Effects of Deoxidation Processes on Inclusions in Environmentally Friendly Free-Cutting Steel

Abstract

Deoxidation of liquid steel is a key link that affects the quality of environmentally friendly free-cutting steel. The selection and addition sequence of deoxidizer affects the composition, size, distribution, and morphology of non-metallic inclusions, which ultimately affect the machinability of free-cutting steel. The effects of deoxidation processes on inclusions in environmentally friendly free-cutting steel were studied by high-temperature experiments and thermodynamic calculations. The content of total oxygen and inclusion characteristics in steel were analyzed by an oxygen and nitrogen analyzer, a metallographic microscope, a scanning electron microscope, and an energy spectrum analyzer. The results show that the inclusions in the process of an initial Si-Fe alloy addition followed by a manganese addition (H1) are mainly Al₂O₃-SiO₂-MnO composite inclusions in the liquid phase. In addition to the liquid phase Al₂O₃-SiO₂-MnO complex inclusions in the MnO-rich region, there are also some solid phase Al₂O₃-MnO inclusions with high Al₂O₃ content in the process of an initial manganese addition followed by a Si-Fe alloy addition (H2). In the two deoxidation experiments, Bi particles mainly exist in the form of adhesion to MnS inclusions. Referring to H2, the average value and median of the aspect ratio is larger and the number of sulfide inclusions with aspect ratio greater than 1.0 increases significantly in H1. In addition, the spheroidization degree of MnS inclusions in the H2 is relatively good.

1. Introduction

Free-cutting steel is a special type of steel that reduces cutting force and enhances the cutting performance of workpieces by adding free-cutting elements such as sulfur, lead, bismuth, tellurium, and selenium, either individually or in combination, to the steel [1,2,3]. This ultimately leads to improved surface quality and machining accuracy, reduced processing costs, and increased processing efficiency [4,5]. Compared with other steels [6,7], the development direction of free-cutting steel is to strike a balance between reducing processing costs and improving cutting performance. Lead exhibits favorable performance in terms of machinability, as its wettability and molten brittleness contribute to the easy fracturing of chips, thereby yielding a good surface finish when machining steel. However, lead poses risks to the environment and human health. Additionally, it tends to cause macro-segregation, which is detrimental to the mechanical properties of steel [8]. In recent decades, bismuth-containing free-cutting steel as an environmentally friendly machinability material has been widely considered by the metallurgical industry with the development of modern industry towards automation, high speed, and precision processing. Bismuth is non-toxic and has physical and chemical properties similar to lead, with small macroscopic segregation. Replacing lead with bismuth has gradually become a novel idea for the development of green and free-cutting steel grades [9,10,11].

The deoxidation of molten steel is indispensable for improving the quality of steel in the steelmaking process. Lots of research has been conducted [12,13,14] which indicates that oxygen content has a great impact on the form, size, and deformation ability of MnS inclusions in free-cutting steel, and reasonable control of the oxygen content in liquid steel is very important to improve the pouring ability. Deng et al. carried out industrial tests on different deoxidation methods of 35CrMo steel and found that the two different deoxidation methods had little effect on the type and size of inclusions in the final product, which were spherical CaS and CaO-MgO-Al₂O₃ composite inclusions [15]. Liu et al. studied the behavior of inclusions in Si-Ca-Ba composite deoxidized 55SiCr spring steel and found that the addition of composite deoxidizer after alloying could effectively control the shape and size of inclusions [16]. Li et al. conducted laboratory experiments and thermodynamic analysis to study the inclusion behavior of high-aluminum steel under different deoxidation methods, and the results showed that different deoxidation methods had a great influence on the composition, morphology, size, and quantity of inclusions [17]. Furthermore, researchers used a new type of deoxidizer AlMnCa alloy to deoxidize low-carbon and low-silicon steel and compared the influence of different components of the deoxidizer on inclusions and total oxygen content in steel [18].

However, few studies have examined the characteristics of the inclusions from the perspective of different deoxidation processes, especially the different sequences of raw materials added to free-cutting steel. Thus, molybdenum disilicide furnace high-temperature melting experiments combined with thermodynamic calculations were carried out in this paper, to explore the influence of different deoxidation processes on the inclusions in bismuth-containing environmentally friendly free-cutting steel, in order to provide practical production suggestions.

