail Fig. 4

Typical 3D morphology of inclusions at three kinds of cooling methods.

3.2 Effect of cooling rate on the size, amount and area fraction of MnS particles

The effect of the cooling rate on the characteristics of MnS particle, including the number density, area fraction and average diameter of inclusions were quantitatively analyzed using automatic SEM scanning and is given in Figure 5. The number density of MnS particles under the condition of water cooling (80.4 K · s−1) and furnace cooling (1.8 K · s−1) was smaller than that of air cooling (3.8 K · s−1). The area fraction of MnS particles increased with decreasing cooling rate, indicating larger MnS particles with smaller cooling rate as indicated by Figure 4c1. Because of this small size MnS was dissolved into steel matrix again during a slower furnace cooling, while the larger size MnS occurs coarsening, which ultimately causes the area fraction to increase and number density to decrease. The average diameter of MnS particles increased with the cooling rate decreased, the reason is that the Ostwald ripening occurred in steel samples due to the cooling rate decreased, resulting in increasing diameter of inclusions [23]. Although the size of MnS in sample measurement was larger than those obtained in the literature [2426], the relationship between the size of MnS particles and the cooling rate follows the trend of the diameter of MnS particles reduces with the increase of cooling rate. The relationship between the average diameter and the area fraction of MnS was nearly exponentially.

thumbnail Fig. 5

Characteristics of MnS particles at three cooling modes.

4 Effect of isothermal holding time and temperature

4.1 Variations of 2D morphologies of MnS particles

The typical 2D morphology of pure MnS particles at different holding time and holding temperature is presented in Figure 6. During holding, there was little change of morphology for MnS particles at 1473 K (1200 °C) holding temperature. At the higher holding temperature of 1573 K (1300 °C), the morphology of MnS changed little before 30 min holding time and the size slightly increased with time after 60 min holding. However, the number and size rapidly decreased with time after 30 min holding at 1673 K (1400 °C). Significant change from a nearly spherical and spindle-like to clustered and irregular of MnS particles was identified only when the holding time exceeded 60 min or 120 min at 1573 K (1300 °C) and 1673 K (1400 °C) holding, respectively.

thumbnail Fig. 6

Morphologies of MnS particles at different isothermal holding time and holding temperature.

4.2 Effect of isothermal holding on the size, amount and area fraction of MnS particles

Figure 7 shows the effect of holding time and holding temperature on the number density of MnS particles. The amount of MnS particles decreased with increasing holding time at 1573 K (1300 °C) and 1673 K (1400 °C), but little changed at 1473 K (1200 °C), which indicates that the quantity of MnS was not sensitive to time at a lower temperature. Figure 8 shows the variation of the area fraction of MnS particles, and their area fraction had a gradually increased tendency as time at 1473 K (1200 °C) holding temperature. Nevertheless, the area fraction of MnS particles decreased with increasing holding time at 1573 K (1300 °C) and 1673 K (1400 °C). Figure 9 shows that the size of MnS particles depended little on the holding time at 1473 K (1200 °C), but decreased linearly with time at 1673 K (1400 °C), especially after 30 min holding, which indicates that MnS particles were redissolved into steel matrix at high temperature. On the contrary, the size of MnS particles presented increasing trend from 1.8 to 2.8 µm at 1573 K (1300 °C) when the holding time exceed 60 min because of Ostwald ripening. After contrast, it was found that the number density and area fraction of MnS particles decreased while the average diameter decreased as the holding time increased from 10 to 120 min at 1673 K (1400 °C). At last, it is concluded that the characteristics of MnS particles had little change at low holding temperature of 1473 K (1200 °C), but significantly changed at higher holding temperature of 1673 K (1400 °C).

