Characteristics,formation mechanisms and removal of large-sized CaO-MgO-Al 2 O 3 inclusions in an aluminum-deoxidized and sulfur-containing XC45 steel

Abstract

XC45 steel, an aluminum-deoxidized and sulfur-containing steel, is widely used for mechanical components, whose fatigue performance and service life are strongly affected by CaO-MgO-Al₂O₃ (CMA) inclusions with large size and low interfacial tension, and commonly identified as D- or DS-type inclusions. To investigate their formation mechanism and control, systematic industrial sampling was conducted during LF refining, VD treatment, and tundish stages, together with inclusion characterization by SEM-EDS and thermodynamic analysis. The results showed that the composition of CMA inclusions in the billet was fixed at tundish stage, while their size was significantly influenced by the CaO content in inclusions. The changes of MgO content in inclusions affected their aggregation and melting/wetting behavior, thereby altering inclusion growth and removal trend. Based on these findings, a control strategy was proposed by maintaining total calcium at 0.0003–0.0005 wt.%, promoting the formation of removable MgO·Al₂O₃ and CaO·2Al₂O₃ inclusions while suppressing CaS precipitation. This ensures the liquid phase fraction of all inclusions exceeding 20%, thereby alleviating nozzle clogging during continuous casting. Industrial trials confirmed a significant reduction in inclusion amounts larger than 5 μm in the billets. This study provides an effective guideline for improving steel cleanliness and product quality of aluminum-deoxidized and sulfur-containing steels.

Keywords

XC45 steel formation mechanism thermodynamic calculation calcium treatment inclusion control continuous casting

Introduction

XC45 is a representative medium carbon, sulfur-containing, and aluminum-deoxidized steel. Due to its high strength, good toughness and low cost, it is widely used for manufacturing some key mechanical parts such as bearings, gears, and connecting rods.¹ The performance of this steel is strongly governed not only by its composition design but even more by its metallurgical cleanliness. Non-metallic inclusions in steel billet/bloom are an important indicator of metallurgical cleanliness and significantly influence the fatigue life and reliability of steel.^2–4 Therefore, clarifying inclusion evolution throughout refining and casting, and developing effective control strategies, are essential for producing high-quality XC45 products.

In industrial production, XC45 steel is commonly deoxidized by aluminum to reduce the dissolved oxygen in steel to a low level,^5–8 and maintain a certain acid-dissolved aluminum to refine the grain size of steel.^9–11 This inevitably generates brittle Al₂O₃ inclusions with high melting point and hardness, which differ markedly from the steel matrix. Such inclusions readily cause local stress concentration and serve as fatigue-crack initiation sites, severely degrading fatigue performance of the steels.^11–15 Moreover, high-melting Al₂O₃ can adhere to nozzle walls during continuous casting of steel, leading to submerged entry nozzle (SEN) clogging, even continuous casting production shutdown. Calcium treatment is therefore widely employed to modify solid Al₂O₃ into low-melting calcium aluminates, to alleviate the SEN clogging.^16–18 Because MgO-C refractories are used in ladles and tundishes, the modified inclusions are often CaO-MgO-Al₂O₃ (CMA) rather than simple CaO-Al₂O₃. These CMA inclusions may be captured by the solid-liquid interface during early solidification of liquid steel and retained as Type D or DS inclusions, which can severely impair steel performance. Suppressing the formation and growth of large CMA inclusions is thus critical for improving the cleanliness and overall quality of XC45 steel. In addition, to improve the cutting performance of steel, sulfur in steel is usually maintained at 0.02–0.05 wt.%.¹⁹ After calcium treatment, the reaction product of S and Ca, CaS readily precipitates on the surface of CMA, increasing its size and harm.

Several studies have been conducted on the formation and evolution of CMA inclusions. Li et al.²⁰ reported that large CMA inclusions in 42CrMo steel for wind-power applications mainly generated due to the reduction of CaO in slag by dissolved Al in molten steel ([Al]) under vacuum condition during vacuum degassing (VD) process. They suggested suppressing this reaction by controlling [Al] content at LF endpoint, slag basicity (CaO/SiO₂), and Al₂O₃ content in slag. Qiao et al.²¹ proposed that CMA inclusions formed via the reactions between dissolved Ca in molten steel ([Ca]) and solid MgO·Al₂O₃ spinel, and observed a sharp increase in number of liquid CMA inclusions when [Ca] reached 0.0003 wt.%. Miao et al.²² revealed that the super-large linear shaped CMA inclusions in bearing steel first originated from the liquid inclusions generated in molten steel during refining process. Following the subsequent solidification, these inclusions underwent aggregation and sintering due to the cavity bridge force of the CaAl₄O₇ shell and finally evolved into super-long linear defects during steel rolling. Qiao et al.²³ further found that MgO·Al₂O₃ inclusions originated from refractory in LF refining process. It dissolved into the molten steel to form [Mg], then reacted with Al₂O₃ to MgO·Al₂O₃. After Ca treatment, MgO·Al₂O₃ transformed into low-melting CaO-MgO-Al₂O₃ but remained difficult to remove, due to the low interfacial tension between the liquid inclusion and molten steel.

