In the continuous casting process for low carbon steel, achieving high casting speeds often leads to significant quality issues, particularly surface longitudinal cracks and slag inclusions. As a researcher focused on metallurgical materials, I encountered a scenario where a domestic steel plant operated at a casting speed of 1.8 m/min, resulting in a high incidence of these defects. The surface longitudinal crack rate reached up to 30%, and after rolling, the hot-rolled coils exhibited severe defects like peeling and black lines, primarily due to slag inclusions. Through preliminary adjustments such as reducing mold water flow and controlling superheat, the crack rate was reduced to 4.5% and the slag inclusion rate to 5%, but this still fell short of customer requirements. This prompted an in-depth investigation into the role of casting powder in mitigating these issues, leading to the design of an optimized powder that balances thermal control and lubrication. In this article, I will detail the analysis, design, and industrial validation of this improved casting powder, emphasizing the reduction of slag inclusions and surface cracks.
The production route for low carbon steel, such as SAE1008, involves iron pre-treatment, basic oxygen furnace (BOF) refining, ladle furnace (LF) treatment, and continuous casting (CC). The caster is a straight-arc slab machine with key parameters summarized in Table 1. The steel composition targets are listed in Table 2. At high casting speeds, the interplay between mold powder functions—thermal insulation, lubrication, and inclusion absorption—becomes critical. Defects like surface longitudinal cracks and slag inclusions often arise from imbalances in these functions, necessitating a tailored approach to powder design.
| Parameter | Value |
|---|---|
| Section Size | 230 mm × (950–1650) mm |
| Casting Speed | 1.6–1.8 m/min |
| Radius | 10.5 m |
| Metallurgical Length | 38.5 m |
| Wide Face Water Flow | 5800 L/min |
| Narrow Face Water Flow | 550 L/min |
| Taper | 1.15% |
| Element | Content (%) |
|---|---|
| C | 0.05 |
| Mn | 0.35 |
| Si | 0.30 |
| P | 0.02 |
| S | 0.015 |
| Als | 0.025 |
The occurrence of defects was systematically analyzed. Surface longitudinal cracks, typically 500–1000 mm long, were concentrated on the wide faces of the slab. Metallographic examination revealed that these cracks originated in the mold, with oxidation and decarburization features indicating high-temperature formation. The primary causes were identified as non-uniform heat transfer due to thin liquid slag layers and high casting speed, which thins the slag film and exacerbates stress concentrations in the initial shell. Additionally, biased flow from the submerged entry nozzle (SEN) contributed to both cracks and slag inclusions. Regarding slag inclusions, defects like peeling and black lines on rolled sheets were analyzed using scanning electron microscopy (SEM) and energy-dispersive spectroscopy (EDS). The inclusions contained elements such as Si, Al, Ca, Fe, Na, and F, confirming their origin from entrained mold powder. These slag inclusions were primarily caused by shear-induced entrapment at the steel-slag interface, especially near the narrow faces where flow patterns promote slag droplet incorporation.
The presence of slag inclusions is a persistent challenge in continuous casting, as they degrade surface quality and lead to downstream defects. In this case, the slag inclusions were characterized by high levels of oxides and fluorides, typical of mold powder entrapment. The mechanism involves turbulent flow at the meniscus, where slag droplets are sheared into the steel and captured by the solidifying shell. To quantify this, the inclusion composition is summarized in Table 3, highlighting the prevalence of slag-related elements. Reducing slag inclusions requires optimizing powder properties to minimize entrapment while maintaining other functions.
| Element | Weight Percentage (%) | Atomic Percentage (%) |
|---|---|---|
| O | 30.48 | 46.81 |
| Fe | 14.66 | 6.45 |
| Na | 10.43 | 11.14 |
| Mg | 5.05 | 5.10 |
| Al | 6.47 | 5.89 |
| Si | 16.90 | 14.78 |
| Ca | 16.02 | 9.82 |
Designing an effective casting powder for high-speed casting involves addressing the contradiction between thermal control and lubrication. For low carbon steel, which has minimal shell shrinkage, good lubrication is essential to prevent stickers and slag inclusions. However, to reduce surface longitudinal cracks, enhanced thermal control is needed to promote uniform shell growth. This is typically achieved by increasing basicity and CaF2 content to raise crystallization temperature and roughness of the slag film, but this can impair lubrication. The design challenge was to balance these aspects without causing operational issues like breakouts. Based on experience, the viscosity should not exceed 0.22 Pa·s at 1300°C, and basicity should be capped at 1.2 to avoid lubrication failures. To overcome these limitations, I explored the addition of Li2O and MnO, which modify powder properties favorably.
The effects of Li2O and MnO on powder characteristics were investigated through laboratory tests. Li2O is known to lower melting point and viscosity while promoting glassy slag films, thereby improving lubrication. MnO can delay crystallization and enhance thermal control by coloring the slag film, affecting heat transfer. To quantify these effects, I varied the content of these oxides in a base powder and measured melting point and viscosity. The results are summarized in Table 4 and depicted through mathematical relationships. The melting point (Tm) as a function of oxide content can be expressed as:
$$ T_m = T_0 – k_L \cdot w_{Li_2O} – k_M \cdot w_{MnO} $$
where \( T_0 \) is the base melting point, \( k_L \) and \( k_M \) are coefficients, and \( w \) denotes weight percentage. Similarly, viscosity (\(\eta\)) at 1300°C follows:
$$ \eta = \eta_0 \cdot \exp\left(-\alpha \cdot w_{Li_2O} – \beta \cdot w_{MnO}\right) $$
with \( \eta_0 \) as base viscosity, and \( \alpha \), \( \beta \) as constants. From experiments, Li2O significantly reduced both parameters, while MnO had a diminishing effect above 3.5%. Therefore, I selected 3% Li2O and 3.5% MnO for optimization.