Comparing the relationship between size distribution and number density of MnS particles at different isothermal holding temperature was showed in Figure 10. It was found that there were many MnS particles at 1473 K (1200 °C), especially for those of less than 3 µm diameter. As expected, it was showed that the number density of less than 3 µm MnS particles sharply decreased by an order of magnitude at 1573 K (1300 °C) holding temperature. However, contrary to the size of MnS increased when the holding time at 120 min, especially appeared those of the size more than 5 or 10 µm, which indicates that the 1 ∼ 3 µm MnS particles were dissolved, but that > 3 µm ones grow up. At 1673 K (1400 °C), the size of MnS particles decreased with increasing holding time and was lower than other isothermal conditions. What is more, the MnS particles of more than 5 µm hardly existed in steel at 60 min holding time, indicating that the MnS particles had dissolved completely in steel matrix at higher holding temperature of 1673 K (1400 °C). It turned out that the longer holding time, the number of less than 3 µm MnS particles, except for those of 1573 K (1300 °C). The size distributions of the MnS particles had a decreased trend as increasing holding temperature. The fact that the number of small MnS particles decreased with increasing holding temperature implies the dissolution of MnS particles during holding, which agrees with the theory of Ostwald ripening.

thumbnail Fig. 7

Number density of MnS particles at three holding temperature.

thumbnail Fig. 8

Area fraction of MnS particles at three holding temperature.

thumbnail Fig. 9

Average diameter of MnS particles at three holding temperature.

thumbnail Fig. 10

Effect of holding temperature and time on the size distribution of MnS particles.

5 Thermodynamic considerations

5.1 Effect of cooling rates on segregation of manganese and sulfur during solidification

In general, the large size of MnS precipitation was caused by the segregation of Mn and S atoms during cooling and solidification, especially under the condition of low sulfur, such as the current work with a 81 ppm sulfur content.

It was reported that the solidification path of some alloy element such as Mn, S, Ti and Al is close to the Scheil model [27]. The solidification path of non-alloy elements such as C, V and N is more close to the Lever-rule model [28]. The diffusion of alloy and non-alloy solute elements in metal should be discussed separately. The suitable selection of precipitation models should consist with the actual situation chemical composition and other practical conditions. For the precipitation theory of MnS in steel, the back-diffusion parameter was proposed [2936] according to the morphology of dendrites. Different segregation models should be adopted for different chemical elements, since different elements have different Fourier number of the solid phase during cooling. For the current heavy rail steel, Ohnaka’s equations are used to calculate the microsegregation of solutes in molten steel during solidification for different cooling rates, the segregation models are represented by equations (1)(5) [3740]. (1) where CL is the concentration of a given solute element in the liquid at the solid–liquid interface, C0 is the initial (nominal) liquid concentration, k is the equilibrium partition coefficient for that element, and fS is the solid fraction. Where and , both of β and α are back-diffusion parameters, where 2α and 4α for the plate and columnar models, respectively. (2) (3) (4)where Di is the diffusion coefficient of solute i in the solid phase in cm2 · s−1, tf is the local solidification time in seconds, λS is the secondary dendrite arm spacing in micron, and RC is the cooling rate in K · s−1, CC is the carbon content in weight percent carbon, TL and TS are liquidus and solidus temperature in K, respectively.

In a region where solid and liquid coexist in a dendrite cell, the temperature at the solid-liquid interface, T, is given by equation (4) [41]. (5) where T0 is the melting point of pure iron [1809 K (1536 °C)].

Data of equilibrium partition coefficients and diffusivity of various solute elements in γ iron employed by Won et al. were accepted in the present study, and reported in Table 2.

Figure 11 illustrates the effect of cooling rate on segregation of manganese and sulfur during solidification. The segregation of sulfur is much stronger than that of manganese, although both manganese and sulfur are strong segregation elements. When the cooling rate changes for three cooling methods, the effect of the cooling rate on the concentration ratio of manganese and sulfur is insignificant.

The equilibrium of Mn with S in molten steel and austenite phase is obtained by equations (6) and (7) [42]. (6) (7) where K is equilibrium constant of MnS, T is temperature of molten steel during solidification in K.

The relationship between actual concentration and equilibrium concentration of manganese and sulfur were calculated based on equations (1) and (7) during solidification as shown in Figure 12. It reveals that the actual concentration of manganese and sulfur in the remnant liquid phase sharply increased with increasing solid fraction. When the solid fraction is close to 0.9567 (the corresponding temperature is 1630 K (1357 °C) by Eq. (5)), MnS begins to precipitate and the actual concentration of manganese and sulfur in the remnant liquid phase far outweighs that of equilibrium concentration.