In addition to specific studies on CMA inclusions, fundamental researches on inclusion agglomeration and deformation behavior have also provided valuable insights. Mu et al.²⁴ systematically compared the agglomeration tendencies of different oxide inclusions and showed that the attractive force between Al₂O₃ particles is significantly stronger than that of Ti–Al complex oxides, highlighting the role of inclusion composition in aggregation behavior. Moreover, Mu et al.²⁵ used high-temperature confocal laser scanning microscopy to observe in situ the deformation, bending, and initial sintering process of Al₂O₃ clusters at the liquid steel/argon interface. Their results reveal that capillary forces drive these processes and the resistance force increases as inclusion particles approach each other. These fundamental findings provide a theoretical basis for understanding the aggregation and removal behavior of CMA inclusions in the present study.

Despite the above advances aimed at elucidating the formation, evolution, and elimination of CMA inclusions, the mechanisms underlying their formation across diverse steel grades and process conditions remain inadequately systematized. Moreover, many studies have neither proposed effective industrial control measures nor reported their application effects. This work focuses on the characteristics, formation mechanism, and control measures of CMA inclusions in industrial XC45 steel. By systematically sampling during the refining and casting stages of steel, and combining scanning electron microscopy (SEM), energy-dispersive spectroscopy (EDS), and FactSage thermodynamic calculations, this study reveals the origin and evolution laws of CMA inclusions and proposes the corresponding process optimization strategies. The findings provide theoretical insights and experimental evidence to support the improvements in the cleanliness and product quality of XC45 steel.

Materials and methods

Production process of XC45 steel

The XC45 steel was produced by the following industrial route: 150 t electric arc furnace (EAF) → ladle furnace (LF) refining → vacuum degassing (VD) → calcium treatment → sulfur treatment → soft argon stirring → continuous casting. The oxygen content was controlled at 0.03 wt.% when EAF melting was completed. The lime, Al particles, and slag formers were added for pre-deoxidation, slag formation, and alloying during EAF tapping. In the LF, additional lime, Al-bearing slag, and Al particles were added to promote desulfurization and deoxidation. The acid-soluble Al ([Al]s) was adjusted to the target range, then the ladle was carried to VD station. After degassing for 15 min at a vacuum degree of ≤ 67 Pa, the vacuum was broken and calcium treatment was performed to modify Al₂O₃ into low-melting calcium aluminates. Sulfur wire was then fed into the molten steel to increase the cutting performance of steel. The molten steel was then softly stirred with argon for about 15 min to promote inclusion flotation and to homogenize composition and temperature of molten steel. Finally, the molten steel was continuously cast through a six-strand tundish into a six-strand curved caster to produce billets with 220 mm × 220 mm cross section. The target chemical composition of XC45 steel is shown in Table 1.

Table 1.

Target chemical composition of XC45 steel (wt.%)

C	Si	Mn	P	S
0.42–0.48	0.10–0.35	0.50–0.80	≤ 0.035	≤ 0.025

Sampling and analysis methods

As illustrated in Figure 1, molten-steel samples were taken with a barrel-shaped sampler at LF arrival (LF1), LF end (LF2), VD vacuum breaking (VD1), VD ladle transfer (VD2), and during steady-state tundish casting (TU). Besides this, a billet specimen of 300 mm long was collected after casting (BT). As shown in Figure 2, the molten-steel and billet samples were machined into metallographic specimens (15 mm × 10 mm × 10 mm) and O/N rods (Φ5 mm × 6 mm).

Figure 1.

Sampling locations along the refining and casting route

Figure 2.

Specimen preparation, (a) billet, (b) barrel-shaped (molten-steel) sample

Metallographic specimens were mounted in conductive resin and prepared by grinding and polishing, then inclusions in them were examined using a ZEISS EVO18 (Carl Zeiss AG, Germany) scanning electron microscope equipped with an energy-dispersive spectrometer. After that, an automated inclusion analysis system was used to quantify the size and composition of inclusions larger than 1 μm over a scanned area of 30 mm². After grinding and ultrasonic cleaning in ethanol, O/N rods were analyzed for total oxygen (T.O) and total nitrogen (T.N) contents using LECO ONH836 (LECO Corporation, USA). Bulk C, Si, Mn, P, total sulfur (T.S), and total Al ([Al]_t) contents were measured by ARL8860 (Thermo Scientific, USA). Total calcium (T.Ca) content was determined by ICP-AES (NCS Testing Technology, China). The Mg content in the steel matrix was not directly measured due to its low concentration. Instead, based on industrial experience and other studies,^9,26 a value of 0.0005 wt.% was adopted for thermodynamic calculations. The results are summarized in Table 2. Thermodynamic calculations were performed with FactSage 8.3 software, choosing the Equilib and Phase Diagram modules and FactPS, FToxid, and FTmisc databases.

