Secondary refining is a major metallurgical process in steelmaking plants for deep deoxidation and desulfurization of molten steel, removal of non-metallic inclusions, and alloying. Considering the differences in corrosive media at various locations within an LF (ladle furnace) refining ladle, the refractory lining can be divided primarily into two regions: the slag line zone and the molten steel bath zone.
The slag line zone of a ladle is subjected to severe corrosion by oxide-based slags. Therefore, MgO-C bricks containing 12–18 wt% carbon are commonly used in this region. Because the refractory lining at the slag line is rapidly eroded by molten slag, the slag line generally needs to be replaced at least once during the service life of the ladle.
The refractory lining in the molten steel bath zone is primarily exposed to erosion by molten steel during refining. Before hot repair of the ladle after steel tapping, the residual slag must be discharged. The refractory material at the slag discharge surface is mainly subjected to chemical attack by the slag and physical erosion caused by slag flow. Refractory materials suitable for the molten steel bath zone include MgO-Al₂O₃-C, Al₂O₃-MgO-C, and carbon-free Al₂O₃-MgO refractories.
During service, the refractory lining in the molten steel bath zone is located in the middle and lower sections of the ladle, where replacement is difficult and generally offers limited economic benefits. Consequently, the service life of the refractory lining in this zone largely determines the overall service life of the ladle.
The degradation mechanisms of refractory materials are relatively complex. Factors such as slag properties, operating temperature, refractory composition, and metallurgical conditions all affect the corrosion process. In industrial ladle linings, refractory degradation is the result of the combined and coupled effects of the temperature field, steel/slag flow fields, and chemical corrosion. Therefore, investigating the localized degradation mechanisms of refractories in the molten steel bath zone is not only scientifically important for understanding refractory corrosion mechanisms but also highly valuable for guiding industrial practice.
1. Investigation of the Current Situation
During the dismantling of a 180 t converter steel ladle at a steel plant, it was observed that the residual thickness of the refractory bricks in the molten steel bath zone at the 12 o'clock position was significantly lower than that of the surrounding bricks (see Fig. 1). This localized wear has limited further improvement in ladle service life, as shown in Table 1.
The main operating conditions at the plant are as follows: the ladles are used exclusively for LF refining, accounting for 100% of their refining operations. Most of the steel grades produced are aluminum-killed steels, and the slag basicity is relatively high, as shown in Table 2.
Regarding the internal configuration of the ladle, the porous plugs are located at the 5 o'clock and 7 o'clock positions (see Fig. 2). During refining, the electrodes are inserted into the ladle at the 1 o'clock, 2 o'clock, and 6 o'clock positions (see Fig. 3).
Figure 1. Thickness of the molten steel bath zone refractory bricks during ladle dismantling
| Residual Thickness | Direction | |||||||
|---|---|---|---|---|---|---|---|---|
| 12 o'clock | 1 o'clock | 3 o'clock | 5 o'clock | 6 o'clock | 7 o'clock | 9 o'clock | 11 o'clock | |
| 1st layer | 50 | 80 | 100 | 130 | 110 | 110 | 120 | 70 |
| 4th layer | 60 | 90 | 130 | 150 | 120 | 120 | 140 | 90 |
| 7th layer | 70 | 110 | 150 | 160 | 125 | 130 | 160 | 110 |
Table 1. Residual Thickness of the Ladle Working Lining Bricks (mm)
| Steel Grade | CaO | MgO | SiO₂ | Al₂O₃ | Fe₂O₃ | Mn₂O₃ |
|---|---|---|---|---|---|---|
| SPHC | 58.5 | 5.44 | 6.59 | 24.96 | 0.42 | - |
Table 2. Composition of the Refining Slag Used in Industrial Operations
As is well known, localized weak points in ladle refractories commonly occur in three areas:
(1) near the electrodes, where the high temperatures generated by the electrodes accelerate the reactions between the slag/molten steel and the refractory materials;
(2) near the porous plugs, where gas purging intensifies the erosion of the refractory lining by molten steel;
(3) at the slag discharge surface, which is subjected to erosion and corrosion by molten steel and slag during ladle emptying at the end of each operating cycle.
