The evaluation of the lost foam casting process for high manganese steel liners, aiming to reduce the tendency of shrinkage porosity and shrinkage cavities. Three different casting processes (Process A, B, and C) were simulated using ProCAST software, and their filling and solidification behaviors were analyzed. The results were summarized in tables and figures to provide a clear comparison.
1. Introduction
Ball mills play a crucial role in mineral processing, building materials, and power industries. The liner, as the core component of ball mills, is subjected to complex forces and thus experiences significant wear. High manganese steel is commonly used for liners due to its excellent plasticity, toughness, and work-hardening ability. However, the high shrinkage rate and low thermal conductivity of high manganese steel increase the tendency of shrinkage defects during casting. Lost foam casting, characterized by high design flexibility, precision, and low cost, was chosen for this study. Yet, it requires careful process optimization to avoid defects.
2. Methodology
2.1. 3D Modeling and Simulation Parameters
The liner castings, approximately rectangular with dimensions of 590 mm × 340 mm and a thickness ranging from 80 mm to 120 mm, were modeled. Three casting processes were designed, as summarized in Table 1.
Table 1. Comparison of Casting Processes
| Process | Pouring Position | Number of Liners per Box | Riser |
|---|---|---|---|
| A | Top | 8 | No |
| B | Step Side | 10 | No |
| C | Side | 4 | Yes |
The chemical composition of the high manganese steel liners is shown in Table 2.
Table 2. Chemical Composition of High Manganese Steel Liners (% w)
| Element | C | Mn | Si | P | S | Cr | Ni | V | Al | Fe |
|---|---|---|---|---|---|---|---|---|---|---|
| Content | 1.40 | 13.35 | 0.70 | 0.038 | 0.005 | 2.10 | 0.03 | 0.025 | 0.005 | Balance |
2.2. Mesh Generation and Process Parameter Settings
The 3D CAD models were imported into ProCAST for mesh generation. The minimum mesh size was 5 mm near the liner grooves, and the maximum was 100 mm at the sand box. The number of tetrahedral meshes for each process is summarized in Table 3.
Table 3. Mesh Information for Each Process
| Process | Number of Tetrahedral Meshes |
|---|---|
| A | 1,352,871 |
| B | 1,575,216 |
| C | 1,181,657 |
Pouring temperature was set at 1420 °C, pouring speed at 20 kg/s, and ambient temperature at 20 °C. The heat transfer coefficients between the metal and the mold materials were adjusted based on the distance.
2.3. Simplification of Heat Transfer during Filling
The heat transfer between the molten metal and the foam pattern was simplified in ProCAST, with adjustments based on the distance between them.
2.4. Defect Prediction Criteria
The POROS criterion and Niyama criterion were used to predict shrinkage porosity and cavities.
3. Simulation Results and Analysis
3.1. Filling Process
The filling processes for the three casting processes.
Turbulence was observed in Processes A and B, likely due to gas generation and complex heat exchange. Process C showed a smooth filling process.
3.2. Solidification Process
The solidification processes for the three casting processes.
All processes showed sequential solidification starting from the edges. Processes A and B lacked risers, leading to hot spots in the center. Process C had a riser, reducing defects but still showing hot spots.
3.3. Defect Prediction
The defect predictions using the POROS criterion and Niyama criterion are shown in Tables 4 and 5, respectively.
Table 4. Defect Prediction using POROS Criterion
| Process | Defect Location | Defect Tendency |
|---|---|---|
| A | Center of liners | Moderate |
| B | Surface and center of liners | High |
| C | Center of liners | Low |
Table 5. Defect Prediction using Niyama Criterion
| Process | Defect Location | G/R Ratio | Defect Tendency |
|---|---|---|---|
| A | Center, close to surface | Low | Moderate |
| B | Surface and inner surface | Low | High |
| C | Center | Higher than A and B | Low |
4. Production Validation
Castings produced using Process C were inspected, and the defect distribution was consistent with the simulation results, with defects concentrated in the center.
5. Conclusion
The lost foam casting process for high manganese steel liners was optimized using ProCAST software. Process C, with side pouring and a riser, showed the best performance, with defects concentrated in the center and a smooth filling process. This study demonstrates the effectiveness of finite element simulation in optimizing lost foam casting processes.