Table 2.

Chemical composition of molten steel at different processing stages (wt.%)

Sample	C	Si	Mn	P	T.S	Cr	Ni	[Al]_t	T.Ca	T.O	T.N
LF1	0.37	0.26	0.70	0.010	0.006	0.062	0.022	0.037	<0.0005	0.0148	0.0083
LF2	0.43	0.30	0.77	0.010	0.003	0.20	0.022	0.034	0.0006	0.0034	0.0040
VD1	0.47	0.32	0.80	0.010	0.002	0.20	0.022	0.018	<0.0005	0.0021	0.0041
VD2	0.47	0.32	0.79	0.010	0.013	0.20	0.021	0.020	0.0012	0.0018	0.0042
TU	0.46	0.29	0.80	0.012	0.014	0.20	0.021	0.017	0.0008	0.0015	0.0049

Results and discussion

Morphology, types, and size distribution of inclusions in the billets

Figure 3 shows the morphologies, chemical compositions, and elements distribution maps of typical inclusions in billet specimens. The large-sized inclusions (>10 μm) in billets consisted of a CaO-MgO-Al₂O₃ oxide core and surrounded by CaS layer. Their two-dimensional morphologies were mostly circular or near-circular. Such morphologies are commonly attributed to the heterogeneous precipitation of CaS on oxide surfaces during solidification.^27,28 Therefore, controlling the formation of the CMA oxide core is the key to reducing these composite inclusions.

Figure 3.

Typical CaS-rimmed CaO-MgO-Al₂O₃ (CMA) inclusions observed in the billet. (a) No.1 inclusion. (b) No.2 inclusion. (c) No.3 inclusion

After removing the CaS content from the composite inclusions, the remaining CaO, MgO, Al₂O₃ contents were normalized and plotted on the ternary phase diagram at 1550 °C calculated using FactSage 8.3, as shown in Figure 4. The region enclosed by the red line denotes the completely liquid phase. It is seen that most inclusions lay in Al₂O₃-rich regions. The inclusions with sizes of 1–3 μm clustered in the MA + CM₂A₈ and A + MA + CM₂A₈ phase regions, while those 3–5 μm in size mainly fell in the transition region between MA + CA₂ + L and CA₂ + CM₂A₈ + MA. The inclusions larger than 5 μm extend deeper toward the CA₂-MA-L three-phase region. Statistically, for CaO-Al₂O₃-based inclusions, higher CaO content correlated with larger particle size, which is consistent with prior observations.^29,30 This suggests that inclusions approaching the low-melting C₁₂A₇ composition tend to grow larger, likely because liquid or semi-liquid particles coalesce more readily in molten steel.

Figure 4.

CaO-MgO-Al₂O₃ compositions of oxide cores in billet inclusions (CaS removed and normalized) plotted on the ternary phase diagram at 1550 °C

Source of CaO-MgO-Al₂O₃ inclusions

To identify the origin of CMA inclusions in billet, the inclusions in molten steel at each processing stage were characterized by SEM-EDS, and the results are shown in Figure 5. At the LF arrival stage (LF1), Al₂O₃ inclusions from Al deoxidation were modified by the high-basicity slag into CMA-type inclusions as shown in Figure 5(a1) and Figure 5(a2). A small MgO content in inclusions originated partly from slag-steel mass transfer, and the other from dissolution or erosion of MgO-C refractories into molten steel. When LF refining was completed (LF2), the activity of dissolved [Mg] increased due to high temperature and reducing atmosphere. It further reacted with CMA inclusions, increasing the MgO content in the inclusions to about 30 wt% (Figure 5(b1) and Figure 5(b2)). After LF desulfurization, the free CaO in the refining slag was transformed into CaS by the reaction [S] + (CaO) = (CaS) + [O], resulting in a decrease in the activity of CaO. After VD operation, the inclusions in steel still maintained MgO-Al₂O₃ rich, as shown in Figure 5(c1) and Figure 5(c2). After calcium treatment, dissolved [Ca] reacted with MgO-Al₂O₃, increasing the CaO content in CMA inclusions and producing low-melting calcium-aluminate liquid inclusions as in Figure 5(d1). What's more, the increased [S] content after feeding sulfur wire led to the formation of CaS-bearing inclusions, as shown in Figure 5(d2). During tundish casting (TU), the temperature of molten steel decreased, which reduced the solubility of [Ca] and [S], causing CaS preferentially precipitated on existing oxide particles and formed the typical CaS-CMA composite inclusions (Figure 5(e1) and Figure 5(e2)).

