The Research on Carbon Deoxygenation of Molten Steel and Its Application in the Converter Steelmaking Process

Abstract

At the steelmaking temperature, carbon has a strong deoxidation ability. Under the vacuum condition, its deoxidation ability can be further improved, and it can become a stronger deoxidation element than aluminum. The product of carbon deoxygenation is CO, which floats up and detaches from the molten steel in the form of bubbles and does not produce oxide inclusions. Under normal pressure, replacing aluminum with carbon to complete partial deoxidation tasks can not only reduce the generation of inclusions and alleviate the pressure of removing inclusions, but also reduce the consumption of aluminum and save deoxidation costs. In this study, the carbon deoxidation process after the converter was investigated. Firstly, the timing of carbon addition was determined through thermodynamic calculations, and it was found that, in oxygen-enriched molten steel, the priority of the reaction of the deoxidation element was [Al] > [Si] > [C] > [Mn]. Through the carbon and oxygen balance calculation, it is known that the carbon deoxidation effect is greatly affected by the carbon content of the molten steel; for low-carbon steel, carbon can be used for pre-deoxygenation, whereas for medium-carbon and high-carbon steel, carbon can complete most of the deoxidation tasks. Finally, with 45 steel as the research object, the carbon deoxidation process was designed and tested in industry. The results showed that, compared with the aluminum deoxidation process, the number of inclusions in the billet casting of the carbon deoxidation process was reduced by 68.8%, and the carbon deoxidation process had fewer large-sized inclusions in the billet casting. In addition, the carbon deoxidation process uses carbon powder instead of the aluminum block for deoxidation during steel tapping from the converter. The deoxidant cost is reduced by CNY 15.47/ton of steel. From a comprehensive point of view, the application of carbon deoxidation after the converter can reduce aluminum consumption and improve the cleanliness of steel, which is an important way for enterprises to reduce costs and increase efficiency.

1. Introduction

As a representative steel grade in high-quality carbon structural steel, 45 steel has a wide range of applications and high demand. It is an indispensable basic material in the fields of industry, agriculture, and national defense construction [1]. Compared with some high-performance alloy steels, 45 steel is a little inferior in some special properties (e.g., corrosion resistance), but it is favored for its economy and versatility. However, with the increasingly fierce market competition and the upgrading of application fields, the quality requirements for 45 steel are becoming higher and higher, and steel enterprises are facing increasing cost pressure. Therefore, many enterprises are committed to improving their production technology to enhance their market competitiveness.

The domestic production of 45 steel mainly adopts the aluminum deoxidation process. Aluminum can quickly reduce dissolved oxygen in steel to a lower level, but it will generate a large number of Al₂O₃ inclusions, which are mostly spherical, point-like, or clustered in morphology [2,3,4]. Al₂O₃ is a high-melting-point and non-deformable inclusion. During the continuous casting process, it is easy to accumulate on the inner wall of the nozzle, causing its blockage [5,6]. In addition, during the rolling process, Al₂O₃ inclusions can cause scratches in the substrate and are prone to cause stress concentrations, which can lead to the formation of voids or cracks, affecting the fatigue life of the material [7,8,9]. In response to the many problems caused by aluminum inclusions, many technologies have been introduced in ladle refining, including ladle bottom blowing [10,11], refining slag modification [12,13,14], and calcium treatment [5,15,16]. Among them, calcium treatment is used to improve the blockage problem at the nozzle. Calcium can react with Al₂O₃ to form calcium aluminate with a lower melting point, making it less likely to accumulate at the nozzle. However, the actual effect of calcium treatment is influenced by many factors. Improper calcium treatment not only fails to achieve the desired effect but also leads to an increase in inclusions in the molten steel.

