In the field of engineering machinery, components like the upper rotary disc play a critical role in supporting loads and facilitating precise movements. As a key transmission part, the upper rotary disc must exhibit high strength, stability, and wear resistance, often manufactured from materials like gray iron HT300 through sand casting processes. However, the complex geometry of such castings, featuring multiple holes and uneven wall thicknesses, poses significant challenges in avoiding defects like shrinkage porosity and cavities during solidification. Traditional trial-and-error methods in sand casting are time-consuming and costly, necessitating the adoption of numerical simulation techniques to optimize the process efficiently. In this study, we leverage advanced simulation tools to analyze and refine the sand casting process for a gray iron upper rotary disc, aiming to minimize defects and enhance casting quality. By integrating computational modeling with practical sand casting principles, we demonstrate a systematic approach to process improvement, which can be applied to similar industrial components.
The upper rotary disc under investigation is fabricated from gray iron HT300, a material chosen for its excellent mechanical properties and castability. The chemical composition of HT300 is critical to achieving the desired microstructure and performance, as outlined in Table 1. This composition ensures a balance of carbon and silicon to promote graphitization while controlling impurities that could compromise integrity. In sand casting, the selection of appropriate molding materials, such as furan resin sand for both the mold and cores, is essential to withstand the thermal stresses during pouring and solidification.
| Material Grade | C | Si | Mn | P | S |
|---|---|---|---|---|---|
| HT300 | 2.80–3.10 | 1.10–1.40 | 1.00–1.20 | < 0.15 | ≤ 0.12 |
The three-dimensional model of the upper rotary disc reveals an intricate design with outer dimensions of 1281 mm × 1269.7 mm × 101 mm and an estimated weight of 502.41 kg. The varying wall thicknesses, ranging from a maximum of 113 mm to a minimum of 15 mm, create thermal gradients that are prone to defects in sand casting. For instance, thicker sections act as hot spots, leading to isolated liquid zones that result in shrinkage issues. To address this, we initially designed a bottom-gating system for the sand casting process, employing a closed-type configuration with one sprue, two runners, and seven ingates. The cross-sectional area ratio was set at 1.15:1.1:1, corresponding to areas of 8.67 cm², 8.29 cm², and 7.54 cm², respectively, with a pouring time of 47 seconds. This setup aimed to ensure a smooth, controlled filling process, reducing turbulence and oxidation—a common concern in sand casting where metal flow dynamics directly impact defect formation.
Numerical simulation serves as a powerful tool in sand casting to predict and analyze the behavior of molten metal during filling and solidification. For this study, we utilized ProCAST software, specifically the iron (gravity) module, to simulate the process. The pre-processing parameters were carefully defined to mirror real-world sand casting conditions. The pouring temperature was set at 1370°C, and the pouring time was maintained at 47 seconds, based on initial design calculations. The thermophysical properties of the materials, including the gray iron HT300 and the furan resin sand mold, were incorporated into the model to ensure accuracy. Boundary conditions played a crucial role in capturing the heat transfer phenomena; for example, the heat exchange coefficient between metal and sand was specified as $$ h = 500 \, \text{W/(m}^2 \cdot \text{K)} $$, while that between metal and chill (if used) was set to $$ h = 2000 \, \text{W/(m}^2 \cdot \text{K)} $$. These parameters are derived from fundamental heat transfer equations, such as Fourier’s law, where the heat flux \( q \) can be expressed as $$ q = -k \frac{dT}{dx} $$, with \( k \) being the thermal conductivity. In sand casting, the use of insulating materials like riser sleeves further modulates solidification rates, and their properties were integrated into the simulation to reflect practical scenarios.
The meshing of the model involved generating 25,636 two-dimensional elements and 311,860 three-dimensional elements, ensuring sufficient resolution to capture detailed thermal and flow fields. A virtual mold approach was adopted to simplify the simulation, as ProCAST allows for this by defining adiabatic conditions on the outer surfaces. However, the virtual mold dimensions were enlarged to prevent computational inaccuracies, a common practice in sand casting simulations to approximate infinite mold boundaries. The governing equations for fluid flow and heat transfer in sand casting include the Navier-Stokes equations for momentum conservation and the energy equation for temperature distribution. For instance, the energy equation can be written as $$ \rho c_p \frac{\partial T}{\partial t} = \nabla \cdot (k \nabla T) + S $$, where \( \rho \) is density, \( c_p \) is specific heat, \( T \) is temperature, \( t \) is time, and \( S \) represents source terms such as latent heat release during solidification. By solving these equations numerically, the simulation provides insights into potential defects, guiding process optimizations in sand casting.
