In continuous casting production of low carbon steel, the inherent characteristics of the steel grade often lead to the occurrence of surface and subsurface slag inclusion defects, while longitudinal surface cracks are less frequent. However, in a domestic steel plant, when the casting speed was increased to 1.8 m/min, a high incidence of longitudinal surface cracks, up to 30%, was observed on the slabs. After rolling, these defects manifested as severe cracks in hot-rolled coils. Through process adjustments such as reducing mold water flow, increasing mold taper, and controlling superheat, the longitudinal crack rate was reduced to approximately 4.5%, and the slag inclusion defect rate to about 5%. Nevertheless, these levels still did not meet customer requirements, necessitating further reduction. Mold flux plays a critical role in continuous casting, encompassing functions such as preventing secondary oxidation, providing thermal insulation, absorbing inclusions, uniformly controlling heat transfer, and improving lubrication. These functions are closely related to slab quality and process stability. Therefore, we designed an optimized mold flux to address both longitudinal cracks and slag inclusion defects in low carbon steel.
The production process for low carbon steel SAE1008 involves: hot metal pretreatment → BOF converter → LF refining → continuous casting. The caster is a straight-arc slab caster designed by Danieli. Key parameters of the caster and the chemical composition of SAE1008 steel are summarized in the following tables.
| Parameter | Value |
|---|---|
| Section Size | 230 mm × (950 – 1650) mm |
| Casting Speed | 1.6 – 1.8 m/min |
| Bending 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 | Control Range (%) | Target Value (%) |
|---|---|---|
| C | 0.03 – 0.07 | 0.05 |
| Mn | 0.2 – 0.5 | 0.35 |
| Si | 0.2 – 0.4 | 0.30 |
| P | 0 – 0.03 | 0.02 |
| S | 0 – 0.03 | 0.015 |
| Als | 0.01 – 0.05 | 0.025 |
The primary defects observed were longitudinal surface cracks on the slabs, concentrated in the middle of the wide faces, with lengths ranging from 500 to 1000 mm. After rolling, the steel plates exhibited numerous defects such as upwarping and black lines, often located 10–500 mm from the edges, with widths of 1–5 mm and lengths of 200–600 mm. These defects are primarily attributed to slag inclusion defect occurrences during casting.
Metallographic analysis of samples taken from the longitudinal cracks revealed that the cracks originated at the surface and extended inward, with depths around 5 mm. The presence of FeO within the cracks and oxidation spots nearby indicated that the cracks formed in the mold during high-temperature conditions. This suggests that uneven initial shell formation at the meniscus region, combined with stress concentration at weak points, led to crack initiation. Key factors contributing to this include: thin liquid slag layer of the mold flux, which caused non-uniform heat transfer; high casting speed reducing slag film thickness; and biased flow from the submerged entry nozzle (SEN). The slag inclusion defect was analyzed using scanning electron microscopy and energy-dispersive spectroscopy, revealing that the inclusions mainly consisted of oxides of Si, Al, Ca, Fe, Na, and Mg, with small amounts of F. The presence of Na and F confirmed that these were mold flux entrapments, likely due to shear-type slag entrainment at the steel-slag interface near the narrow face, exacerbated by SEN flow patterns.
The design of mold flux for simultaneously mitigating longitudinal cracks and slag inclusion defect presents a significant challenge. To reduce longitudinal cracks, the flux must enhance heat transfer control, typically by increasing basicity and CaF2 content to promote crystallization and increase slag film roughness. However, this can degrade lubrication, leading to issues like large slag bands and breakouts. Conversely, to minimize slag inclusion defect, higher viscosity is needed to increase surface tension and reduce entrapment, but at high casting speeds (1.8 m/min), viscosity above 0.25 Pa·s can trigger alarms and breakouts. Thus, a balance is required. We introduced Li2O and MnO to address this矛盾. Li2O lowers melting point and viscosity while promoting glassy slag films, improving lubrication. MnO delays crystallization, increases crystallization temperature, and affects heat transfer by coloring the slag film. We conducted experiments to study their effects on physical properties, as shown in the following formulas and tables.
The viscosity of mold flux can be described by the Arrhenius equation: $$\eta = A \exp\left(\frac{B}{T}\right)$$ where $\eta$ is viscosity, $T$ is temperature in Kelvin, and $A$ and $B$ are constants dependent on composition. Adding Li2O decreases $B$, reducing viscosity, while MnO modifies the activation energy for crystallization. The relationship between composition and melting point can be approximated by: $$T_m = \sum_{i} x_i \cdot T_{m,i} + \Delta H_{mix}$$ where $x_i$ is the mole fraction of component $i$, $T_{m,i}$ is its pure component melting point, and $\Delta H_{mix}$ is the mixing enthalpy.