In addition to the treatment of impurities caused by aluminum deoxidation, aluminum is a high-energy-consuming product with high carbon emissions during its production process. Therefore, many scholars have explored clean purification deoxidation methods. The product of carbon deoxidation is CO, which will detach from the molten steel system in the form of bubbles [17], without producing oxide inclusions. Watanabe [18] carried out vacuum carbon deoxidation experiments through a 15 kg vacuum induction furnace, and the oxygen content of the steel could be reduced from 400 × 10⁻⁶ to 7~8 × 10⁻⁶ in 90 min. Xue [19,20] et al. used a 25 kg vacuum induction furnace for vacuum carbon deoxidation, which can also reduce oxygen in steel to below 10 × 10⁻⁶, and even below 5 × 10⁻⁶. In addition, some scholars have adopted vacuum carbon deoxidation technology for industrial production. Li [21] proposed the treatment of low silicon rotor steel with “aluminum pre-deoxidation + vacuum carbon deoxidation”, which stably produced rotor steel with oxygen content less than 30 × 10⁻⁶. Dong [22] adopted the vacuum carbon deoxidation smelting and vacuum casting process to produce high-purity large steel ingots with T[O] ≤ 15 × 10⁻⁶. Ba [23] adopted a combination of vacuum carbon deoxidation and aluminum deoxidation, which reduces the oxygen content of steel ingots by about 17.6% based on the original silicon and aluminum deoxidation process. Xin [24] used a vacuum carbon deoxidation process to produce 17CrNiMo6 high-strength gear steel, and the final product showed significant improvements in gas content, number of inclusions, and mechanical properties. In summary, research on carbon deoxidation is mostly focused on vacuum conditions, with relatively less research on atmospheric carbon deoxidation. This is because, compared with the aluminum deoxidation process, its safety, efficiency, and universality are slightly insufficient. However, with the continuous advancement of metallurgical technology and equipment, atmospheric carbon deoxidation can be combined with other deoxidation methods to achieve the target deoxidation goal.

The authors of this study explored and applied the carbon deoxidation process after the converter. During the steel tapping from the converter, if deoxidation can be achieved through carbon, it can not only reduce the generation of inclusions from the source but also reduce the consumption of ferroalloys. In this study, the theoretical feasibility of carbon deoxidation was first investigated. Then, taking 45 steel as the research object, the carbon deoxidation process after the converter was designed, and the smelting effect was quantitatively characterized by sampling in the whole process, which provides a reference for the actual production.

2. Experimental Materials and Methods

2.1. Experimental Materials

The production process of 45 steel is as follows: BOF–LF–CC. The steel composition requirements for each process node are shown in

Table 1. Requirements for the composition of molten steel at each node/wt%.

Process Node		C	Si	Mn	P	S	Alt
BOF smelting endpoint	lower limit	0.08	-	-	-	-	-
	upper limit	0.15	-	-	0.020	-	-
Before LF refining begins	lower limit	0.35	0.10	0.45	-	-	-
	upper limit	0.40	0.20	0.55	-	-	-
Finished product	target value	0.46	0.22	0.62	≤0.025	≤0.012	0.010

.

According to Table 1, at the end of BOF smelting, the carbon content of the molten steel is required to be between 0.08% and 0.15%, and before LF refining begins, the carbon content of the molten steel is required to be between 0.35% and 0.40%. This indicates that a large amount of carburant is added when steel tapping from the converter. The proximate analysis of the carburant is listed in

Table 2. Proximate analysis of carburant on an air-dry basis/wt%.

Testing Items	Fixed Carbon	Ash	Volatile	Moisture	S
Mass fraction/wt%	92.14	5.68	1.62	0.34	0.22

. The fixed carbon, ash, volatile, and moisture are obtained according to GB/T 2001-2013 [25]. The S element content is obtained according to GB/T 2286-2017 [26].

2.2. Experimental Methods

2.2.1. Deoxygenation Process and Sampling Plan

The aluminum deoxidation process involves adding the aluminum block to the ladle for deoxidation during steel tapping from the converter, and then adding carburant and Si-Mn alloy for alloying. When the weight of the poured molten steel reaches about 3/4, lime, refined slag, and fluorite are added for slag-making. After entering the ladle refining, calcium carbide, silicon carbide, and aluminum particles are added for diffusion deoxidation. When the composition of the molten steel reaches the expected range, calcium wire is added to the molten steel. Different from the aluminum deoxidation process, the key to the carbon deoxidation process is to exert the deoxidation effect of carbon. The specific process is shown in Figure 1.

As shown in Figure 1, the carbon deoxygenation process has several key operations:

① During steel tapping from the converter, the carburant is evenly added to the ladle by manual throwing after the molten steel covers the bottom of the ladle.

② When the steel tapping is about to end, lime and refined synthetic slag are added to the ladle to make the initial refined slag. By adjusting the ratio of lime and refined synthetic slag, the composition of refined slag is adjusted to be consistent with the aluminum deoxidation process. At the same time, increasing the bottom blowing intensity of the ladle promotes deoxidation.