However, the localized weak point observed at this plant is not located in any of these three conventional areas, making its investigation particularly significant.
Post-service refractory bricks were collected from both the weak and non-weak areas of the molten steel bath zone at the plant, as shown in Figs. 4 and 5. The effective residual thicknesses of the bricks from the weak and non-weak areas were 160 mm and 60 mm, respectively. As shown in the figures, a slag layer was present on the surface of the refractory bricks from the non-weak area, whereas almost no slag layer was observed on the bricks from the weak area. In addition, two cracks parallel to the hot face were observed in the bricks from the non-weak area, and these cracks had been filled with slag.
The retrieved refractory bricks were subsequently analyzed to identify the underlying degradation mechanisms.
Figure 2. In-Service View of the Ladle Showing the Locations of the Porous Plugs
Figure 3. Schematic Diagram of the Relative Positions of the Graphite Electrodes Inside the Ladle
Figure 4. Appearance of the Post-Service Refractory Bricks from the 5 o'clock Position of the Ladle Wall
Figure 5. Appearance of the Post-Service Refractory Bricks from the 12 o'clock Position of the Ladle Wall
2. Inspection and Analysis of Post-Service Refractory Bricks
2.1 Inspection Methods and Procedures
Samples measuring 25 mm × 25 mm × 25 mm were cut at 25 mm intervals from the hot face toward the cold face for physical and chemical property testing. The hot-face region was subjected to microstructural analysis.
2.2 Physicochemical Analysis of Post-Service Refractory Brick Samples
Table 3 presents the physical properties of samples collected at different distances from the hot face. In the 0–25 mm region, the apparent porosity and bulk density of the refractory bricks from the weak and non-weak areas of the molten steel bath zone showed little difference. However, the bulk density of the sample from the weak area was slightly higher than that from the non-weak area.
In the 30–55 mm region, the sample from the weak area exhibited a bulk density of 2.80 g/cm³, significantly higher than the 2.74 g/cm³ measured in the sample from the non-weak area. Its apparent porosity was 14.8%, markedly lower than the 16.2% measured in the sample from the non-weak area.
These results indicate that, at the same distance from the hot face, samples from the weak area exhibited a higher bulk density and lower apparent porosity. This indirectly suggests that the refractory bricks in the weak area had a thinner slag penetration layer or decarburized layer than those in the non-weak area.
| Distance from hot face / mm | Sample 1 (5 o'clock direction) | Sample 2 (12 o'clock direction) | ||
|---|---|---|---|---|
| Apparent Porosity / % | Bulk Density / g·cm⁻³ | Apparent Porosity / % | Bulk Density / g·cm⁻³ | |
| 0~25 | 17.0 | 2.75 | 17.1 | 2.76 |
| 30~55 | 16.0 | 2.74 | 14.8 | 2.80 |
| 60~85 | 14.0 | 2.81 | - | - |
| 90~115 | 12.2 | 2.89 | - | - |
Table 3. Apparent Porosity and Bulk Density of Samples at Different Distances from the Hot Face
| Distance from hot face / mm | Chemical composition / wt% | ||||||
|---|---|---|---|---|---|---|---|
| Direction | CaO | MgO | SiO₂ | Al₂O₃ | Fe₂O₃ | C | |
| ≤25 | 5 o'clock | 8.93 | 58.16 | 5.71 | 15.05 | 0.85 | 7.23 |
| 12 o'clock | 3.35 | 66.60 | 4.16 | 14.16 | 0.76 | 6.08 | |
| 30–55 | 5 o'clock | 1.57 | 66.58 | 5.03 | 12.27 | 0.69 | 10.9 |
| 12 o'clock | 1.29 | 64.80 | 4.07 | 14.12 | 0.62 | 11.13 | |
| 60–85 | 5 o'clock | 1.30 | 65.97 | 4.75 | 12.72 | 0.68 | 11.0 |
| 12 o'clock | - | - | - | - | - | - | |
Table 4. Chemical Composition of Samples at Different Distances from the Hot Face
Table 4 presents the chemical composition of samples collected at different distances from the hot face. The results show that the CaO content gradually decreases with increasing distance from the hot face. For samples collected at the same distance, the CaO content in the non-weak area is higher than that in the weak area. The CaO content of the original, unaltered refractory layer is estimated to be approximately 1.3 wt%. In the non-weak area at the 5 o'clock position, the CaO content of the sample collected 30–55 mm from the hot face was 1.57%, exceeding the original level of 1.3 wt%. This indicates that the slag penetration depth in the refractory bricks from the non-weak area was at least 25 mm.