Figure 5.

Inclusion morphologies at different stages

To further trace the source of inclusions, CaS content was removed from the measured inclusions at each stage, and the remaining CaO, MgO, and Al₂O₃ contents were normalized and projected onto the CaO-MgO-Al₂O₃ ternary diagram, as shown in Figure 6. It is seen that the inclusions in tundish (TU) closely matched those observed in the billet (Figure 4), whereas inclusions in TU sample exhibited higher Al₂O₃ content and lower CaO content than those in VD2 sample. As shown in Table 2, T.N content increased from 0.0042 wt.% (VD2) to 0.0049 wt.% (TU), while [Al]_t content decreased from 0.020 wt.% to 0.017 wt.%. This phenomenon indicated that the molten steel was re-oxidized by air, leading to nitrogen pickup and aluminum consumption. Reoxidation promoted Al₂O₃ formation, thereby increasing the Al₂O₃ content in inclusions in TU sample. It is worth noting that although the variations in T.N and [Al]_t contents indicated the reoxidation of molten steel, the T.O content decreased from 0.0018 wt.% (VD2) to 0.0015 wt.% (TU), which was possibly caused by the different sampling and detection position.³¹ The reference³¹ showed that different sample position will significantly affect the T.O content, but less on T.N content. As not enough samples in TU station can be detected again, this phenomenon will be paid more attention in the future work.

Figure 6.

Evolution of CaO-MgO-Al₂O₃ oxide-core compositions during refining and casting (CaS removed and normalized)

In summary, the CMA inclusions observed in billet originated from refining process and were transformed into low-melting calcium aluminates after Ca treatment. However, they underwent compositional shifts driven by the reoxidation in the tundish, leading to an increase in Al₂O₃ content. Subsequently, these CMA inclusions acted as heterogeneous nucleation core for CaS precipitation during liquid steel solidification. Therefore, improving tundish casting protection to suppress reoxidation occurrence is crucial for controlling the final CMA + CaS composite inclusions.

Compositional and physicochemical characteristics of CaO-MgO-Al₂O₃ inclusions

After identifying the source of CMA inclusions, their compositional features were analyzed to provide a basis for controlling the type of inclusion. Figure 7 shows that most CMA inclusions in billet contained 70–90 wt.% Al₂O₃ and 10–20 wt.% MgO. Within this range, Al₂O₃ and MgO contents varied only weakly with inclusion size, whereas CaO content showed a clear positive correlation with size. As the CaO content increased, inclusion size showed an increasing trend.

Figure 7.

Composition-size relationships of billet CaO-MgO-Al₂O₃ inclusions. (a) Al₂O₃ and CaO vs. particle size. (b) MgO and CaO vs. particle size

To clarify the melting behavior of inclusions with different compositions at steelmaking temperatures, FactSage 8.3 (Equilib module, FToxid database) was used to calculate the initial-melting temperature of CaO-MgO-Al₂O₃ inclusions as a function of CaO and MgO contents. The result is shown in Figure 8, in which the curves of various colors correspond to MgO mass fractions of 0, 5, 10, 15, and 20 wt.% respectively. When CaO content was 0∼30 wt.%, higher MgO lowered the initial-melting temperature at a given CaO content, particularly in the CA₂ + CA region. As the CaO content further increased, the influence of MgO on the initial-melting temperature weakened. In the CA + C₃A region, the initial-melting temperature exhibited a decrease followed by an increase trend with rising MgO content. Overall, increasing MgO reduced the critical CaO content required to form a liquid phase, implying that MgO can promote melting of inclusions by altering phases. The low-melting point inclusions are difficult to remove and can grow by collision and coalescence in molten steel,³² thus limiting MgO content in CMA may be favorable to avoid the formation of fully liquid inclusions.

Figure 8.

Effect of CaO and MgO contents on the initial-melting temperature of CaO-MgO-Al₂O₃ inclusions