③ After the ladle enters the LF station, reducing agents such as calcium carbide, silicon carbide, and aluminum particles are added to the surface of the refining slag for diffusion deoxidation.

④ After the modification of refined slag is completed, reducing slag is obtained. Then, Si-Mn alloy is added to the ladle for alloying.

⑤ After the composition of the molten steel is qualified, calcium wire is added to the molten steel, and soft blowing of the ladle is carried out. After the soft blowing, the ladle is lifted to the turntable for casting.

To investigate the changes in the cleanliness of molten steel during the carbon deoxidation process, steel samples were taken at some key nodes. The sampling method and sampling nodes are shown in Figure 2.

According to Figure 2, steel samples were taken at the end of the converter, after carbon deoxidation, before the start of ladle refining, after LF diffusion deoxidation, before adding calcium wire, after soft blowing of the ladle, in the tundish, and at the cast billet. At the end of the converter smelting, a bucket-shaped sampler was used for sampling, and it was found that the molten steel was boiling and surging out of the steel mold, making it impossible to obtain dense steel ingots. An attempt was also made using a racket sampler, but dense samples could not be obtained. Before the start of ladle refining, there are many pores in the steel sample, indicating that the carbon–oxygen reaction is still ongoing, and there are many bubbles in the molten steel. After LF diffusion deoxidation, the steel sample is dense, indicating that the carbon–oxygen reaction is already very weak.

2.2.2. Sample Testing Method

The inert gas melting infrared absorption method was used to detect oxygen and nitrogen content in steel. The ASPEX automatic scanning electron microscope is composed of Aztec Steel software (Carl Zeiss Microscopy Ltd., Cambridge, Cambridgeshire, UK) and X-Max spectrometer hardware (Oxford Instruments Nanotechnology Tools Ltd., High Wycombe, Buckinghamshire, UK). It was employed to perform inclusion statistics on metallographic samples with a size of 10 mm × 10 mm × 10 mm, automatically scanning the number of non-metallic inclusions with a size greater than or equal to 1 μm, with a scanning area of approximately 32.9 mm². Finally, SEM-EDS was used to observe the morphology of inclusions and analyze their composition.

3. Results and Discussion

3.1. Theoretical Calculation of the Carbon Deoxygenation Process

3.1.1. Analysis of the Deoxygenation Ability of Different Elements

The deoxidation elements commonly used for deoxidation after the converter include Al, Si, Mn, and C. The equation for their reaction with dissolved oxygen in molten steel and the variation value of the standard Gibbs free energy calculation formula [27] are shown in

Table 3. Reaction equation and the variation value of the standard Gibbs free energy calculation formula.

Deoxidation Element	Reaction Equation	Variation Value of Standard Gibbs Free Energy (J/mol)
Al	$2\left[\mathrm{A}\mathrm{l}\right]+3\left[\mathrm{O}\right]=\mathrm{A}{\mathrm{l}}_2{\mathrm{O}}_3$	$\mathsf{\Delta }{G}^{\mathsf{\theta }}=-1202910+385.815T$
Si	$\left[\mathrm{S}\mathrm{i}\right]+2\left[\mathrm{O}\right]=\mathrm{S}\mathrm{i}{\mathrm{O}}_2$	$\mathsf{\Delta }{G}^{\mathsf{\theta }}=-574350+218.83T$
Mn	$\left[\mathrm{M}\mathrm{n}\right]+\left[\mathrm{O}\right]=\mathrm{M}\mathrm{n}\mathrm{O}$	$\mathsf{\Delta }{G}^{\mathsf{\theta }}=-288805+126.835T$
C	$\left[\mathrm{C}\right]+\left[\mathrm{O}\right]=\mathrm{C}\mathrm{O}$	$\mathsf{\Delta }{G}^{\mathsf{\theta }}=-21755-38.74T$

.

The variation value of the standard Gibbs free energy of different deoxygenated element was calculated at different temperatures, and the calculation results are shown in Figure 3a. The variation value of the standard Gibbs free energy of the reaction of Al, Si, and Mn with dissolved oxygen shows an increasing trend with increasing temperature, whereas that of the reaction of carbon with dissolved oxygen shows a decreasing trend with increasing temperature. Thermodynamically, when the temperature increases to a certain value, [C] will have the ability to reduce the oxides of Si and Al. However, at 1873~1923 K, the priority of elements reacting with dissolved oxygen can be described as [Al] > [Si] > [C] > [Mn].