2. Materials and Methods

The experiments were carried out in a tube electric resistance furnace (Jinzhou Electric Furnace Co., Ltd., Jinzhou, China) under laboratory conditions, as shown in Figure 1. The target components (mass fractions, wt.%) of environmentally friendly free-cutting steels were as follows: C, 0.05~0.10%; Si, 0.02~0.06%; Mn, 1.0~1.3%; P, 0.04~0.08%; S, 0.25~0.35%; and Bi, 0.05~0.10%. A corundum crucible (Φ55 × 70 mm) with 600 g industrial pure iron was placed in a graphite crucible and then these were placed together in the constant temperature zone of the tube resistance furnace. Argon with a flow rate of 6 L/min was injected into the furnace through the bottom blow tube. The steel samples were heated to 1873 K under an argon gas protection atmosphere and heat preservation. The temperature-measuring instrument used was a double platinum rhodium thermocouple (PtRH30/PtRH6S). After melting, a four-hole corundum tube (Φ4 mm) was used to transfer oxygen for ten seconds, to maintain a certain amount of dissolved oxygen in the molten steel. Three minutes later, a quartz tube (Φ6 mm) was inserted into the molten steel to conduct the first sample extraction. In the H1 heat, the Si-Fe alloy was initially added. Subsequently, electrolytic manganese (purity 99.6%) was introduced to initiate the deoxidation process. A molybdenum rod was used to stir after the deoxidizer was melted. Samples were collected after each deoxidizer addition. Thereafter, FeS and Bi were added in sequence. The molten steel was maintained at 1873 K for 30 min and then cooled in the furnace. After it had completely solidified, the steel ingot was cut for sampling purposes. In the H2 heat, electrolytic manganese (purity 99.6%) was added first, followed by the addition of Si-Fe alloy for deoxidation. Sampling was conducted after each deoxidizer addition. Subsequent alloy addition sequences and sampling procedures were consistent with those employed in Heat H1. The experimental schematic diagram and sampling process are shown in Figure 2. The compositions of alloys used in the experiments are listed in

Table 1. Composition of alloys used in the experiments/wt.%.

Alloy	C	Si	Mn	S	Al	Bi	P	Fe
Si-Fe	0.03	72.2	-	0.003	1.2	-	0.008	25.6
Mn	0.04	0.1	99.6	0.05	-	-	0.005	0.15
FeS	0.03	0.3	0.08	27.3	-	-	0.03	72
Bi	0.002	-	-	-	-	99.99	-	0.005

.

After the samples were cut and ground, the total oxygen content of the samples was detected by a LECO-TC500 Nitrogen-Oxygen Analyzer (LECO Co., St. Joseph, MI, USA). The contents of carbon and sulfur were detected by a LETO-CS2300 Carbon-Sulfur Analyzer (LECO Co., St. Joseph, MI, USA), and the contents of silicon, manganese, phosphorus, aluminum, and bismuth were measured by Inductively Coupled Plasma Optical Emission Spectrometry (ICP-OES, Focused Photonics Inc., Hangzhou, China). The morphology, chemical composition, size, and distribution of all samples were analyzed by a scanning electron microscope equipped with energy-dispersive spectroscopy (SEM-EDS, FEI Czech LTD, Brno, Czech Republic) when the inclusion samples were inlaid, ground and polished. Inclusions were observed by inspecting randomly selected areas on the polished surface of the samples at different magnifications. During the detection of inclusions, the acceleration voltage of the scanning electron microscope was set at 15 KV, and the entire area of the specific inclusion was analyzed using an energy spectrometer to obtain the average composition of the inclusions. Meanwhile, the different phases existing in the inclusions were also analyzed for composition. The inclusions were classified by size using statistical software.

3. Results

3.1. Change in Molten Steel Composition During Deoxidation

Table 2. Sample compositions of different deoxidation processes/wt.%.

No.	C	Si	Mn	S	Al	Bi	Fe
H1	0.052	0.047	1.21	0.34	0.0034	0.068	Bal.
H2	0.061	0.049	1.20	0.32	0.0028	0.077	Bal.

shows the measured chemical compositions of the final samples from different experimental furnaces. The main compositions (C, Si, Mn, and S) at the end of smelting were similar for the same steel grades. It can be seen from Table 2 that the compositions of the samples meet the expected component requirements. The measurement errors in these chemical analyses were within the usual deviation of ICP-OES and the Carbon-Sulfur Analyzer.

Figure 3 shows the change in the total oxygen mass fraction in molten steel with different deoxidation processes. In the H1 experiment, the initial total oxygen content was relatively high at 0.0370 wt.%. After the addition of ferrosilicon, the oxygen content decreased to 0.0329 wt.%. Subsequently, when manganese was added, the oxygen content became 0.0254 wt.%. In the H2 experiment, the initial total oxygen content in the steel was relatively low at 0.0337 wt.%. Due to the weak deoxidation ability of manganese, the decrease of oxygen content in the steel was not significant [19]. Then after the addition of ferrosilicon, the oxygen content in the steel continued to decline. The total oxygen content of the final samples in each experimental furnace was 0.0082 wt.% and 0.0098 wt.%, respectively. The results highlight that the total oxygen content in molten steel under the two deoxidation processes is not very different, and the deoxidation efficiency of the two processes is equivalent.