Attraction tendency among inclusions is a useful indicator for evaluating the ability of inclusions to collide, agglomerate, and coalesce into large particles. Figure 9 summarizes reported attraction forces for common oxide inclusions.^33–35 The Al₂O₃ exhibits the strongest attraction, followed by Al₂O₃-rich inclusions (such as CA₆, 92 wt.% Al₂O₃-8 wt.% CaO, and MA), suggesting a high propensity for clustering. For CaO-Al₂O₃ inclusions (CA₂, CA, C₁₂A₇), attraction force decreases as CaO content increases. Among the listed inclusions in Figure 9, C₁₂A₇ and 55 wt.% Al₂O₃-40 wt.% CaO −5 wt.% MgO show the weakest attraction. Compared to CA (65 wt.% Al₂O₃-35 wt.% CaO), the 60 wt.% Al₂O₃-30 wt.% CaO-10 wt.% MgO inclusions exhibited a smaller lower limit of attraction force due to its 10 wt.% MgO content. Although the 60 wt.% Al₂O₃-30 wt.% CaO-10 wt.% MgO and 55 wt.% Al₂O₃-40 wt.% CaO-5 wt.% MgO inclusions displayed the same lower limit of attraction force, the former held a higher upper limit owing to its increased MgO content. For CaO-MgO-Al₂O₃ (CMA) inclusions, reasonably reducing the MgO content can decrease the attractive force between inclusions. For example, forming 55 wt.% Al₂O₃-40 wt.% CaO-5 wt.% MgO inclusions instead of 60 wt.% Al₂O₃-30 wt.% CaO-10 wt.% MgO can suppress their aggregation and growth.

Figure 9.

Reported attraction forces between different oxide inclusions in molten steel

An effective control on CMA inclusions requires balancing its resistance to agglomeration and efficient removal. Removal tendency of inclusions is strongly influenced by its wettability, which is commonly characterized by the contact angle. A larger contact angle implies poorer wettability and higher interfacial energy, favoring the separation of inclusions from molten steel. Table 3 summarizes reported contact angles between molten steel and typical inclusions.^36–39 For CaO-Al₂O₃ inclusions, increasing CaO content reduces the contact angle. The Al₂O₃, along with CA₆, and CA₂ inclusions where the CaO content is below 23 wt.%, all possess high melting points with 2050 °C, 1850 °C, and 1750 °C, respectively, and their contact angles all exceed 120°. When CaO in inclusions exceeds 35 wt.%, liquid phases such as CA (1605 °C) and C₁₂A₇ (1455 °C) form. Due to their contact angles below 70°, it is hard to separate them from the molten steel. The MA inclusion also shows a large contact angle (>120°), implying that it can be removed more readily than liquid calcium aluminates. Considering both attraction force (Figure 9) and wettability, CA₆ and CA₂ inclusions possess high contact angles and low attraction, leading them easier to remove and less prone to agglomerate. Therefore, steering inclusions toward CA₆, CA₂, MA-type is expected to be favorable for reducing both their size and number density.

Table 3.

Contact angles between typical inclusions and molten steel

Oxides	Temperature (°C)	Contact angle (°)	Ref.
Al₂O₃	1600	135	37,38
MgO·Al₂O₃ (28:72)	1600	134	37,39
CaO·6Al₂O₃ (8:92)	1600	>120	36
CaO·2Al₂O₃ (22:78)	1600	120	37,39
CaO-Al₂O₃ (36:64) (CA)	1600	65	37,38
CaO-Al₂O₃(50:50) (C₁₂A₇)	1600	58	37,38
CaO-Al₂O₃(58:42)	1600	54	37,38

It should be pointed out that low attraction tendency of inclusions mainly suppresses their agglomeration and growth, whereas the actual removal of inclusions is controlled by a combination of its wettability and size, and flow of molten steel and stirring conditions.

Control strategy for CaO-MgO-Al₂O₃ inclusions

Based on the above analysis, reducing CaO and MgO contents in inclusions is critical to suppress the formation of fully liquid CMA inclusions. Since inclusion composition is governed by dissolved [Ca] and [Al] contents, the stability phase diagram of major inclusion in the VD2 steel at 1600 °C was calculated using FactSage 8.3 software. The result is shown in Figure 10, in which the red region represents the fully liquid region and the blue region denotes the desired control window where CA₂ and MA can coexist. The composition of VD2 steel after Ca treatment is projected into this figure. It is seen that it (purple point) fell in the fully liquid phase region. Although the liquid inclusions help avoid SEN clogging, they are typically larger and more difficult to remove. Conversely, controlling all inclusions to be solid CA₂ or MA could also increase the SEN clogging risk because of their relatively high melting points.^40,41 Therefore, an intermediate strategy that maintains a controlled liquid fraction is required.

Figure 10.