Carbon can reduce Al₂O₃ under certain conditions with the following reaction expression:

$$3\left[\mathrm{C}\right]+\mathrm{A}{\mathrm{l}}_2{\mathrm{O}}_3=2[\mathrm{A}\mathrm{l}]+3\mathrm{C}\mathrm{O}\hspace{1em}\mathsf{\Delta }{G}^{\mathsf{\theta }}=1137645-502.035T$$

(1)

$$\hspace{1em}\mathsf{\Delta }G=\mathsf{\Delta }{G}^{\mathsf{\theta }}+RT\ln {\displaystyle \frac{P_{CO}\cdot a_{[Al]}^2}{a_{[C]}^3}}$$

(2)

Reaction (1) proceeds positively when $\mathsf{\Delta }G<0$. The reaction temperature of the system was set to be 1873 K. The critical conditions for the reaction were calculated as shown in Figure 3b. At standard atmospheric pressure, when the [Al] activity in the molten steel is 0.005, the [C] activity in the molten steel must be greater than 2 to meet the thermodynamic conditions for reducing Al₂O₃. When the activities of [C] and [Al] remain constant, reducing the CO partial pressure can promote Reaction (1). However, even under high vacuum conditions, the reduction of Al₂O₃ by [C] is relatively small, which is mainly limited by the kinetic conditions. Therefore, when carbon deoxidation is carried out under standard atmospheric pressure, [C] cannot reduce Al₂O₃. If carbon is added to the molten steel after Al, the dissolved oxygen in the molten steel will preferentially combine with Al to form Al₂O₃, whereas carbon cannot achieve a deoxidation effect and can only increase the carbon content of the molten steel. The alloying process was adjusted in this study based on the above theoretical calculation results. During the steel tapping from the converter, a sufficient amount of carburizer is added for deoxidation and to increase the carbon content of the molten steel, and then the alloying of other elements is transferred to the LF station.

3.1.2. Limit Analysis of Carbon Deoxygenation Reaction

According to the carbon–oxygen reaction equation in Table 1, the equilibrium constant K is calculated as follows:

$$K={\displaystyle \frac{P_{CO}}{a_C\cdot a_O}}={\displaystyle \frac{P_{CO}}{f_C\cdot f_O\cdot w_{{[C]}_{\%}}\cdot w_{{[O]}_{\%}}}}$$

(3)

In the formula, $P_{CO}$ represents the dimensionless partial pressure; $a_C$ represents the activity of [C] in molten steel; $a_O$ represents the activity of [O] in molten steel; $w_{{[C]}_{\%}}$ represents the mass fraction of [C] in molten steel; $w_{{[O]}_{\%}}$ represents the mass fraction of [O] in molten steel; $f_C$ represents the activity coefficient of [C] in molten steel; and $f_O$ represents the activity coefficient of [O] in molten steel.

When the mass fraction of carbon in the molten steel is 0.02~2 wt%, the value of $f_C\cdot f_O$ is close to 1 [28]. The expression can be simplified into Formula (4):

$$K={\displaystyle \frac{P_{CO}}{w_{{[C]}_{\%}}\cdot w_{{[O]}_{\%}}}}$$

(4)

The relationship between temperature and the equilibrium constant K is shown in Formula (5):

$$\lg K={\displaystyle \frac{1168}{T}}+2.07$$

(5)

When the temperature of the molten steel is 1873 K, the carbon–oxygen balance relationship in the molten steel is shown in Formula (6):

$$w_{{[C]}_{\%}}\cdot w_{{[O]}_{\%}}=0.002025P_{CO}$$

(6)

Similarly, assuming $f_{Al}=f_{Si}=f_O=f_{Mn}=1$ [28] in the molten steel, the equilibrium relationship between [Al], [Si], [Mn], and [O] can be obtained as shown in Equations (7)–(9):

$$w_{{[Al]}_{\%}}\cdot w_{{[O]}_{\%}}^{1.5}=2.01\times {10}^{-7}$$

(7)

$$w_{{[Si]}_{\%}}\cdot w_{{[O]}_{\%}}^2=2.59\times {10}^{-5}$$

(8)

$$w_{{[Mn]}_{\%}}\cdot w_{{[O]}_{\%}}=0.03725$$

(9)

The thermodynamic equilibrium curves of dissolved oxygen with deoxygenated elements such as carbon, aluminum, silicon, and manganese at 1873 K are shown in Figure 4a, and the carbon–oxygen equilibrium curves at different vacuum degrees are shown in Figure 4b.

