In the field of mechanical engineering, structural components like beams are critical for applications in machine tools, cranes, and other heavy-duty equipment. These sand casting products must withstand significant loads, demanding high quality in terms of internal integrity and mechanical properties. Steel castings, particularly medium-carbon steels, offer an excellent balance of strength, toughness, and cost-effectiveness, making them ideal for such applications. However, the casting of steel presents challenges due to its high melting point, narrow solidification range, high volumetric shrinkage, and poor fluidity, often leading to defects such as shrinkage porosity, hot tears, and stress concentrations. Traditional casting process design relies heavily on empirical knowledge and iterative trials, which are time-consuming and costly. To address this, numerical simulation has emerged as a powerful tool for predicting defects, optimizing processes, and enhancing the efficiency of producing sand casting products. In this study, I employ numerical simulation to analyze and optimize the sand casting process for a steel beam, aiming to minimize defects and improve quality.
The beam casting, as a representative sand casting product, has a relatively simple geometry with uniform wall thickness. Its overall dimensions are approximately 1900 mm in length, 260 mm in width, and 190 mm in height, classifying it as a medium-sized casting. For mass production, sand casting with self-hardening sodium silicate sand was selected. To improve productivity, a two-cavity mold layout was adopted. The material chosen is ZG270-500 (equivalent to ZG35), a medium-carbon steel with the chemical composition detailed in Table 1. The composition influences solidification behavior, particularly the peritectic transformation, which contributes to significant linear shrinkage and susceptibility to defects. The casting must be free from cold shuts, cracks, shrinkage cavities, and through-defects, adhering strictly to the MC2000 ultrasonic testing standard to ensure internal quality. This emphasis on quality underscores the importance of advanced design methods for sand casting products.
| Element | C | Si | Mn | P | S | Mo | Cr | Ni | Fe |
|---|---|---|---|---|---|---|---|---|---|
| Content | 0.32–0.42 | 0.20–0.45 | 0.50–0.80 | ≤0.04 | ≤0.04 | ≤0.20 | ≤0.35 | ≤0.30 | Bal. |
The initial casting process was designed using conventional methods. Given the characteristics of steel, the gating system must be simple with large cross-sectional areas to ensure rapid and smooth filling, typically employing a pouring basin (ladle) for open gating. The diameter of the ladle nozzle was set at 45 mm based on the casting weight. The gating system components were sized proportionally: the total cross-sectional area ratio of the nozzle, ingate, runner, and sprue was set as ∑Anozzle : ∑Aingate : ∑Arunner : ∑Asprue = 1 : 1.9 : 1.9 : 2.2. The dimensions are summarized in Table 2. For riser design, thermal hotspots were identified at the beam ends and reinforcing ribs. The modulus method was applied, with end moduli calculated as 1.9 cm and rib moduli as 2.4 cm. Accordingly, circular top risers (diameter: 120 mm, height: 180 mm) were placed at the ends, and side risers (diameter: 140 mm, height: 280 mm) were used for the ribs to serve both cavities. This initial setup aimed to achieve directional solidification, but as with many sand casting products, theoretical designs often require validation.
| Component | Cross-Sectional Shape | Dimensions (mm) |
|---|---|---|
| Sprue | Circular | Diameter: 60 |
| Runner | Trapezoidal | Top: 70, Bottom: 50, Height: 60 |
| Ingate | Rectangular | Width: 40, Height: 30 |
To evaluate the initial process, numerical simulation was conducted using ViewCast software. The 3D model of the casting and mold assembly was created and imported as an STL file. Mesh generation was performed with approximately 2 million elements to balance accuracy and computational time—a critical consideration in simulating sand casting products. The casting parameters were set as follows: pouring temperature of 1550°C, mold initial temperature of 20°C, and material properties assigned based on ZG270-500 steel and sand mold data. The governing equations for solidification simulation include the heat transfer equation:
$$ \rho c_p \frac{\partial T}{\partial t} = \nabla \cdot (k \nabla T) + Q $$
where \( \rho \) is density, \( c_p \) is specific heat, \( T \) is temperature, \( t \) is time, \( k \) is thermal conductivity, and \( Q \) represents latent heat release during phase change. The simulation tracked the liquid fraction over time to predict shrinkage defects. The results for the initial process revealed the solidification sequence: edges solidified first at 172 s, followed by the gating system at 352 s, thin walls between ribs and ends at 712 s, and isolated liquid pockets in the beam center by 981 s. By 1342 s, solidification nearly completed, but residual liquid remained at riser necks, indicating inadequate feeding. Defect prediction, as shown in Figure 5 of the reference, highlighted shrinkage porosity in areas beyond riser feeding distances and at riser junctions, common issues in sand casting products when补缩 is insufficient.