Stability fields of inclusion phases in the VD2 steel at 1600 °C calculated by FactSage 8.3

Previous studies^42,43 suggest that nozzle clogging can be effectively mitigated when the liquid fraction of inclusions exceeds 20%. Accordingly, the variation of inclusion type and amount with T.Ca content in the VD2 steel (after Ca treatment) at 1600 °C was calculated using FactSage 8.3 software. The result is shown in Figure 11(a). The change of major constituents (Al₂O_3(l), CaO_(l), MgO_(l), SiO_2(l), and CaS_(l)) in the liquid phase with T.Ca is shown in Figure 11(b). The vertical dashed line marked the T.Ca content at which the calculated liquid fraction reached 20%. It is approximately 0.00029 wt.%. From Figure 11(a), when T.Ca content was low (0-0.0001 wt.%), inclusions were dominated by CaO·2MgO·8Al₂O₃, MgO·Al₂O₃, and CaO·2Al₂O₃. As T.Ca increased, liquid inclusions appeared and its content rose rapidly before approaching stable. Al₂O₃ and CaO become the dominant constituents in liquid inclusions, as shown in Figure 11(b). At a high T.Ca content (about 0.00145 wt.%), solid CaS started to form. It often precipitates on oxide inclusions to expand their size, also readily causing SEN clogging. Therefore, maintaining T.Ca content at 0.0003–0.0005 wt.% is recommended, in which the liquid fraction remains above 20% to help reducing nozzle clogging risk, and solid MgO·Al₂O₃ and CaO·2Al₂O₃ are still present. Moreover, CaS precipitation is suppressed. In addition, the total amount of oxide inclusions in the liquid inclusion phase was relatively low in the T.Ca content range, which helped reduce the overall inclusion population.

Figure 11.

Inclusion evolution with calcium content at 1600 °C. (a) Phase fractions of solid and liquid inclusions. (b) Major constituents of the liquid phase

Optimized industrial trial based on the above control strategy of inclusions

An optimized industrial trial was conducted based on the above analysis. The T.Ca content after calcium treatment was controlled at approximately 0.0004 wt.%. Because Section 3.2 indicates that the component content in CMA inclusions was finally fixed at tundish stage, only inclusions in tundish (TU) and billet (BT) were statistically analyzed. Inclusion compositions (excluding CaS) were normalized and plotted on the CaO-MgO-Al₂O₃ ternary phase diagrams, as shown in Figure 12. The comparison of Figure 12(a) with Figure 6 reveals that the distribution of inclusions in tundish remained similar after optimization, but the inclusion population was markedly decreased. More importantly, the comparison of Figure 12(b) with Figure 4 shows that billet inclusions become significantly finer, and the number of inclusions larger than 5 μm decreased substantially.

Figure 12.

CaO-MgO-Al₂O₃ distributions of oxide cores after process optimization. (a) Tundish inclusions. (b) Billet inclusions with size classification

These results indicate that maintaining a relatively low T.Ca content promotes the formation of MgO·Al₂O₃ and CaO·2Al₂O₃ and keeps the liquid fraction above 20%. This dual effect reduces both the number and size of inclusions and no obvious nozzle clogging occurred in a sequence of eight continuous casting heats. Therefore, the control strategies of calcium treatment and inclusions are essential for improving XC45 steel cleanliness and castability. These metallurgical mechanisms will be further investigated in future work.

Conclusions

Through systematical analysis of the characteristics, formation mechanisms and control strategies of CaO-MgO-Al₂O₃ type of inclusion in sulfur-containing and aluminum-deoxidized XC45 steel, the following conclusions are drawn.

The major large-sized inclusions in XC45 billets were composite inclusions with CaO-MgO-Al₂O₃ oxide as core, surrounded by a CaS shell. Their two-dimensional morphology was typically circular or near-circular. Controlling the formation of CaO-MgO-Al₂O₃ oxide core was the key to reduce their damage to steel performance. Every component content in CaO-MgO-Al₂O₃ oxide varied in different refining stages but it was finally fixed at tundish stage, as indicated by thermodynamic calculations and analysis of inclusion composition evolution.

The CaO content was the primary factor influencing the size of CMA inclusion. A higher CaO corresponded to a larger particle size, whereas Al₂O₃ and MgO content showed a relatively weak size dependence in the investigated range. Reducing MgO from 10 wt.% to 5 wt.% in CMA is beneficial for decreasing the attraction tendency among inclusions and suppressing their agglomeration. However, the inclusion containing MgO will reduce its initial melting temperature and promote low-melting liquid phase formation, which potentially hindering the flotation and removal of inclusions.

Thermodynamic calculations indicated that maintaining T.Ca at 0.0003-0.0005 wt.% ensured an inclusion liquid fraction exceeding 20%, which helped to alleviate nozzle clogging. This T.Ca content window also favored the formation of easily removable solid MgO·Al₂O₃ and CaO·2 Al₂O₃, while suppresses the harmful CaS inclusion precipitation.

Industrial trials confirmed that the proposed control strategy of calcium treatment significantly reduced the number of inclusions larger than 5 μm in billets, thereby improving steel cleanliness and providing an economically feasible option for actual XC45 steel production.

Footnotes

Acknowledgements

The authors gratefully acknowledge Baoshan Iron & Steel Co., Ltd., and Nanjing Iron & Steel Co., Ltd. for their valuable support in conducting the industrial trials.