When P_CO = 1, as the carbon content in the molten steel increases, the equilibrium oxygen content decreases. When the carbon content in the molten steel is 0.1%, the equilibrium oxygen content is 230 × 10⁻⁶. When the carbon content in the molten steel is 0.5%, the equilibrium oxygen content is 46 × 10⁻⁶. When the carbon content in the molten steel is 1.0%, the equilibrium oxygen content is 23 × 10⁻⁶. Therefore, carbon has a strong deoxidation ability under normal pressure, but the deoxidation effect is greatly affected by the carbon content of the molten steel. In summary, carbon can be used for the pre-deoxidation of low-carbon steel, while for medium-carbon and high-carbon steel, carbon can complete the majority of deoxidation tasks. However, for medium-carbon steel, although the oxygen content can theoretically be reduced to around 50 × 10⁻⁶, its oxygen content still exceeds the target range of the product. This indicates that after carbon deoxidation, other deoxidation methods still need to be used to complete the final deoxidation.

Vacuum conditions can improve the deoxygenation ability of carbon. The deoxidation ability of carbon increases with the decrease in vacuum degree. When P_CO = 0.01, the deoxygenation ability of carbon is almost the same as that of aluminum, whereas when P_CO = 0.001, carbon’s deoxygenation ability is stronger. When a vacuum treatment system is available, the carbon deoxidation process can meet the cleanliness requirements of most steel grades, including low-carbon, medium-carbon, and high-carbon steel.

3.1.3. The Allocation Pathway of Carbon in the Carbon Deoxygenation Process

Without considering burning loss, there are two ways to utilize carbon after it is added to oxygen-enriched molten steel: carbonization and deoxidation. The authors of this study utilized a thermodynamic software, named FactSage 7.2, selected the Equilib module, selected the FTmisc and FactPS databases, and used the open module for carbon–oxygen balance calculation. The system temperature was set to 1873 K, the initial mass of molten steel to 100 g, the initial oxygen content to 500 × 10⁻⁶, and the initial carbon content to 0. In total, 0.01 g of carbon was added to the calculation system each time, it was cycled 100 times, and the carbon and oxygen content of the molten steel and the mass of CO produced after each carbon addition were calculated. The changes are shown in Figure 5a, and the proportion of carbon used for carbonization and deoxidation is shown in Figure 5b.

For oxygen-enriched molten steel, there is a brief period of constant oxygen content in the molten steel with the addition of carbon, during which all the carbon added is used for carbonization. When the amount of carbon added exceeds 0.03 g, the oxygen content in the molten steel begins to decrease, CO begins to be produced, and the growth rate of carbon content in the molten steel decreases. At this stage, the added carbon participates in both deoxidation and carbonization. With the addition of carbon, the proportion of its mass used for carbonization and deoxidation shows a stage change. In the initial stage, oxygen-enriched molten steel does not contain carbon, and all the added carbon is used to increase the carbon content. When the carbon and oxygen in the molten steel reach equilibrium, carbon is added again, and CO begins to be produced. The maximum proportion of carbon added for deoxidation can reach 43.83%. Afterwards, as the oxygen content in the molten steel decreases, the proportion of added carbon participating in deoxidation becomes smaller and smaller. When the weight of carbon added reaches 1 g, the average mass proportion used for carbonization is 96.42%, and the average mass proportion used for deoxidation of molten steel is 3.58%.

In addition, from Figure 5a, it can be seen that the carbon deoxidation rate is affected by the dissolved oxygen content in the molten steel. As the dissolved oxygen content in the molten steel decreases, the carbon deoxidation rate slows down. Therefore, in actual production, measures need to be taken in the later stage of carbon deoxidation to promote the deoxidation effect. In the process design of this study, after converter tapping, the bottom blowing flow of the ladle is increased to promote deoxidation. After the ladle enters the LF station, the refining slag is modified to promote deoxidation.

3.2. Application Effect of the Carbon Deoxidation Process in 45 Steel

3.2.1. Changes in Oxygen and Nitrogen Content in Steel

Due to the presence of many pores inside the steel samples obtained at nodes such as the end of converter smelting (B-1), after carbon deoxidation (B-2), and before LF refining begins(L-1), oxygen and nitrogen content detection could not be performed. The changes in oxygen and nitrogen content in the steel at subsequent nodes are shown in Figure 6.