3.2. Inclusion Characteristics

3.2.1. Morphology and Composition of Inclusions

The morphology and composition of the inclusions were analyzed by SEM-EDS. Figure 4 shows the morphology and composition of the inclusions in the H1 experiment. The main types of inclusions include separately precipitated Al₂O₃ and SiO₂ inclusions, spherical SiO₂-MnO inclusions, large Al₂O₃-MnO inclusions, spherical Al₂O₃-SiO₂-MnO inclusions, and Al₂O₃-MnO + SiO₂-MnO composite inclusions. In addition, spherical or spindle MnS inclusions with a size between 1 and 3 μm were also found in the sample. The forms of Bi that existed in the free-cutting steel could be summarized as follows: (1) separately precipitated Bi particles, (2) Bi adhered to sulfides, and (3) Bi completely encapsulated by sulfides. Similar findings have been reported in previous studies [20,21].

Figure 5 shows the morphology and composition of the inclusions in the H2 experiment. The inclusions are mainly spherical or ellipsoidal Al₂O₃-MnO inclusions, spherical Al₂O₃-SiO₂-MnO inclusions, and fusiform Al₂O₃-MnO + Al₂O₃-SiO₂-MnO composite inclusions. Spherical or spindle MnS inclusions precipitated separately were also found in this experimental steel, with a size between 1 and 2 μm. Furthermore, there were two forms of Bi particles in this experimental steel: separately precipitated Bi particles and Bi adhering to sulfide.

3.2.2. Quantity, Density, and Size Distribution of Inclusions

Figure 6 shows the quantity, density, and particle size distribution of the inclusions in steel with different deoxidation processes. As is evident from Figure 6, the quantity and density of inclusions in H2 was relatively low, at 154 per mm². The average diameters of the inclusions in the two deoxidation experiments were 2.59 μm and 2.61 μm, respectively. Most of the inclusions were concentrated in the range of 1–3 μm across the two deoxidation experiments. In the H2 experiment, there were a large number of inclusions with a size distribution in the range of 3–6 μm, resulting in a larger average diameter of inclusions. The proportion of inclusions larger than 6 μm in the two deoxidation experiments was very small, and most of the inclusions were distributed in the range of 1–4 μm.

4. Discussion

4.1. Effect of Deoxidation Processes on Oxide Inclusions in Steel

In molten steel, the reaction of metal elements with oxygen could generally be expressed as [22]:

$$x\mathrm{M}+y\mathrm{O}={\mathrm{M}}_x{\mathrm{O}}_y(\mathrm{s})$$

(1)

$$K_{\mathrm{M}-\mathrm{O}}={\displaystyle \frac{a_{{\mathrm{M}}_x{\mathrm{O}}_y}}{a_{\mathrm{M}}^x\cdot a_{\mathrm{O}}^y}}=\exp (-{\displaystyle \frac{\Delta G_{\mathrm{M}-\mathrm{O}}^{\mathrm{\theta }}}{\mathrm{R}T}})$$

(2)

where M represents dissolved Si, Al, and Mn in molten steel; $a_{\mathrm{M}}$ and $a_{\mathrm{O}}$ are the activity of M and O, referred to as a diluted solution of 1 mass pct in a standard state; $a_{{\mathrm{M}}_{\mathrm{x}}{\mathrm{O}}_{\mathrm{y}}}$ is the activity of deoxidized product referred to as a pure material in a standard state, with a value of 1; $\Delta G_{\mathrm{M}-\mathrm{O}}^{\mathrm{\theta }}$ represents the change in standard Gibbs free energy, J·mol⁻¹.

The activity of the main element $a_i$ in the molten steel, and the activity coefficient $f_i$ could be expressed as:

$$a_i=f_i\cdot [\%i]$$

(3)

$$\lg f_i={\displaystyle \sum e_i^j}\cdot [\%j]$$

(4)

where $f_i$ represents the activity coefficient of element i, 1; $e_i^j$ is the interaction coefficient of j on i, 1; [%i] and [%j] represent the mass fraction of i and j, respectively, %.