Based on the simulation analysis, the process was optimized to enhance the quality of sand casting products. Two key modifications were implemented: First, chill plates were added in the central region and thin-wall sections to locally increase cooling rates and extend the feeding range of risers. The chills, made of cast iron, were positioned to promote directional solidification toward the risers. Second, the conventional risers were replaced with insulated sleeve risers to improve feeding efficiency by reducing heat loss. The optimized riser dimensions were adjusted based on recalculated moduli, with side risers enlarged to a diameter of 150 mm and height of 300 mm. The placement of chills and risers is summarized in Table 3. These changes aimed to achieve better thermal control, a crucial aspect in producing defect-free sand casting products.
| Element | Type | Dimensions/Details | Purpose |
|---|---|---|---|
| Top Riser | Insulated Sleeve | Diameter: 130 mm, Height: 190 mm | Enhanced feeding at ends |
| Side Riser | Insulated Sleeve | Diameter: 150 mm, Height: 300 mm | Improved补缩 for ribs |
| Chill Plates | Cast Iron | Thickness: 20 mm, Placed centrally | Accelerate cooling, extend feeding |
| Gating System | Unchanged | As per initial design | Maintain filling characteristics |
The optimized process was simulated again under identical conditions. The filling simulation showed smooth and orderly mold filling, completing in 12.7 s, which meets the requirement for rapid pouring in steel sand casting products. The solidification simulation demonstrated a significant improvement: chills accelerated cooling in critical areas, leading to a more controlled sequence. Liquid fraction plots indicated that solidification progressed from the chills outward, with isolated liquid zones forming only within risers by 1478 s. By 1598 s, the final solidification was confined to the risers, confirming effective directional solidification. Defect prediction for the optimized process, as seen in Figure 9 of the reference, showed a drastic reduction in shrinkage porosity, with only minor defects on the bottom surface that could be addressed during machining. This outcome highlights the value of simulation in refining processes for sand casting products.
To validate the simulation results, actual castings were produced using the optimized process. The sand casting products were inspected via ultrasonic testing, revealing no significant internal defects. For mechanical and microstructural analysis, samples were taken from attached test blocks. Hardness measurements yielded an average HRC of 50.8, with values of 50, 47, 53, 52, and 52, all meeting the specified requirements. Microstructural examination, after etching with 4% nitric alcohol solution, revealed a typical hypoeutectoid steel structure consisting of ferrite and pearlite. However, some Widmanstätten morphology was observed, where acicular ferrite penetrated into pearlite colonies. This structure, resulting from coarse as-cast grains and relatively fast cooling, can reduce toughness and impact strength. It can be eliminated through subsequent heat treatments like normalizing or forging, but for as-cast sand casting products, it indicates the need for controlled cooling rates. The successful production underscores how simulation-driven optimization can elevate the quality of sand casting products.
The benefits of numerical simulation extend beyond defect reduction. For instance, the process yield improved from an initial estimate to 72% in the optimized setup, reducing material waste. Key formulas used in optimization include the modulus calculation for riser design:
$$ M = \frac{V}{A} $$
where \( M \) is the modulus (cm), \( V \) is volume (cm³), and \( A \) is surface area (cm²). For feeding distance estimation, an empirical relation for steel castings in sand molds is:
$$ L_f = k \cdot \sqrt{T} $$
where \( L_f \) is the feeding distance (mm), \( k \) is a material constant (approximately 15 for carbon steels), and \( T \) is section thickness (mm). Incorporating chills modifies this to:
$$ L_{f,chill} = L_f + \Delta L_c $$
where \( \Delta L_c \) represents the extended distance due to chilling effect, often derived from simulation data. These mathematical approaches complement simulation in designing robust processes for sand casting products.
In conclusion, this study demonstrates the effective integration of numerical simulation into the sand casting process for steel beams. By analyzing initial designs with ViewCast software, I identified defect-prone areas and optimized the process through chill additions and insulated risers. The optimized process achieved orderly filling, directional solidification, and a significant reduction in shrinkage defects, as confirmed by both simulation and实际 production. The resulting sand casting products exhibited satisfactory internal quality and mechanical properties, with a process yield of 72%. This approach not only enhances the reliability of sand casting products but also reduces development time and costs. Future work could explore advanced simulation aspects like stress analysis or multi-objective optimization to further improve the performance of sand casting products in demanding applications.