As shown in Figure 6a, the oxygen content shows a significant decreasing trend. After diffusion deoxidation, the total oxygen content in the steel decreases to 49 × 10⁻⁶. Before adding calcium wire, the total oxygen content in the steel decreases to 19.3 × 10⁻⁶, and the final total oxygen content of the cast billet is 21.5 × 10⁻⁶. The nitrogen content in steel shows a continuous upward trend. After LF diffusion deoxidation, the nitrogen content is 23.9 × 10⁻⁶, and the nitrogen content in the cast billet increases to 55.1 × 10⁻⁶. Comparing the cast billet produced by the carbon deoxidation process with that of the aluminum deoxidation process, as shown in Figure 6b, the total oxygen content of the cast billet produced by the former is 21.5 × 10⁻⁶, and the total oxygen content of the cast billet produced by the latter is 25 × 10⁻⁶. In addition, the nitrogen content of the carbon deoxidation process cast billet is 55.1 × 10⁻⁶, whereas that of the aluminum deoxidation process cast billet is 69 × 10⁻⁶. Based on the comprehensive oxygen and nitrogen content, it can be concluded that the carbon deoxidation process has certain advantages in both deoxidation and nitrogen control. This is because many CO bubbles are generated inside the molten steel during the process of carbon deoxidation, which has a certain denitrification effect. In addition, during diffusion deoxidation, oxygen elements in the molten steel cross the steel slag interface and enter the slag for reduction. As the free oxygen content in the molten steel decreases, some alloy elements enter the molten steel through refining slag and react with residual oxygen in the molten steel to produce inclusions, which are concentrated in the middle and upper parts of the ladle and are more likely to float up. This may be the reason why deoxidation has certain advantages.

3.2.2. Study on the Morphology and Evolution of Inclusions in the Carbon Deoxidation Process

To investigate the morphology and elemental distribution of inclusions during the carbon deoxidation process, SEM-EDS was used to observe the metallographic samples of different sampling nodes, as shown in Figure 7. After LF diffusion deoxidation, the inclusions in the molten steel appear as clusters with distinct edges. According to the surface scan results, their main elements are Al and O. Before adding calcium wire, the inclusions appear approximately circular, and the surface scan results show that their main elements are Al, Ca, and O, with a small amount of Mg. At the end of soft blowing and inside the tundish, the inclusions in the molten steel appear nearly circular, and the surface scan results show that their main elements are Al, Ca, and O, with a small amount of Mg or Si. In the cast billet, the inclusions are also approximately circular, and the surface scan results show that their main elements are Al, Ca, and O. Additionally, a layer of CaS is formed around the inclusions, which is generated during the cooling process of the cast billet.

To investigate the compositional evolution of inclusions, EDS spectroscopy was used to detect inclusions, and their composition was mapped onto a ternary phase diagram, as shown in Figure 8. After LF diffusion deoxidation, the inclusions in the molten steel are mainly composed of Al₂O₃. Before adding calcium wire, the inclusions were still mainly Al₂O₃, but the content of MgO and CaO in the inclusions slightly increased. After adding calcium wire, the mass proportion of CaO in the inclusions significantly increased, forming a composite inclusion mainly composed of calcium aluminate and containing a small amount of MgO and SiO₂. The melting point of the inclusion moved to the low-melting-point zone. Therefore, in the carbon deoxidation process, the evolution of inclusions can be summarized as Al-(Ca)-(Mg)-O→(S)-Ca-Al-(Mg)-(Si)-O. Overall, the evolution of inclusion types in carbon deoxidation processes shows similarities with many reports on aluminum deoxidation processes [29].

To explain the reasons for the evolution of inclusions, FactSage7.2 was used to calculate the phase diagram of the inclusion precipitation advantage zone and the effect of calcium treatment on the inclusion type under the “carbon deoxidation” mode. The steel composition of key nodes in the carbon deoxidation process is shown in

Table 4. Composition of molten steel at key nodes of the carbon deoxygenation process.

Composition of Molten Steel	C/wt%	Si/wt%	Mn/wt%	Al/wt%	O/×10−6
Smelting endpoint of the converter	0.06	0.025	0.09	0.0038	432
After LF diffusion deoxygenation	0.34	0.025	0.09	0.0128	49
Before adding calcium wire	0.44	0.250	0.60	0.0140	19.3

.