2.3 Microstructural Analysis
Figure 6 shows the microstructure of the post-service refractory bricks from the molten steel bath zone at the 5 o'clock position. The hot-face region of the bricks can be clearly divided into four layers: a slag layer, a reaction layer, a decarburized layer, and a penetration layer. The slag layer was 3.4 mm thick, while the reaction layer ranged from 2.7 to 4.6 mm in thickness. The decarburized layer was 3.4 mm thick. The CaO content remained relatively high even at locations farther from the hot face, indicating a considerable depth of slag penetration. Chemical analysis of the samples presented in Table 5 further revealed that slag penetration occurred throughout the entire 25 mm sample.
The reaction layer on the surface of the refractory bricks from the molten steel bath zone at the 5 o'clock position was relatively thick. The aggregates and matrix were sintered together, forming a dense structure with high strength. This layer impeded further oxidation of the bricks by oxygen and oxides in the slag while also providing a certain degree of resistance to erosion.
Table 5 presents the results of the energy-dispersive X-ray spectroscopy (EDS) analysis. Magnesium aluminate spinel was identified in the reaction layer, indicating that spinel formed through the reaction between alumina and magnesia at high temperatures in highly basic slag. As a high-temperature phase, magnesium aluminate spinel can effectively protect the ladle refractory bricks against further corrosion.
Figure 6. Microstructure of the Post-Service Refractory Bricks at the 5 O'clock Position
A. Slag layer B. Reaction layer C. Decarburized layer D. Penetration layer.
| MgO | Al₂O₃ | SiO₂ | CaO | Fe₂O₃ | |
|---|---|---|---|---|---|
| Zone 1 | 14.43 | 27.88 | 9.14 | 48.55 | 1.25 |
| Zone 2 | 83.58 | 11.23 | 1.76 | 3.43 | - |
| Zone 3 | 69.62 | 15.04 | 9.03 | 4.36 | 1.39 |
Table 5. EDS Analysis of the Different Regions Shown in Figure 6
Figure 7 shows the microstructure of the post-service refractory bricks from the molten steel bath zone at the 12 o'clock position. The slag layer was 0.6 mm thick, the combined thickness of the reaction and decarburized layers was approximately 0.6 mm, and the penetration layer was approximately 2.5 mm thick.
Unlike the microstructure of the post-service ladle wall bricks from the 5 o'clock position, the slag, reaction, decarburized, and penetration layers at the weak point were all very thin, with the original refractory layer reached at a relatively shallow depth. This observation is consistent with the physicochemical analysis presented in Tables 3 and 4. The reaction layer formed between the slag and the refractory was discontinuous and lacked a dense structure.
Table 6 presents the chemical composition of the different regions. The results show that the CaO content of the refractory bricks decreased rapidly with increasing distance from the hot face. In Region 5, the CaO content had already reached the level characteristic of the original refractory layer, indicating a relatively shallow slag penetration depth.