The deoxidation reactions of the traditional deoxidizers Al, Si, and Mn in actual steel production could be described by Equation (5) to Equation (7) [23,24,25,26]. In accordance with the thermodynamic data provided in the literature [27,28,29,30] and the steel compositions in Table 2 and Equation (4), the activity coefficient of the main elements in molten steel can be calculated as shown in

Table 3. The activity coefficient of components at 1873 K.

fC	fSi	fMn	fS	fAl	fBi	fO
1.03	1.09	0.95	0.93	0.91	0.61	0.77

.

$$2[\mathrm{A}\mathrm{l}]+3[\mathrm{O}]=\mathrm{A}{\mathrm{l}}_2{\mathrm{O}}_3(\mathrm{s}) \Delta G^{\mathrm{\theta }}=-\mathrm{1,202,000}+386.3T$$

(5)

$$[\mathrm{S}\mathrm{i}]+2[\mathrm{O}]=\mathrm{S}\mathrm{i}{\mathrm{O}}_2(\mathrm{s}) \Delta G^{\mathrm{\theta }}=-\mathrm{581,900}+221.8T$$

(6)

$$[\mathrm{Mn}]+[\mathrm{O}]=\mathrm{MnO}(\mathrm{s}) \Delta G^{\mathrm{\theta }}=-\mathrm{288,100}+128.3T$$

(7)

Substituting Equation (3) into Equation (2) obtains Equation (8). It is assumed that the alloy reacts completely with oxygen to form a simple oxide. With the mass fraction of the alloy as the horizontal coordinate and the mass fraction of oxygen as the vertical coordinate, the equilibrium curve of the alloy element and oxygen at 1873 K can be obtained, as shown in Figure 7. According to Figure 7, it can be seen that once the mass fraction of Al reaches 0.01%, the mass fraction of O decreases to 0.001%. When the mass fraction of Si is 0.2%, the mass fraction of O is 0.0132%. However, the equilibrium O mass fraction is as high as 0.0698% when Mn mass fraction is about 0.91%. It is attributed to the limited deoxidation capacity of Mn, and similar views have been reported in previous studies [15].

$${[\%\mathrm{M}]}^x\cdot {[\%\mathrm{O}]}^y=1/\left[\exp (-{\displaystyle \frac{\Delta G_{\mathrm{M}-\mathrm{O}}^{\mathrm{\theta }}}{\mathrm{R}T}})\cdot f_{\mathrm{M}}^x\cdot f_{\mathrm{O}}^y\right]$$

(8)

In the H1 experiment, after adding Si-Fe alloy to molten steel, Al₂O₃ and SiO₂ inclusions precipitated separately in the steel. Since the Si-Fe alloy added to the molten steel contained some aluminum, it reacted with dissolved oxygen to form Al₂O₃. Then, adding manganese, the SiO₂-MnO inclusions, Al₂O₃-MnO galaxite inclusions, and Al₂O₃-SiO₂-MnO inclusions were formed. The generation reactions of these inclusions are shown as Equation (9) to Equation (11) [31,32,33,34]. In addition, the Al₂O₃-MnO + SiO₂-MnO composite inclusion was formed. In the H2 experiment, the oxygen content in the steel decreased very little when manganese was first added to the molten steel. There were no separately precipitated MnO inclusions found in the steel specimen. When silicon was delivered into the molten steel, spherical or ellipsoidal Al₂O₃-MnO inclusions, spherical Al₂O₃-SiO₂-MnO inclusions, and large elliptical or fusiform Al₂O₃-MnO + Al₂O₃-SiO₂-MnO composite inclusions were formed.

$$2{\mathrm{MnO}\left(\mathrm{s}\right)+\mathrm{SiO}}_2\left(\mathrm{s}\right)=(2\mathrm{M}\mathrm{n}\mathrm{O}\cdot \mathrm{S}\mathrm{i}{\mathrm{O}}_2)$$

(9)

$$\mathrm{MnO}(\mathrm{s})+{\mathrm{Al}}_2{\mathrm{O}}_3(\mathrm{s})=\mathrm{M}\mathrm{n}\mathrm{O}\cdot {\mathrm{Al}}_2{\mathrm{O}}_3$$

(10)

$$x{\mathrm{Al}}_2{\mathrm{O}}_3(\mathrm{s})+y{\mathrm{SiO}}_2(\mathrm{s})+z\mathrm{MnO}(\mathrm{s})=x{\mathrm{Al}}_2{\mathrm{O}}_3\cdot y{\mathrm{SiO}}_2\cdot z\mathrm{M}\mathrm{n}\mathrm{O}$$

(11)