In the initial sand casting process design, the filling analysis revealed a sequential bottom-up progression of molten metal, with the liquid front advancing steadily from the ingates to the top of the casting. At 25% filling, the metal entered the cavity底部 with minimal冲击, reducing the risk of sand erosion—a frequent issue in sand casting when high-velocity flow occurs. By 50% filling, the metal had spread to the central regions, maintaining a stable flow pattern. At 75% and full filling, the cavity was completely occupied without any short shots, indicating that the gating system was adequate for complete mold filling in this sand casting setup. However, the solidification analysis exposed significant shrinkage porosity and cavities, particularly in the thicker sections like the top surface, base, and central areas. These defects arose due to thermal hotspots, where slower cooling led to isolated liquid pools that could not be fed adequately. The Niyama criterion, often used in casting simulations to predict shrinkage, can be expressed as $$ G / \sqrt{R} $$, where \( G \) is the temperature gradient and \( R \) is the cooling rate. Values below a threshold indicate a high risk of microporosity, which aligned with our observations in the initial sand casting simulation.
To address these issues, we optimized the sand casting process by introducing risers and modifying the gating system. Riser design in sand casting is critical for compensating solidification shrinkage, and we employed both top-neck risers and blind risers based on standard guidelines. The dimensions and quantities are summarized in Table 2 and Table 3, ensuring that the risers act as reservoirs to feed the thick sections during solidification. Additionally, we switched to a top-gating system with one sprue, one runner, and four ingates, having cross-sectional areas of 38.5 cm², 33.0 cm², and 31.7 cm², respectively. This change aimed to reduce flow distance and improve temperature distribution, key factors in minimizing defects in sand casting. The optimization was driven by empirical rules and simulation feedback, such as Chvorinov’s rule for solidification time, $$ t_s = B \left( \frac{V}{A} \right)^2 $$, where \( V \) is volume, \( A \) is surface area, and \( B \) is a mold constant. By placing risers in regions with high \( V/A \) ratios, we ensured that they solidified last, effectively drawing shrinkage away from the casting.
| Quantity | RD | d | RH |
|---|---|---|---|
| 5 | 180 | 90 | 300 |
| Quantity | d | h |
|---|---|---|
| 5 | 80 | 180 |
After implementing these changes, the optimized sand casting process was re-simulated to evaluate its effectiveness. The filling process remained smooth and complete, with no evidence of cold shuts or misruns, confirming the robustness of the top-gating system in sand casting. The solidification sequence showed that areas far from the risers solidified first, followed by the regions beneath the risers, and finally the risers themselves. This progressive solidification is ideal in sand casting, as it allows the risers to supply liquid metal to compensate for shrinkage. The solidification fraction over time demonstrated a controlled pattern, with the risers maintaining liquid metal until the end, thereby reducing defects. Post-optimization defect analysis revealed that nearly all shrinkage porosity and cavities were transferred to the risers, leaving the casting itself largely free of such imperfections. Only one minor defect spot remained in the casting, which is acceptable for practical applications, especially since the risers are positioned on machinable surfaces and can be removed during post-processing. This outcome underscores the value of numerical simulation in sand casting for achieving high-quality components with minimal experimental iterations.
In conclusion, the integration of numerical simulation into the sand casting process for the gray iron upper rotary disc has enabled a significant reduction in defects. Through initial analysis, we identified critical areas prone to shrinkage and implemented strategic modifications, including the addition of risers and an optimized gating system. The simulated results confirm that these adjustments effectively转移 defects to the risers, enhancing the overall integrity of the casting. This approach not only improves the efficiency of sand casting but also reduces costs associated with scrap and rework. For future work, further refinements could explore the use of different molding materials or advanced cooling techniques in sand casting to achieve even better performance. Overall, this study highlights the transformative potential of simulation-driven optimization in modern sand casting practices, paving the way for more reliable and economical manufacturing of complex cast components.