The precipitation temperature of individual MnS for the current steel with composition in Table 1 using the FactSage 6.4 thermodynamic software was calculated as shown in Figure 13. The formation temperature of MnS in the steel is 1627 K (1354 °C). In addition, the liquidus and solidus temperatures were also calculated by this software and their temperature are 1731 and 1619 K (1458 and 1346 °C), respectively. This indicates that the above calculation results of Figure 12 are in good agreement with the results of Figure 13.

Table 2

Equilibrium partition coefficients (k), and their diffusion coefficients (D) in γ phases.

thumbnail Fig. 11

Effect of three cooling methods on the segregation of (a) sulfur and (b) manganese during solidification.

thumbnail Fig. 12

Variations of manganese and sulfur concentration product with solid fraction.

thumbnail Fig. 13

Equilibrium precipitation of MnS in the current steel during solidification.

5.2 Growth of MnS particles during solidification

To describe the growth of MnS particles, it is assumed that the precipitate is spherical, a stationary diffusion state is reached and each precipitate grows independently without interaction with other precipitates [43]. Because the segregation of manganese and sulfur occurs during solidification, when the reaction of manganese and sulfur reaches equilibrium at the solidifying front, MnS particles begin to precipitate and grow at the same time. Assuming that the only one MnS was formed in each interdendritic volume. When the MnS begins to precipitate, the actual concentration of manganese in the remnant liquid phase far outweighs that of sulfur. Judged from the Figure 11, the growth of MnS is assumed to be controlled by the diffusion of sulfur.

MnS particles are precipitated and grow with enrichment of solute at the solidifying front of molten steel (i.e. solid/liquid interface of steel) because of microsegregation of elements during solidification. MnS grows when the concentration solubility product of Mn and S exceeds the equilibrium value of MnS. The mass balance in the unit volume surrounding one MnS is obtained by considering the decrease in solute content in molten steel caused by MnS growth.

The driving force of MnS growth can be evaluated by the difference (CL− Ceq) between the solute content CL increased in molten steel by microsegregation between dendrite arms and the solute content Ceq in equilibrium with the MnS. Assuming the equilibrium between the MnS and the molten steel shown by equation (6) and considering that MnS particles growth is caused by the diffusion of solute element in the molten steel, the following equation of the diffusion growth was used, as expressed by equation (8) [44]. (8) where r is the radius of MnS (µm), MMnS is the molar mass of MnS (0.087 kg · mol−1), is the molten steel density (7070 kg · m−3), Di is the confusion coefficient of sulfur in molten steel (cm2 · s−1), MFe is the molar mass of steel (0.056 kg · mol−1), ρMnS is the density of MnS (3990 kg · m−3), Ceq is the solute content in equilibrium of MnS.

The results of radius can be obtained by equation (9). It is obvious that the radius of MnS particles is greatly affected by the local growth time (as shown in Eq. (3)), which is determined by the cooling rate. (9)

Figure 14 shows the calculating results of the diameter of MnS particles during solidification for three cooling methods. The solid fraction at which MnS is entrapped by solid/liquid interface of steel is considered to be influenced by the cooling rate. Therefore, it is difficult to directly compare the calculated diameter of MnS growth with the actual value of MnS, but relative comparison is possible. In the case of the high cooling rate (water cooling), the local solidification time is short, and the diffusion time for solute element is also short. Therefore, the MnS growth is limited and the diameter of MnS is about 4.2 µm at a solid fraction of 0.96. When the cooling method is air cooling, the local solidification time is long, and the MnS diameter at a solid fraction of 0.96 (the corresponding temperature is 1629 K (1356 °C) by Eq. (4)) is about 20 µm. On the furnace cooling condition, local growth time is the longest, and the diameter of MnS is about 100 µm at a solid fraction of 0.98 (1351 °C). This is because the molten steel is flowing at all times in the two-phase zone, and the MnS had enough time and space of growth in the front of solidification. Moreover, MnS that have been precipitated can be pushed forward by solidified dendrites resulting in growing up of MnS constantly. However, the size of MnS particles which were captured by the secondary dendrite is smaller than the secondary arm spacing. Therefore, a lot of more than 100 µm MnS particles were obtained in sample of furnace cooling, and it were formed in the front of solidified dendrite. Of course, 1 ∼ 10 µm MnS were also found and it were captured by the secondary dendrite. These values are close to the actually measured value as shown in Figures 4a1, 4b1 and 4c1.

thumbnail Fig. 14

Variations of the diameter of MnS particles in steel samples for three cooling methods.