We tested variations in Li2O and MnO content while keeping other components constant. The results are summarized below:
| Li2O Content (%) | Melting Point (°C) | Viscosity at 1300°C (Pa·s) |
|---|---|---|
| 0 | 1100 | 0.18 |
| 1 | 1085 | 0.16 |
| 2 | 1070 | 0.14 |
| 3 | 1055 | 0.12 |
| 4 | 1040 | 0.10 |
| MnO Content (%) | Melting Point (°C) | Viscosity at 1300°C (Pa·s) |
|---|---|---|
| 0 | 1100 | 0.18 |
| 1.5 | 1080 | 0.17 |
| 3.0 | 1065 | 0.16 |
| 3.5 | 1060 | 0.15 |
| 4.0 | 1058 | 0.15 |
Based on these findings, we selected 3% Li2O and 3.5% MnO for the optimized flux, designated as Flux B, compared to the original Flux A. The physical and chemical properties are detailed in the following tables.
| Property | Flux A | Flux 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 |
| Component | Flux A (%) | Flux 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 |
The increase in melting and crystallization temperatures for Flux B enhances heat transfer control, reducing thermal gradients and promoting uniform shell growth. The higher viscosity, balanced by Li2O’s lubricating effect, helps minimize slag entrapment. The liquid slag layer thickness is expected to increase due to reduced flux infiltration into the mold-strand gap, further mitigating slag inclusion defect.
Industrial trials were conducted on a No. 2 caster under standard operating procedures. We used 5 tons of Flux B over 7 heats of SAE1008 steel. During the trials, Flux B exhibited good spreadability and fluidity in the mold. The total slag layer was 50–70 mm, with liquid slag thickness increasing from 9–11 mm (Flux A) to 14–17 mm. This improved isolation of the steel from the powder layer, reducing slag inclusion defect potential. 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 better heat transfer control. Mold surface flames were moderate, and flux consumption averaged 0.35–0.4 kg/t, ensuring adequate lubrication. No significant slag bands or mold alarms occurred.
Slab surface quality was assessed after conditioning. Slabs cast with Flux B showed no long longitudinal cracks. After rolling, the longitudinal crack rate dropped from 4.5% to 0.2%, and the comprehensive slag inclusion defect rate decreased from 5% to 0.3%, meeting customer requirements. This confirms that Flux B effectively addresses both defect types.
To further elaborate on the mechanisms, we consider the heat transfer through the slag film. The heat flux $q$ can be expressed as: $$q = \frac{T_{shell} – T_{mold}}{R_{total}}$$ where $T_{shell}$ is the shell surface temperature, $T_{mold}$ is the mold temperature, and $R_{total}$ is the total thermal resistance of the slag film. $R_{total}$ comprises conductive resistance through the liquid and crystalline layers: $$R_{total} = \frac{d_l}{k_l} + \frac{d_c}{k_c}$$ where $d_l$ and $d_c$ are thicknesses, and $k_l$ and $k_c$ are thermal conductivities of the liquid and crystalline layers, respectively. By increasing crystallization, Flux B raises $k_c$ and adjusts $d_c$, thereby modulating $q$ to reduce thermal stress and crack formation.
Regarding slag entrapment, the critical velocity for shear entrainment $v_c$ can be estimated using: $$v_c = \sqrt{\frac{2 \sigma_{slag} g \Delta \rho}{\rho_{steel}^2}}$$ where $\sigma_{slag}$ is the slag surface tension, $g$ is gravity, $\Delta \rho$ is the density difference between steel and slag, and $\rho_{steel}$ is steel density. Higher viscosity increases $\sigma_{slag}$, raising $v_c$ and making entrainment less likely. However, excessive viscosity hinders lubrication. Our design optimizes this balance.
We also analyzed the effect of Li2O and MnO on crystallization kinetics. The crystallization rate constant $k$ can be modeled by: $$k = k_0 \exp\left(-\frac{E_a}{RT}\right)$$ where $k_0$ is a pre-exponential factor, $E_a$ is activation energy, $R$ is the gas constant, and $T$ is temperature. Li2O reduces $E_a$ for glass formation, while MnO increases $E_a$ for crystallization, delaying it. This combination allows for a slag film with a glassy layer near the strand for lubrication and a crystalline layer near the mold for heat control.
In summary, the simultaneous occurrence of longitudinal cracks and slag inclusion defect in high-speed casting of low carbon steel stems from imbalances in mold flux properties. The thin liquid slag layer and non-uniform heat transfer promote cracks, while shear flows at the steel-slag interface cause slag entrapment. The addition of Li2O (3%) and MnO (3.5%) to the mold flux formulation effectively reconciles lubricity and heat transfer control. Li2O lowers viscosity and promotes glassy behavior, enhancing lubrication, while MnO increases crystallization temperature and modulates heat transfer. Industrial trials demonstrated that the optimized flux reduces longitudinal crack rate from 4.5% to 0.2% and slag inclusion defect rate from 5% to 0.3%, significantly improving slab quality and process stability. This approach highlights the importance of tailored flux design in mitigating complex defect patterns like slag inclusion defect in continuous casting.