ORCID iD

Yisheng Chen

Ethical considerations

Not applicable. This study does not involve human participants, human data, or human tissue.

Funding

The authors received no financial support for the research, authorship, and/or publication of this article.

Declaration of conflicting interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Data availability

The data that support the findings of this study are available from the corresponding author upon reasonable request.

References

Chai

Guo

. Two flowing stress models for hot deformation of XC45 steel at high temperature. Mater Sci Eng A 2012; 534: 101–110.

Xing

Wang

, et al. Bending fatigue behaviors analysis and fatigue life prediction of 20Cr2Ni4 gear steel with different stress concentrations near non-metallic inclusions. Mater 2019; 12: 3443.

Babu

Michelic

. Analysis of non-metallic inclusions by means of chemical and electrolytic extraction—a review. Mater 2022; 15: 3367.

Merder

Warzecha

Pieprzyca

, et al. New design solution for porous plug in the argon steel purging process. Sci Rep 2025; 15: 21540.

Wang

, et al. A multi-step thermodynamic model for alumina formation during aluminum deoxidation in fe-O-al melt. Acta Metall Sin Engl Lett 2015; 28: 272–280.

Wang

Xiao

Zhao

, et al. Atomic cluster aggregates in nucleation of solid alumina inclusion in the aluminum deoxidation for liquid iron. Metall Mater Trans B 2018; 49: 282–290.

Wang

, et al. Evidence of multi-step nucleation leading to various crystallization pathways from an fe-O-al melt. Sci Rep 2014; 4: 5082.

Lei

Xiao

Wang

, et al. Thermodynamic insight into the growth of nanoscale inclusion of al-deoxidation in fe-O-al melt. Sci Rep 2020; 10: 16909.

Wang

Tang

Jia

, et al. Effect of deoxidizer addition sequence on cleanliness of al-killed and S-containing steel: industrial trials and thermodynamic analysis. ISIJ Int 2025; 65: 1289–1301.

10.

Wang

Tang

Chen

, et al. Effect of refining slag basicity on transformation and removal of inclusions in al-killed S-containing steel. J Mater Res Technol 2025; 39: 5225–5237.

11.

Wang

Tang

Wang

, et al. Formation mechanism and control of a kind of pit-shaped defect in Cf53 rolled products. Chem Eng J 2024; 495: 153273.

12.

Long

Wang

, et al. Inclusion characteristics in 95CrMo steels with different calcium and sulfur contents. Mater 2020; 13: 619.

13.

Juan

Wang

Lian

, et al. Quantifying the comprehensive characteristics of inclusion-induced defects using an integrated destructive and non-destructive method. Mater 2021; 14: 1475.

14.

Yan

Chen

. Quality control of high carbon steel for steel wires. Mater Basel Switz 2019; 12: 846.

15.

Wang

. A kinetic model for the modification of Al₂O₃ inclusions during calcium treatment in high-carbon hard wire steel. Mater 2021; 14: 1305.

16.

Shen

. Morphology study on inclusion modifications using mg-ca treatment in resulfurized special steel. Mater Basel Switz 2019; 12: 197.

17.

Pyzalski

Sujak

Durczak

, et al. The effect of biological corrosion on the hydration processes of synthetic tricalcium aluminate (C₃A). Mater Basel Switz 2023; 16: 2225.

18.

Wang

Chen

, et al. Effect of mg treatment on the nucleation and ostwald growth of inclusions in fe-O-al-mg melt. Mater Basel Switz 2020; 13: 3355.

19.

Zhou

Zhang

, et al. Research and practice of calcium treatment processes insulfur-containing free-cutting steel. Iron Steel 2024; 59: 68–80.

20.

Zhang

Qian

, et al. Formation mechanism and control of large-size CaO-MgO-Al₂O₃ inclusions in steel used for wind power. ISIJ Int 2023; 63: 1941–1950.

21.

Qiao

Cheng

Huang

, et al. A study on formation mechanism of CaO-MgO-Al₂O₃ inclusions in 42CrMo4 steel. Jiangxi Metall 2022; 42: 8–15.

22.

Miao

Long

Cheng

, et al. Agglomeration and clustering of CaO-Al₂O₃-MgO leading to super large-size line-shape inclusions in high carbon chromium bearing steel. Metall Mater Trans B 2022; 53: 512–525.

23.

Qiao

Cheng

Huang

, et al. Formation and removal mechanism of nonmetallic inclusions in 42CrMo4 steel during the steelmaking process. Metals (Basel) 2022; 12: 1505.

24.

Xuan

. Agglomeration mechanism of complex ti-al oxides in liquid ferrous alloys considering high-temperature interfacial phenomenon. Metall Mater Trans B 2019; 50: 2694–2705.