5.3 Effects of isothermal holding on the thermodynamic behavior of MnS precipitation

The diffusion rate of Mn and S atoms in steel matrix increased with the increase of temperature. The variation in the area fraction of MnS particles was mainly due to the diffusion caused by the difference of concentration between MnS particles and the variation of solid solubility caused by the variation of temperature. High holding temperature and long holding time favors the dissolution of MnS particles which can be applied to the actual production in reheating furnace.

Previous experimental results showed that MnS precipitated in the solid and liquid phase zone and the concentration product of [Mn] and [S] in the liquid and solid steel was equations (10) and (11) [42]. (10) (11) where [%Mn] and [%S] are the mass percentages of Mn and S in steel, respectively. fMn and fS are the activity coefficients of elements in molten steel at 1873 K (1600 °C).

Assuming that Mn and S in steel had been changed into MnS except the solid solution and the following relationship was expressed on the basis of stoichiometric ratio of MnS in equation (12) [45]. (12) where Mn and S are the original mass fraction of Mn (0.91%) and S (0.0081%) in steel, [Mn] and [S] represent the solid solution Mn and S, AMn and AS are the atomic mass (AMn = 54.938, AS = 32.06), respectively.

Combining equations (10)(12), the variation of [Mn] and [S] and the amount of MnS precipitation could be calculated at different temperatures, as shown in Figure 15. The amount of MnS precipitation was calculated about 0.0218% by FactSage 6.4 with almost equal the results (0.0219%) by calculation of equations (10) and (11) as shown in Figures 13 and 15, respectively. What’s more, the total amount of Mn and S solid solution in steel is very little relative to those of initial content under different temperature. The higher the isothermal holding temperature, the better the effect of homogenization treatment on MnS particles was observed. It can be also proved that the diffusion of Mn or S in steel matrix was considered as the controlled factor for the Ostwald ripening rather than dissolution of solute, which is consistent with the growth of MnS controlled by the diffusion of sulfur during solidification.

Assuming that the precipitation which can amount up to 92% was a large amount, the temperature range of a large amount of MnS precipitation was from 1649 K (1376 °C) to 1473 K (1200 °C), as shown in Figure 15. Previous experimental results [46,47] showed that the size of MnS particles kept small due to continuous formation of new nucleuses during the precipitation of MnS. At the same time, the MnS particles grew up rapidly because of constantly dissolved smaller MnS particles during coarsening or isothermal holding process. However, the average size of MnS particles continued to increase throughout the precipitation period. Therefore, subsequent isothermal holding temperature should be above 1473 K (1200 °C) in order to promote ultrafine dispersive precipitation MnS particles with hindering grain growth.

thumbnail Fig. 15

Variation of [%Mn] and [%S] and amount of MnS precipitation at different temperatures.

5.4 Suggestions and improvements in industrial production

From the above experiment results that the area fraction of MnS particles increased at 1473 K (1200 °C) with increasing holing time, resulting from the new nucleation is promoted and what is important is that the solid solubility temperature of MnS had not yet arrived. However, on one hand, the number density of MnS particles distinctly decreased from roughly 87 per mm2 to nearly 14 per mm2 and 0.07 per mm2 at 1573 K (1300 °C) and 1673 K (1400 °C) with increasing holing time, on the other hand, The area fraction of MnS particles obviously reduced from roughly 116 ppm to nearly 11.4 ppm and 0.19 ppm at 1573 K (1300 °C) and 1673 K (1400 °C) with increasing holing time. Thus, the reason is the effect of together the diffusion of Mn and S element and solution which the MnS was again dissolved into steel matrix. Nevertheless, at 1573 K (1300 °C), the average size of MnS particles distinctly rose from 2.32 to 2.85 µm in Figure 9, it is proved that the larger MnS in the original steel matrix had not been dissolved. On the contrary, the size of these particles had grown up, which is in accordance with the theory of Ostwald ripening.