In summary, the post-service refractory bricks from the molten steel bath zone at the 5 o'clock position exhibited distinct reaction and decarburized layers. The reaction layer was relatively dense and continuous, effectively inhibiting further degradation of the refractory by slag and oxygen and reducing erosion by molten steel.
In contrast, a stable reaction layer could not form on the surface of the post-service refractory bricks at the 12 o'clock position. Continuous exposure to the surrounding atmosphere and slag caused the newly formed reaction layer to be rapidly removed by erosion, repeatedly exposing fresh decarburized refractory material to further reactions. Consequently, the refractory underwent a cyclic process of corrosion–oxidation–corrosion/erosion. This degradation pattern is also consistent with descriptions of slag line refractory corrosion reported in numerous previous studies.
Figure 7. Microstructure of the Post-Service Refractory Bricks from the Molten Steel Bath Zone at the 12 O'clock Position
A. Slag layer B. Reaction layer C. Decarburized layer D. Penetration layer E. Original refractory layer.
| MgO | Al₂O₃ | SiO₂ | CaO | Fe₂O₃ | |
|---|---|---|---|---|---|
| Zone 1 | 7.63 | 38.45 | 4.82 | 45.96 | 3.13 |
| Zone 2 | 71.80 | 12.06 | 9.45 | 5.48 | 1.22 |
| Zone 3 | 65.39 | 16.84 | 13.33 | 2.93 | 1.51 |
| Zone 4 | 76.65 | 12.75 | 6.22 | 2.54 | 1.84 |
| Zone 5 | 72.08 | 13.52 | 11.81 | 1.34 | 1.26 |
The formation of a dense reaction layer may be related to the operating practices at the plant. After steel tapping is completed, the ladle is transferred from the continuous casting platform, and the residual slag and molten steel are discharged by tilting the ladle. During slag discharge, the slag discharge surface is covered by slag, intensifying chemical corrosion. However, the slag coverage also provides protection against thermal shock. Meanwhile, the highly basic slag reacts with the refractory bricks to form a relatively dense and continuous reaction layer, which inhibits further attack by air and slag.
At the 12 o'clock position, a thin slag layer adheres to the refractory surface as the slag level decreases during steel tapping. After tapping is completed, the refractory is exposed to oxidation by air and thermal shock, preventing the formation of a dense reaction layer. During the subsequent operating cycle, erosion by molten steel removes the reaction layer, ultimately resulting in an accelerated rate of refractory wear at this location.
3. Conclusions
In this study, the localized accelerated erosion of refractory bricks in the molten steel bath zone of an LF refining ladle at a steel plant was investigated. The erosion morphologies of post-service refractory bricks collected from different locations within the ladle bath zone were systematically analyzed. The following conclusions were drawn.
The metallurgical operating process is the primary factor responsible for the localized accelerated wear of the refractory lining in the LF refining ladle. At the slag discharge surface, coverage by molten slag promotes the formation of a relatively dense reaction layer, which inhibits further attack by air and slag and improves the erosion resistance of the refractory bricks. In contrast, the localized abnormal erosion of refractories at areas not exposed to slag coverage during slag discharge is mainly attributed to the loose and porous structure of the reaction layer formed between the refractory bricks and slag. During the subsequent refining cycle, this layer is more susceptible to erosion by molten steel. With repeated metallurgical operations, the alternating cycle of corrosion–oxidation–corrosion/erosion becomes the primary degradation mechanism of carbon-containing refractory linings.
Based on the above analysis, the following measures are recommended to improve the service life of the refractory lining:
1. Improve the density of the refractory lining in areas not exposed to slag coverage during slag discharge to promote the formation of a denser reaction layer between the refractory bricks and slag.
2. Enhance the high-temperature flexural strength of the refractory materials to improve the erosion resistance of the ladle bricks.
3. Improve the oxidation resistance of the refractory lining in areas not exposed to slag coverage during slag discharge to reduce oxidation-induced degradation during metallurgical operations.