25.

Dogan

Coley

. In situ observation of deformation behavior of chain aggregate inclusions: a case study for Al₂O₃ at a liquid steel/argon interface. J. Mater. Sci 2018; 53: 13203–13215.

26.

Wang

Chen

Wang

, et al. Characteristics and evolution of non-metallic inclusions in GCr15 bearing steel. Steelmaking 2024; 40: 89–97.

27.

Wang

Valdez

Sridhar

. Formation of CaS on Al₂O₃-CaO inclusions during solidification of steels. Metall Mater Trans B 2002; 33: 625–632.

28.

Sidorova

Karasev

Kuznetsov

, et al. Investigation of the initial corrosion destruction of a metal matrix around different non-metallic inclusions on surfaces of pipeline steels. Mater Basel Switz 2022; 15: 2530.

29.

Cheng

, et al. Characteristics and formation mechanism of duplex (ca,mn)S inclusions in commercial ca-treated resulfurized steel. Met Mater Int 2023; 29: 1019–1033.

30.

Zhao

Bao

, et al. Inclusion evolution during modification of alumina inclusions by calcium in liquid steel and deformation during hot rolling process. ISIJ Int 2015; 55: 2115–2124.

31.

Han

Feng

. Effects of sample preparation and sampling methods on analysis results of oxygen and nitrogen in steel. Res. Explor. Lab 2016; 35: 49–55.

32.

Yin

Shibata

Emi

, et al. Characteristics of agglomeration of various inclusion particles on molten steel surface. ISIJ Int 1997; 37: 946–955.

33.

Schröder

Fischer

Schmidt

, et al. Interactions between exogenous magnesia inclusions with endogenous inclusions in a high alloyed steel melt. Adv Eng Mater 2017; 19: 1700146.

34.

Nakajima

Mizoguchi

. Capillary interaction between inclusion particles on the 16Cr stainless steel melt surface. Metall Mater Trans B 2001; 32: 629–641.

35.

Ren

, et al. In situ observation of the agglomeration of MgO-Al₂O₃ inclusions on the surface of a molten GCr15-bearing steel. Metall Mater Trans B 2023; 54: 1159–1173.

36.

Bao

Cheng

Zhang

, et al. Characteristics, formation mechanism, and removal of large-sized CaO-Al₂O₃-SiO₂ inclusions in D2 high-speed railway wheel steel. Steel Res Int 2024; 95: 2400497.

37.

Deng

Liu

Zhu

, et al. Formation, evolution and removal of MgO·Al₂O₃ spinel inclusions in steel. ISIJ Int 2021; 61: 1–15.

38.

Cramb

Jimbo

. Calculation of the interfacial properties of liquid steel - slag systems. Steel Res 1989; 60: 157–165.

39.

Kang

Sahebkar

Scheller

, et al. Observation on physical growth of nonmetallic inclusion in liquid steel during ladle treatment. Metall Mater Trans B 2011; 42: 522–534.

40.

Zhang

. Nonmetallic inclusions in steels: industrial practice. 1st ed. Beijing: Metallurgical Industry Press, 2019, p.290.

41.

Liu

Zhang

. Relationship between dissolved calcium and total calcium in al-killed steels after calcium treatment. Metall Mater Trans B 2018; 49: 1624–1631.

42.

Zhang

Luo

, et al. Effect of calcium treatment on nonmetallic inclusions in 20CrMnTiH gear steel. Chin J Eng 2021; 43: 825–834.

43.

Wang

, et al. Dependence of the clogging possibility of the submerged entry nozzle during steel continuous casting process on the liquid fraction of non-metallic inclusions in the molten al-killed ca-treated steel. Metals (Basel) 2020; 10: 1205.

Characteristics,formation mechanisms and removal of large-sized CaO-MgO-Al 2 O 3 inclusions in an aluminum-deoxidized and sulfur-containing XC45 steel

Abstract

Keywords

Introduction

Materials and methods

Production process of XC45 steel

Sampling and analysis methods

Results and discussion

Morphology, types, and size distribution of inclusions in the billets

Source of CaO-MgO-Al2O3 inclusions

Compositional and physicochemical characteristics of CaO-MgO-Al2O3 inclusions

Control strategy for CaO-MgO-Al2O3 inclusions

Optimized industrial trial based on the above control strategy of inclusions

Conclusions

Footnotes

Acknowledgements

ORCID iD

Ethical considerations

Funding

Declaration of conflicting interests

Data availability

References

Source of CaO-MgO-Al₂O₃ inclusions

Compositional and physicochemical characteristics of CaO-MgO-Al₂O₃ inclusions

Control strategy for CaO-MgO-Al₂O₃ inclusions