The phase diagram of the inclusion precipitation advantage zone was calculated using the composition of molten steel at each key node as the initial calculation condition, with T = 1600 °C, an oxygen content ranging from 0 to 500 × 10⁻⁶, and an Al content ranging from 0 to 300 × 10⁻⁶. The effect of different amounts of calcium addition on the transformation of inclusion types in the molten steel was also calculated using the composition of the molten steel before adding calcium wire as the initial condition, with a Ca content range of 0~30 × 10⁻⁶. The calculation results are shown in Figure 9.

According to Figure 9a, the phase diagram can be divided into two regions: a “high oxygen and low aluminum” region and a “high aluminum and low oxygen” region. In the “high oxygen and low aluminum” region, slag-type inclusions are preferentially precipitated, and these are mainly Al-Si-Mn-O-type low-melting-point inclusions. In the “high aluminum and low oxygen” region, M₂O₃ inclusions preferentially precipitate, and these are high-melting-point Al₂O₃ inclusions. At the end of the converter smelting process, the dissolved oxygen content in the steel is 432 × 10⁻⁶, and the residual content of Al, Si, and Mn elements in the steel is very low. This molten steel condition falls in the “high oxygen and low aluminum” region. Therefore, some oxygen will form a small number of Al-Si-Mn-O-type inclusions with residual elements such as Al, Si, and Mn, while the remaining oxygen elements still exist in the form of [O]. As carbon deoxidation progresses, the dissolved oxygen content in the molten steel decreases. After the ladle enters the LF station, diffusion deoxidation is carried out by adding calcium carbide, silicon carbide, and aluminum particles to the slag surface. The added aluminum particles not only quickly adjust the oxidation of the slag but also partially penetrate the slag layer and enter the molten steel. According to Figure 9b,c, after LF diffusion deoxidation, the dissolved oxygen in the steel decreases to 49 × 10⁻⁶, and the Al content in the molten steel reaches 128 × 10⁻⁶. This molten steel condition falls in the “high aluminum and low oxygen” region, which meets the precipitation conditions for Al₂O₃. At the same time, aluminum elements will also promote the transformation of non-floating Al-Si-Mn-O inclusions into Al₂O₃. Therefore, after LF diffusion deoxidation and before adding calcium wire, the type of inclusions should be mainly Al₂O₃ inclusions, which is consistent with the actual detection results. This also indicates that, in the carbon deoxidation process, the final deoxidation in the molten steel is completed by the Al element. Figure 9d shows the effect of calcium content on the types of inclusions in molten steel. When the calcium content is less than 9 × 10⁻⁶, calcium can transform Al₂O₃ into the high-melting-point calcium aluminate salt, whereas when the calcium content is 9~16 × 10⁻⁶, calcium can transform Al₂O₃ into the low-melting-point calcium aluminate salt. When the calcium content is greater than 16 × 10⁻⁶, Ca₂SiO₄ begins to form, and when it is greater than 24 × 10⁻⁶, CaS begins to form. Therefore, the liquid window area of calcium treatment is 9~16 × 10⁻⁶.

3.2.3. Comparison of the Cleanliness of Cast Billets Produced by Different Deoxidation Processes

Due to the presence of many pores inside the steel samples obtained at nodes such as B-1, B-2, and L-1, inclusion scanning could not be performed. The number density and size of inclusions in steel at other nodes were statistically analyzed, and the results are shown in Figure 10.

According to Figure 10a, the number density of inclusions shows a trend of first decreasing and then increasing. After LF diffusion deoxidation, the number density of inclusions in the molten steel is 3.4/mm²; before adding calcium wire, it drops to 1.6/mm², and at the end of soft blowing, it drops to the lowest point, only 0.9/mm². The number density of inclusions in the tundish is 1.0/mm², whereas that in the cast billet slightly increases to 1.9/mm². According to Figure 10b, there are large-sized inclusions in the early stage of LF refining. In the later stage of LF refining, the number of large-sized inclusions decreases, and the size of inclusions in the cast billet is mainly concentrated below 5 μm, with no large-sized inclusions larger than 10 μm. To observe the effect of the carbon deoxidation process on inclusions, samples were taken at some key nodes of the aluminum deoxidation process. The comparison results of the two deoxidation processes are shown in Figure 11.