The Factsage 8.2 software was occupied to calculate the phase equilibrium of the Al₂O₃-SiO₂-MnO ternary system, as shown in Figure 8. It was worth noting that the stable regions of each phase were indicated with solid black lines, and the liquidus at 1873 K was marked with solid red lines in Figure 8. The solid red dots which represent components of oxide inclusions detected in the samples with two deoxidation experiments were mapped to the ternary phase diagram. Since the EDS analysis of oxygen content in the inclusions was uncertain, metallic elements were applied to determine the compositions of inclusions. It can be seen that most of the inclusions are distributed in the liquid phase region of 1873 K in H1. The component points of the inclusions fall in the phase stable regions of 2MnO∙SiO₂ and 3MnO∙Al₂O₃∙3SiO₂, indicating that these inclusions appear as liquid at steelmaking temperature. The interfacial tension between these liquid inclusions and the molten steel is relatively small, making them prone to collision and floating. However, thanks to the short holding time, the liquid inclusions failed to float and remove themselves in time and so remained in the steel [17,18]. This may be one of the reasons for the high inclusion quantity and density in the H1 experiment. In the H2 experiment, most inclusions are in the liquid phase at steelmaking temperature and have component points near the MnO-rich region. There are some Al₂O₃-MnO inclusions with high Al₂O₃ content, which are out of the liquid-phase region of 1873 K, indicating that their melting points are relatively high. From the above results, it can be seen that deoxidation processes have a great influence on the oxide inclusion composition in environmentally friendly free-cutting steel.

4.2. Effect of Deoxidation Processes on Non-Oxide Inclusions in Steel

After a scanning electron microscope equipped with energy-dispersive spectroscopy (SEM-EDS) was used for detection, the two experimental steels were found to also include manganese sulfide inclusions, bismuth particles, and Bi-MnS inclusions (as shown in Figure 4 and Figure 5). Manganese sulfide inclusions always precipitate in the later period of steel solidification. The generation reactions of manganese sulfide inclusions are shown in Equation (12) [35].

$$[\mathrm{M}\mathrm{n}]+[\mathrm{S}]=\mathrm{M}\mathrm{n}\mathrm{S}(\mathrm{s})$$

(12)

The spheroidization degree of MnS inclusions is one of the most important factors affecting the cutting performance of steel, and the spheroidization degree of inclusions can be expressed by the ratio of the maximum side length to the minimum side length, called the aspect ratio. The expression of aspect ratio is shown in Equation (13).

$$\mathrm{Aspect}\mathrm{ratio}={\displaystyle \frac{d_{\max }}{d_{\min }}}$$

(13)

The aspect ratio of MnS inclusions is high, and large size MnS inclusions have the risk of forming a rod-like form more easily [36,37]. The closer the aspect ratio is to 1, the higher the spheroidization degree of MnS inclusions. The aspect ratio of the sulfide inclusions in the two deoxidation furnaces was calculated, and a violin plot with a box was drawn according to the statistical results, as shown in Figure 9. The range of the box is between 25% and 75%. It is worth noting that the median is equal to the arithmetic average of the third quartile and first quartile, the value of which was represented by a black font. The larger the external shape area of the violin map, the denser the distribution of data points, whereas the interquartile (IQR) is a reflection of the central tendency. It appears that the average value and median of the aspect ratio in the H1 experiment are larger than those in the H2 experiment. The number of inclusions increased significantly in the range of aspect ratios 1.5–2.0 in the H1 experiment. It could be concluded that the spherodization degree of MnS inclusions in the steel with first manganese addition followed by Si-Fe alloy addition is relatively well.

5. Conclusions

The characteristics of inclusions in environmentally friendly free-cutting steel under different deoxidation processes were studied by laboratory experiments and thermodynamic calculations. The findings are summarized as follows:

• The oxygen content exhibits a similar decreasing trend in both deoxidation processes, and there is no significant difference in total oxygen content between the final samples, indicating comparable deoxidation efficiencies for the two methods;
• In the process where the Si-Fe alloy is added first, followed by manganese, the inclusions in the molten steel are predominantly liquid-phase Al₂O₃-SiO₂-MnO composite inclusions. In contrast, when manganese is added first, followed by the Si-Fe alloy, in addition to liquid-phase Al₂O₃-SiO₂-MnO composite inclusions in the MnO rich region, there are also some solid-phase Al₂O₃-MnO inclusions with higher Al₂O₃ content;
• In both deoxidation furnaces, Bi particles are mainly present adhering to MnS inclusions. The process involving adding the Si-Fe alloy followed by adding manganese results in higher average and median aspect ratios. In the process where manganese is added first followed by the Si-Fe alloy, the spheroidization degree of MnS inclusions is relatively better.