Based on the chemical compositions of present experimental steel, the starting temperature of precipitation of MnS was about 1630 K (1357 °C). In the case of device capability of steel plants, the isothermal holding temperature in the process should be as high as possible, preferably above 1649 K (1376 °C) (The higher limit of temperature of a large amount of MnS precipitation) for this experimental steel according to Figure 15. Thus, it could promote solid solution of MnS particles in steel matrix, but increase the energy costs of production. In additional, it is also possible to adjust the content of sulfur by less than 16 ppm. Only in this way can it assure that the larger MnS which formed during solidification redissolves in the steel matrix. Otherwise, the continuous casting bloom would be seriously damaged by increasing the heating temperature above the solidus. Above all, once they dissolved in steel, solid solubility would decrease with decreasing temperature of the heavy rail steel during rolling process. They will precipitate again in a finely dispersive state and hinder annealing grain growth and finally make for the improvement of the toughness property of the steel.

6 Conclusions

In the current work, the effect of the cooling rate and isothermal holding temperature and time on the characteristics of MnS particles in a heavy rail steel were investigated through experiments and theoretical analysis, the following conclusions were obtained:

  1. The cooling rate has a major effect on the morphology, area fraction and size of MnS. When the cooling rates decreases, the 3D morphology of MnS had changed from a nearly spherical into rod-like. The area fraction and average diameter of MnS increased with decreasing cooling rate;

  2. Effect of isothermal holding temperature on the morphology of MnS was obtained and changed little at 1473 K (1200 °C), but the variation of their shape from a nearly spherical and spindle-like to irregular was observed at higher holding temperature 1673 K (1400 °C) when the holding time exceeded 60 min. In additional, the number density and area fraction of MnS decreased with increasing holding time at 1573 K (1300 °C) and 1673 K (1400 °C), respectively. Hence, higher temperature favors the diffusion of Mn and S in steel matrix;

  3. The effect of cooling rate on the segregation of Mn and S is insignificant, but the segregation of sulfur is much stronger than that of manganese. When the solid fraction is close to 0.9567 [the corresponding temperature is 1630 K (1357 °C)], MnS begins to precipitate and grow up. Thus, the growth rate of MnS is considered to be controlled by the diffusion of sulfur. It has been found that the increase of cooling rate gives rise to the decreased of MnS diameter because the growth time of MnS is short;

  4. Through theoretical calculations and isothermal holding experimental validation, the results show that high holding temperature and long holding time favors the dissolution of MnS particles. Therefore, the way to control the large size MnS is to decrease sulfur content by less than 16 ppm, not by increasing the heating temperature which is above 1649 K (1376 °C).

Acknowledgements

The authors are grateful for support from the National Key Research and Development Program and National Natural Science Foundation of China (Grant No. 2017YFB0701802 and No. 51701044), the High Quality Steel Consortium (HQSC) and Green Process Metallurgy and Modeling (GPM2) at the School of Metallurgical and Ecological Engineering at University of Science and Technology Beijing (USTB), China.

