In the mining, construction, and power generation industries, ball mill liners play a critical role in protecting the mill shell from impact and wear caused by grinding media and materials. The performance of these liners directly influences the operational safety, reliability, and service life of ball mills. High manganese steel casting is widely used for liners due to its excellent toughness and work-hardening properties under impact loads. However, high manganese steel casting exhibits high shrinkage susceptibility and low thermal conductivity, leading to a propensity for shrinkage porosity and cavity defects during solidification, especially in lost foam casting (LFC) processes. LFC offers advantages such as design flexibility, high dimensional accuracy, and reduced environmental impact, but it requires careful optimization to mitigate defects. Traditional trial-and-error methods are time-consuming and costly. Therefore, finite element simulation using ProCAST software is employed to evaluate and optimize the LFC process for high manganese steel casting liners, focusing on minimizing defects like shrinkage cavities and porosities.
This study investigates three distinct LFC processes for producing high manganese steel casting liners: Process A (top-gating with eight liners per box), Process B (step-side-gating with ten liners per box), and Process C (side-gating with four liners per box and risers). The objective is to simulate the filling and solidification processes, analyze temperature fields, fluid flow, and solidification behavior, and predict defect formation using criteria such as POROS and Niyama. The simulations aim to identify the optimal process that reduces defects and enhances the quality of high manganese steel casting components. The methodology involves detailed modeling of thermophysical properties, mesh generation, and parameter settings to replicate real-world casting conditions.
The high manganese steel casting material used is ZGMn13, with chemical composition as shown in Table 1. The thermophysical properties, including density, enthalpy, thermal conductivity, and solid fraction, are calculated using ProCAST’s database module, considering a cooling rate of 5°C/s. The EPS foam pattern has a density of 25 kg/m³, and the mold material is resin sand with specific thermal properties. Key simulation parameters include a pouring temperature of 1,420°C, pouring rate of 20 kg/s, and a vacuum pressure of 0.045 MPa. The interface heat transfer coefficients are set to account for interactions between metal, foam, and sand.
| Element | C | Mn | Si | Cr | Ni | V | Al | S | P | Fe |
|---|---|---|---|---|---|---|---|---|---|---|
| Content (wt%) | 1.40 | 13.35 | 0.70 | 2.10 | 0.03 | 0.025 | 0.005 | 0.005 | 0.038 | Bal. |
The mesh generation for the three processes results in varying numbers of tetrahedral elements: 1,352,871 for Process A, 1,575,216 for Process B, and 1,181,657 for Process C. The minimum mesh size is 5 mm at critical regions like liner grooves, ensuring accuracy in simulating fluid flow and heat transfer. The filling process is modeled with dynamic heat transfer between the metal and foam, where the heat transfer coefficient increases to a maximum of 250 W/(m²·K) upon contact. Defect prediction employs the POROS criterion, which identifies isolated liquid regions with shrinkage rates above 1%, and the Niyama criterion, expressed as:
$$ \text{Niyama} = \frac{G}{\sqrt{R}} $$
where \( G \) is the temperature gradient and \( R \) is the cooling rate. A lower Niyama value indicates a higher risk of shrinkage porosity in high manganese steel casting.
The filling process simulation reveals significant differences among the three processes. In Process A, metal flow becomes turbulent at approximately 30% filling, particularly in three liners, due to gas gap formation from foam decomposition. This turbulence persists until 90% filling, with the last-filled regions at the bottom of four liners. Process B exhibits more severe flow disturbances, with chaotic metal advancement in two liners during the latter stages, leading to inconsistent filling. In contrast, Process C demonstrates stable and uniform filling without turbulence, as the fewer liners per box allow efficient gas evacuation. The temperature distribution at the end of filling is most uniform in Process C, whereas Processes A and B show variations up to 50°C between liners, exacerbating defect risks in high manganese steel casting.
Solidification analysis indicates that all processes follow sequential solidification, starting from the liner surfaces and progressing inward. In Processes A and B, the absence of risers results in isolated liquid zones at the centers, forming hot spots that lead to shrinkage defects. Process C, with risers, provides better feeding to the centers, though minor hot spots still occur. The temperature fields during solidification highlight that liner centers remain hotter than edges, with Process B showing the largest temperature disparities. The solid fraction evolution, modeled by the equation:
$$ f_s(T) = \frac{1}{1 + \exp\left(-k(T – T_l)\right)} $$
where \( f_s \) is the solid fraction, \( k \) is a constant, and \( T_l \) is the liquidus temperature, confirms that centerline solidification last in all cases, but the extent of isolation varies. Process C achieves more controlled solidification, reducing defect formation in high manganese steel casting.
Defect prediction using the POROS criterion shows that all processes exhibit shrinkage cavities, but their distribution differs. Process A has concentrated defects in the centers of the last-filled liners, while Process B displays widespread defects, including near the surfaces of the middle six liners, which could impair mechanical performance. Process C confines defects to the core regions, minimizing surface impact. The Niyama criterion further validates these findings, with low \( G/\sqrt{R} \) values in Process B liners indicating high porosity risk, especially at surfaces. Process C maintains higher Niyama values in critical areas, reducing defect propensity. The overall results emphasize that Process C is optimal for high manganese steel casting, as it balances filling stability and solidification control.
| Process | Filling Behavior | Solidification Hot Spots | POROS Defect Location | Niyama Critical Zones |
|---|---|---|---|---|
| A (Top-Gating, 8 liners) | Turbulent flow, gas gaps | Center isolated zones | Concentrated in centers | Low in centers and near surfaces |
| B (Step-Side-Gating, 10 liners) | Highly turbulent, inconsistent | Multiple isolated zones | Dispersed, including surfaces | Very low at surfaces and centers |
| C (Side-Gating with Risers, 4 liners) | Stable, uniform | Minor center hot spots | Confined to cores | Higher in cores, minimal surface risk |
The thermophysical properties of high manganese steel casting, such as density \( \rho(T) \), enthalpy \( H(T) \), and thermal conductivity \( k(T) \), are derived from ProCAST calculations and fitted to polynomial functions for simulation accuracy. For instance, density variation with temperature can be expressed as:
$$ \rho(T) = \rho_0 – \alpha (T – T_0) $$
where \( \rho_0 \) is the initial density, \( \alpha \) is the thermal expansion coefficient, and \( T_0 \) is the reference temperature. Similarly, thermal conductivity follows a relation like \( k(T) = k_0 + \beta T \), where \( k_0 \) and \( \beta \) are material constants. These properties critically influence the solidification morphology and defect formation in high manganese steel casting. The integration of these parameters into the simulation ensures realistic prediction of shrinkage behavior.
In discussion, the superiority of Process C for high manganese steel casting is attributed to its reduced liner count per box, which minimizes gas entrapment and flow instability. The inclusion of risers enhances feeding during solidification, countering the inherent shrinkage of high manganese steel. In contrast, Processes A and B suffer from turbulent filling and inadequate compensation for volume contraction, leading to defects that could compromise liner durability. The finite element approach effectively captures these phenomena, providing a cost-effective tool for optimizing high manganese steel casting processes. Future work could explore variations in pouring parameters or foam properties to further refine the LFC technique for high manganese steel casting applications.
In conclusion, the finite element simulation of lost foam casting for high manganese steel liners demonstrates that Process C, with side-gating and risers for four liners per box, yields the best outcomes. It ensures stable filling, controlled solidification, and minimal defect formation compared to Processes A and B. This study underscores the value of ProCAST in advancing high manganese steel casting technology, reducing reliance on empirical methods, and enhancing product quality in industrial settings.