| Oxide Content (%) | Melting Point (°C) | Viscosity at 1300°C (Pa·s) |
|---|---|---|
| 0% Li2O, 0% MnO | 1100 | 0.18 |
| 3% Li2O, 0% MnO | 1080 | 0.15 |
| 0% Li2O, 3.5% MnO | 1090 | 0.17 |
| 3% Li2O, 3.5% MnO | 1075 | 0.14 |
Based on this, I designed an improved casting powder (Powder B) compared to the original (Powder A). The key was to increase melting and crystallization temperatures for better thermal control while maintaining adequate lubrication through adjusted viscosity. The composition and properties are detailed in Table 5 and Table 6. Powder B is a pre-fused hollow granular type with higher basicity and controlled oxide additions. The crystallization temperature was raised to enhance the slag film’s ability to regulate heat transfer, reducing shell unevenness and crack initiation. Concurrently, the viscosity was optimized to ensure sufficient slag inflow for lubrication without promoting slag inclusions. The relationship between viscosity and temperature is critical; for Powder B, the viscosity-temperature curve shows a steeper slope, indicating improved stability at high speeds. This can be modeled as:
$$ \ln(\eta) = A + \frac{B}{T} $$
where \( A \) and \( B \) are material constants, and \( T \) is temperature in Kelvin. For Powder B, \( B \) is higher, reflecting better crystallization behavior.
| Component | Powder A | Powder B |
|---|---|---|
| R (Basicity, CaO/SiO2) | 0.95 | 1.20 |
| SiO2 | 31.4 | 29.5 |
| MgO | 2.3 | 1.5 |
| CaO | 29.8 | 35.3 |
| Al2O3 | 5.5 | 6.4 |
| Li2O | 0 | 3.0 |
| MnO | 0 | 3.5 |
| Na2O | 8.5 | 4.5 |
| F– | 6.5 | 3.2 |
| Property | Powder A | Powder B |
|---|---|---|
| Melting Temperature (°C) | 1100 | 1130 |
| Crystallization Temperature (°C) | 1120 | 1195 |
| Crystallization Rate (%) | 0 | 30 |
| Viscosity at 1300°C (Pa·s) | 0.18 | 0.22 |
| Melting Speed (s) | 25 | 26 |
| Carbon Content (%) | 4.5 | 5.0 |
The industrial trial was conducted on a caster under standard operating conditions for SAE1008 steel. Powder B was used over multiple heats, totaling 5 tons. During casting, the powder exhibited excellent spreadability and fluidity in the mold. The total slag layer thickness was 50–70 mm, with a liquid slag layer of 14–17 mm, significantly thicker than the 9–11 mm with Powder A. This improved isolation of the steel from the powder layer, reducing the risk of slag inclusions. Heat flux measurements showed a decrease from 1.8 MW/m² to 1.5 MW/m² on the wide faces and from 1.7 MW/m² to 1.4 MW/m² on the narrow faces, indicating enhanced thermal control. The powder consumption averaged 0.35–0.40 kg/t, ensuring adequate lubrication without excessive usage. No sticker alarms or breakouts occurred, and slag rims were minimal, confirming good operational stability.
The quality of the cast slabs was rigorously assessed. Surface inspection revealed a dramatic reduction in longitudinal cracks. The crack rate dropped from 4.5% to 0.2%, as quantified over the trial heats. More importantly, the incidence of slag inclusions on rolled sheets was drastically reduced. Defects like peeling and black lines, previously linked to slag inclusions, decreased from 5% to 0.3%. This improvement is attributed to the optimized powder’s ability to balance lubrication and thermal control, minimizing both crack initiation and slag entrapment. The mechanisms can be summarized using a defect reduction model:
$$ \text{Defect Rate} = f(\eta, T_c, H) $$
where \( \eta \) is viscosity, \( T_c \) is crystallization temperature, and \( H \) is liquid slag layer thickness. For Powder B, higher \( T_c \) and \( H \) reduce thermal stress and slag inclusion probability, while controlled \( \eta \) maintains lubrication. The success underscores the importance of tailored powder design in addressing simultaneous defects like surface cracks and slag inclusions.
In conclusion, the integration of Li2O and MnO into casting powder formulation effectively resolves the conflict between thermal control and lubrication at high casting speeds. The optimized powder enhances crystallization behavior for uniform shell growth, reducing surface longitudinal cracks, while its viscosity and slag layer management mitigate slag inclusions. Industrial trials validated these benefits, with crack rates falling to 0.2% and slag inclusion rates to 0.3%. This approach not only improves slab quality but also ensures casting process stability, meeting stringent customer demands. Future work could explore dynamic powder adjustment models based on real-time casting parameters to further minimize defects like slag inclusions in varying operational conditions.
To generalize, the reduction of slag inclusions is a multifaceted challenge requiring careful control of powder physicochemical properties. The key parameters include viscosity, basicity, and oxide additions, which influence slag-steel interaction and heat transfer. From this study, the following empirical correlations were derived for defect prediction:
$$ \text{Slag Inclusion Index} = k_1 \cdot \eta^{-0.5} + k_2 \cdot R^{1.2} $$
where \( k_1 \) and \( k_2 \) are constants, \( \eta \) is viscosity, and \( R \) is basicity. Lower values indicate fewer slag inclusions. Similarly, for crack susceptibility:
$$ \text{Crack Susceptibility} = \frac{\Delta T}{\eta \cdot H} $$
with \( \Delta T \) as temperature gradient. Minimizing this ratio through powder design leads to better outcomes. These insights can guide future innovations in continuous casting powders, particularly for high-speed applications where defects like slag inclusions remain a critical concern.