References

  1. F. Hernández, G. Plascencia, K. Koch, Rail base corrosion problem for North American transit systems[J], Eng. Fail. Anal. 16(1), 281–294 (2009) [Google Scholar]
  2. C. Sims, Relative deoxidizing powers of some deoxidizing for steel[J], Mater. Trans. 185(11), 814–825 (1949) [Google Scholar]
  3. Y. Ito, N. Masumitsu, K. Matsubara, Formation of manganese sulfide in steel[J], Trans. Iron Steel Inst. Jpn. 21(7), 477–484 (1981) [CrossRef] [Google Scholar]
  4. K. Oikawa, H. Ohtani, K. Ishida, T. Nishizawa, The control of the morphology of MnS inclusions in steel during solidification[J], ISIJ Int. 35(4), 402–408 (1995) [CrossRef] [Google Scholar]
  5. C. Liu, M. Bassim, S. Lawrence, Evaluation of fatigue-crack initiation at inclusions in fully pearlitic steels[J], Mater. Sci. Eng.: A. 167(1), 107–113 (1993) [CrossRef] [Google Scholar]
  6. M. Wakoh, T. Sawai, S. Mizoguchi, Effect of S content on the MnS precipitation in steel with oxide nuclei[J], ISIJ Int. 36(8), 1014–1021 (1996) [CrossRef] [Google Scholar]
  7. H. Yaguchi, Manganese sulfide precipitation in low-carbon resulfurized free-machining steel[J], Metall. Trans. A. 17(11), 2080–2083 (1986) [CrossRef] [Google Scholar]
  8. B. Garbarz, J. Marcisz, J. Wojtas, TEM analysis of fine sulphides dissolution and precipitation in steel[J], Mater. Chem. Phys. 81(2), 486–489 (2003) [Google Scholar]
  9. M. Li, F. Wang, C. Li, Z. Yang, Q. Meng, S. Tao, Effects of cooling rate and Al on MnS formation in medium-carbon non-quenched and tempered steels[J], Int. J. Miner. Metall. Mater. 22(6), 589–597 (2015) [CrossRef] [Google Scholar]
  10. H. Goto, K. Miyazawa, W. Yamada, K. Tanaka, Effect of cooling rate on composition of oxides precipitated during solidification of steels[J], ISIJ Int. 35(6), 708–714 (1995) [CrossRef] [Google Scholar]
  11. T. Baker, J. Charles, Type II manganese sulphides: Their deformation and effect on steel fracture[J], J. Iron Steel Inst. 211(3), 187–192 (1973) [Google Scholar]
  12. K. Matsubara, On the behaviors of the precipitated sulfide inclusions in solid steel[J], Tetsu-to-Hagane 51(12), 2220–2232 (1965) [CrossRef] [Google Scholar]
  13. M. Nishida, T. Kato, T. Tanaka, Spheroidization of manganese sulfide inclusion during heating and its effect on hydrogen induced cracking in rolled steel[J], Tetsu-to-Hagane 67(9), 1533–1541 (1981) [CrossRef] [Google Scholar]
  14. R. Heckel, R. Degregorio, The growth and shrinkage rates of second-phase particles of various size distributions[J], AIME Met. Soc. Trans. 233(11), 2001–2011 (1965) [Google Scholar]
  15. F. Nichols, On the spheroidization of rod-shaped particles of finite length[J], J. Mater. Sci. 11(6), 1077–1082 (1976) [Google Scholar]
  16. A. Segal, J. Charles, Influence of particle size on deformation characteristics of manganese sulphide inclusions in steel[J], Met. Technol. 4(1), 177–182 (1977) [CrossRef] [Google Scholar]
  17. D. Gnanamuthu, T. Kattamis, M. Flemings, R. Mehrabian, Effect of homogenization on sulfide inclusions in ferrous alloys[J], Metall. Trans. 5(12), 2557–2567 (1974) [CrossRef] [Google Scholar]
  18. X. Shao, X. Wang, M. Jiang, W. Wang, F. Huang, Effect of heat treatment conditions on shape control of large-sized elongated MnS inclusions in resulfurized free-cutting steels[J], ISIJ Int. 51(12), 1995–2001 (2011) [CrossRef] [Google Scholar]
  19. H. Kim, H. Lee, K. Oh, Precipitation behavior of MnS on oxide inclusions in Si/Mn deoxidized steel[J], Met. Mater. 6(4), 305–310 (2000) [CrossRef] [Google Scholar]
  20. Y. Ren, Y. Wang, S. Li, L. Zhang, X. Zuo, S. Lekakh, K. Peaslee, Detection of non-metallic inclusions in steel continuous casting billets[J], Metall. Mater. Trans. B. 45(4), 1291–1303 (2014) [CrossRef] [Google Scholar]
  21. K. Fang, R. Ni, Research on determination of the rare-earth content in metal phases of steel[J], Metall. Trans. A. 17(2), 315–323 (1986) [CrossRef] [Google Scholar]
  22. H. Kim, H. Lee, K. Oh, MnS precipitation in association with manganese silicate inclusions in Si/Mn deoxidized steel[J], Metall. Mater. Trans. A. 32(6), 1519–1525 (2001) [CrossRef] [Google Scholar]
  23. C.K.L. Davies, P. Nash, R.N. Stevens, The effect of volume fraction of precipitate on Ostwald ripening[J], Acta Metall. 28(2), 179–189 (1980) [CrossRef] [Google Scholar]
  24. H. Liu, D. Hu, J. Fu, Analysis of MnS inclusions formation in resulphurised steel via modeling and experiments[J], Materials 12(12), 2028 (2019) [CrossRef] [Google Scholar]