According to Figure 11a, it can be seen that the carbon deoxidation process has certain advantages in controlling inclusions. In terms of the number density of inclusions in the cast billet, the aluminum deoxidation process has an inclusion density of 6.1/mm², while the carbon deoxidation process has an inclusion density of only 1.9/mm², a decrease of 68.8%. According to Figure 11b, it can be seen that the size of inclusions in the cast billets produced by both processes is concentrated below 5 μm. Among them, the average size of inclusions in the carbon deoxidation process cast billet is larger. However, within the same scanning area, there are more large-sized inclusions with a diameter greater than 10 μm in the aluminum deoxidation process cast billet, while the size distribution of inclusions in carbon deoxidation process castings is relatively uniform. Taken together, the carbon deoxidation process demonstrates certain advantages in the control of the number of inclusions and the control of size uniformity.

3.2.4. Comparison of Deoxygenation Costs for Different Deoxygenation Processes

Aluminum is a high-cost, high-energy-consuming product. Eliminating the addition of aluminum blocks during converter steelmaking and using carbon for deoxidation helps to reduce deoxidation costs and carbon emissions. The types of deoxidants used in the two deoxidation processes include aluminum blocks, calcium carbide, silicon carbide, aluminum particles, and carbon powder. The consumption of deoxidants and the cost per ton of steel are shown in

Table 5. Cost of deoxidants for different processes.

Different Deoxidation Processes	Deoxidant						Cost Per Ton of Steel/CNY
	Types	Aluminum Blocks	Calcium Carbide	Silicon Carbide	Aluminum Particles	Carbon Powder
	Unit Price/(CNY/kg)	20.02	3.92	4.54	20.52	2.64
Aluminum deoxidation	Consumption/(kg/t)	0.75	1.38	0.50	0.25	4.89	40.74
Carbon deoxidation		0	0.59	1.18	0.29	4.12	24.74

.

According to the table, the cost of deoxidant for the aluminum deoxidation process is CNY 40.74/ton of steel, while that for the carbon deoxidation process is only CNY 24.74/ton of steel, a decrease of CNY 16/ton of steel. In summary, for 45 steel, the carbon deoxidation process is an effective cost reduction and efficiency improvement measure.

4. Conclusions

With the continuous advancement of metallurgical technology and equipment, the deoxidation process of converters has been significantly improved. The development and application of aluminum deoxidation technology have greatly promoted the growth and progress of the steelmaking industry, but the problem of inclusion control is also prominent. The authors of this study explore the feasibility of the carbon deoxidation process when steel is tapped from a converter. The effectiveness of the carbon deoxidation process is discussed using both theoretical calculations and industrial experiments. The research conclusions are as follows:

• At the steelmaking temperature, the priority of deoxidation element reaction in the molten steel is [Al] > [Si] > [C] > [Mn]. Therefore, when carbon deoxidation is carried out during steel tapping, carbon must be added to the molten steel before alloy elements such as Al and Si. Under normal pressure, the carbon deoxidation effect is greatly affected by the carbon content of the molten steel. Without vacuum treatment conditions, carbon deoxidation cannot be used as the final deoxidation method, and other such methods still need to be combined to complete the final deoxidation of molten steel.
• After reaching carbon oxygen balance in oxygen-rich molten steel, the added carbon participates in both carbonization and deoxidation. When the carbon content of the molten steel ranges from 0.038% to 0.12%, the proportion of carbon added for deoxidation is relatively high. As the dissolved oxygen content in the molten steel decreases, the carbon deoxidation rate slows down. Therefore, in actual production, measures need to be taken in the later stage of carbon deoxidation to promote the deoxidation effect.
• In the carbon deoxidation process, the final deoxidation of molten steel in LF is still completed by the Al element. Compared with the aluminum deoxidation process, the carbon deoxidation process has shown certain advantages in oxygen and nitrogen control. In addition, the evolution law of inclusions in the carbon deoxidation process can be summarized as Al-Si-Mn-O→Al-(Ca)-(Mg)-O→Ca-Al-(Mg)-(Si)-O.
• Compared with the aluminum deoxidation process, the number of inclusions in the cast billet produced by the carbon deoxidation process is reduced by 68.8%, and there are fewer large-sized inclusions in the billet. In addition, the carbon deoxidation process eliminates the addition of aluminum blocks after the converter, reducing the cost of deoxidants by CNY 15.47/ton of steel. In summary, the carbon deoxidation process has a better inclusion control effect and lower deoxidation costs.