  25. R. Diederichs, W. Bleck, Modelling of manganese sulphide formation during solidification, part I: Description of MnS formation parameters[J], Steel Res. Int. 77(3), 202–209 (2006) [CrossRef] [Google Scholar]
  26. Y. Tanaka, F. Pahlevani, S. Moon, In situ characterisation of MnS precipitation in high carbon steel[J], Sci Rep. 9(1), 1–12 (2019) [CrossRef] [PubMed] [Google Scholar]
  27. T. Battle, R. Pehlke, Mathematical modeling of microsegregation in binary metallic alloys[J], Metall. Trans. B. 21(2), 357–375 (1990) [CrossRef] [Google Scholar]
  28. J. Dupont, A. Marder, M. Notis, C. Robino, Solidification of Nb-bearing superalloys: Part II. Pseudoternary solidification surfaces[J], Metall. Mater. Trans. A. 29(11), 2797–2806 (1998) [CrossRef] [Google Scholar]
  29. I. Ohnaka, Mathematical analysis of solute redistribution during solidification with diffusion in solid phase[J], Trans. Iron Steel Inst. Jpn. 26(12), 1045–1051 (1986) [CrossRef] [Google Scholar]
  30. V. Voller, A semi-analytical model of microsegregation in a binary alloy[J], J. Cryst. Growth 197(1), 325–332 (1999) [Google Scholar]
  31. V. Voller, C. Beckermann, A unified model of microsegregation and coarsening[J], Metall. Mater. Trans. A. 30(8), 2183–2189 (1999) [CrossRef] [Google Scholar]
  32. V. Voller, C. Beckermann, Approximate models of microsegregation with coarsening[J], Metall. Mater. Trans. A. 30(11), 3016–3019 (1999) [CrossRef] [Google Scholar]
  33. D. Xu, A unified microscale-parameter approach to solidification-transport phenomena-based macrosegregation modeling for dendritic solidification: Part I. Mixture average-based analysis[J], Metall. Mater. Trans. B. 32(6), 1129–1141 (2001) [CrossRef] [Google Scholar]
  34. D. Xu, A unified microscale parameter approach to solidification-transport process-based macrosegregation modeling for dendritic solidification: Part II. Numerical example computations[J], Metall. Mater. Trans. B. 33(3), 451–463 (2002) [CrossRef] [Google Scholar]
  35. S. Kobayashi, A mathematical model for solute redistribution during dendritic solidification[J], Trans. Iron Steel Inst. Jpn. 28(7), 535–542 (1988) [CrossRef] [Google Scholar]
  36. L. Nastac, D. Stefanescu, An analytical model for solute redistribution during solidification of planar, columnar, or equiaxed morphology[J], Metall. Trans. A. 24(9), 2107–2118 (1993) [CrossRef] [Google Scholar]
  37. S. Choudhary, A. Ghosh, Mathematical model for prediction of composition of inclusions formed during solidification of liquid steel[J], ISIJ Int. 49(12), 1819–1827 (2009) [CrossRef] [Google Scholar]
  38. T. Clyne, W. Kurz, Solute redistribution during solidification with rapid solid state diffusion[J], Metall. Trans. A. 12(6), 965–971 (1981) [CrossRef] [Google Scholar]
  39. M. Flemings, Behavior of metal alloys in the semisolid state[J], Metall. Trans. B. 22(3), 269–293 (1991) [CrossRef] [Google Scholar]
  40. Y. Won, B. Thomas, Simple model of microsegregation during solidification of steels[J], Metall.Mater. Trans. A. 32(7), 1755–1767 (2001) [CrossRef] [Google Scholar]
  41. Y. Wang, J. Yang, Y. Bao, Characteristics of BN precipitation and growth during solidification of BN free-machining steel[J], Metall. Mater. Trans. B. 45(6), 2269–2278 (2014) [CrossRef] [Google Scholar]
  42. Y. Ueshima, K. Isobe, S. Mizoguchi, H. Maede, H. Kajioka, Analysis of the rate of crystallization and precipitation of MnS in the resulphurized free-cutting steel[J], Tetsu-to-Hagane 74(3), 465–472 (1988) [CrossRef] [Google Scholar]
  43. Y. Wang, J. Yang, X. Xin, R. Wang, L. Xu, The effect of cooling conditions on the evolution of non-metallic inclusions in high manganese TWIP steels[J], Metall. Mater. Trans. B. 47(2), 1378–1389 (2016) [CrossRef] [Google Scholar]
  44. H. Goto, K. Miyazawa, K. Yamaguchi, S. Ogibayashi, K. Tanaka, Effect of cooling rate on oxide precipitation during solidification of low carbon steels[J], ISIJ Int. 34(5), 414–419 (1994) [CrossRef] [Google Scholar]
  45. Y. Gong, C. Cai, J. Chen, S. Zheng, H. Li, Three-dimensional duplex morphology of MnS-AlN and thermodynamic analysis, Charact. Miner. Met. Mater.: John Wiley & Sons, Inc., 2013, pp. 91–97 [Google Scholar]
  46. S. Mishra, V. Kumar, Co-precipitation of copper-manganese sulphide in Fe-3%Si steel[J], Mater. Sci. Eng.: B. 32(3), 177–184 (1995) [CrossRef] [Google Scholar]
  47. W. Sun, J. Jonas, Influence of dynamic precipitation on grain boundary sliding during high temperature creep[J], Acta Metall. Mater. 42(1), 283–292 (1994) [CrossRef] [Google Scholar]

Cite this article as: Xuewei Zhang, Caifu Yang, Lifeng Zhang, Effects of cooling rate and isothermal holding on the characteristics of MnS particles in high-carbon heavy rail steels, Metall. Res. Technol. 117, 110 (2020)

All Tables

Table 1

Chemical composition of the rail steel material.

Table 2

Equilibrium partition coefficients (k), and their diffusion coefficients (D) in γ phases.

All Figures

thumbnail Fig. 1

Schematic of sample preparation (I: electrolytic sample; II: metallographic specimen; III: isothermal holding samples).

In the text
thumbnail Fig. 2

Diagrammatic sketch for the process of cooling method and isothermal holding.

In the text
thumbnail Fig. 3

Typical 2D morphology of inclusions at three cooling modes.

In the text
thumbnail Fig. 4

Typical 3D morphology of inclusions at three kinds of cooling methods.

In the text
thumbnail Fig. 5

Characteristics of MnS particles at three cooling modes.

In the text
thumbnail Fig. 6

Morphologies of MnS particles at different isothermal holding time and holding temperature.

In the text
thumbnail Fig. 7

Number density of MnS particles at three holding temperature.

In the text
thumbnail Fig. 8

Area fraction of MnS particles at three holding temperature.

In the text
thumbnail Fig. 9

Average diameter of MnS particles at three holding temperature.

In the text
thumbnail Fig. 10

Effect of holding temperature and time on the size distribution of MnS particles.

In the text
thumbnail Fig. 11

Effect of three cooling methods on the segregation of (a) sulfur and (b) manganese during solidification.

In the text
thumbnail Fig. 12

Variations of manganese and sulfur concentration product with solid fraction.

In the text
thumbnail Fig. 13

Equilibrium precipitation of MnS in the current steel during solidification.

In the text
thumbnail Fig. 14

Variations of the diameter of MnS particles in steel samples for three cooling methods.

In the text
thumbnail Fig. 15

Variation of [%Mn] and [%S] and amount of MnS precipitation at different temperatures.

In the text

Current usage metrics show cumulative count of Article Views (full-text article views including HTML views, PDF and ePub downloads, according to the available data) and Abstracts Views on Vision4Press platform.

Data correspond to usage on the plateform after 2015. The current usage metrics is available 48-96 hours after online publication and is updated daily on week days.

Initial download of the metrics may take a while.