In the field of modern manufacturing, the sand casting foundry process remains one of the most versatile and cost-effective methods for producing complex aluminum alloy components. The aluminum alloy cover under investigation in this study is a key part used in aerospace hydraulic systems, requiring high dimensional accuracy, excellent mechanical properties, and absolute leak-tightness under elevated pressures. However, due to its intricate geometry with numerous stiffening ribs, varying wall thicknesses, and internal cavities, the sand casting foundry process often suffers from shrinkage porosity, gas entrapment, and incomplete filling. This article presents a comprehensive numerical simulation and systematic process optimization of the sand casting foundry operation for this specific cover, using the AnyCasting software package. The primary objective is to eliminate internal defects and improve the overall casting quality, thereby reducing the rejection rate in mass production. Through detailed analysis of flow behavior, solidification sequence, and defect formation mechanisms, a series of modifications to the gating system, chilling strategy, and riser placement are proposed. The optimized sand casting foundry parameters are validated by both simulation predictions and actual production trials.
| Parameter | Value |
|---|---|
| Material | AlSi10Mg (Al-10%Si-0.3%Mg) |
| Overall dimensions (mm) | 263 × 220 × 103 |
| Average wall thickness (mm) | 5 |
| Minimum wall thickness (mm) | 3 (rib regions) |
| Mold material | Alkaline phenolic resin sand |
| Core material | Cold-box resin sand |
| Pouring temperature (°C) | 690 – 710 |
| Initial pouring time (s) | 3.5 |
| Metal – mold heat transfer coefficient (W/(m²·K)) | 420 |
| Gating system type | Bottom-gated, unpressurized |
The sand casting foundry process begins with a thorough analysis of the original casting design. The cover features a rectangular base and a circular top surface connected by a curved body. Numerous longitudinal and transverse ribs are distributed on the exterior to enhance stiffness while maintaining low weight. The interior is hollow, requiring a complex sand core to form the internal geometry. In the original sand casting foundry layout, the cover was positioned with the rectangular base at the top, and the gating system introduced metal from the bottom of the rectangular face. A single riser was placed on the circular top surface. The initial simulation using AnyCasting with approximately 1.3 million uniform cubic cells revealed severe hot spots at the junctions between the ribs and the main body. Figure 1 (inserted later) illustrates the typical sand casting foundry setup.
The solidification sequence in the original sand casting foundry scheme showed that the thin ribs solidified first, while the thicker rib roots and the central bottom region remained liquid for a longer time. This non-uniform solidification led to isolated liquid islands that could not be fed by the riser, resulting in shrinkage porosity. The defect prediction module of AnyCasting, based on the Niyama criterion and the temperature gradient method, confirmed that the most critical defects appeared in the rib roots and the lower part of the cover far from the riser. The Niyama criterion is defined as:
$$ N = \frac{G}{\sqrt{R}} $$
where \(G\) is the temperature gradient (K/mm) and \(R\) is the cooling rate (K/s). A low Niyama value (\(N < 1\)) indicates a high risk of shrinkage porosity. In the initial sand casting foundry simulation, the Niyama values in the defective regions were as low as 0.2–0.5, confirming the susceptibility to microporosity. Additionally, the fill simulation showed that the molten metal entering from the bottom had to travel upward through the narrow rib channels, causing turbulence and potential gas entrapment. The core, being completely surrounded by metal, had limited gas venting, further increasing the risk of blowholes.
| Parameter | Initial design | Optimized design |
|---|---|---|
| Casting orientation | Rectangular base at top | Rectangular base at bottom (cover opening upward) |
| Gating type | Single bottom gate (rectangular side) | Two ingates from rectangular base edges |
| Cross-sectional area ratio (sprue:runner:ingate) | 1:1.2:1.2 | 1:2:1.5 |
| Choke area (sprue bottom, cm²) | 2.54 | 2.54 (unchanged) |
| Number of risers | 1 (circular top) | 2 (circular top) |
| Chills | None | Internal chills at rib roots + bottom plate |
| Pouring time (s) | 3.5 | 2.5 |
| Filter | None | Filter at sprue base |
Based on the initial simulation results, a comprehensive optimization of the sand casting foundry process was undertaken. The most significant change was the reversal of the casting orientation: the rectangular base was placed at the bottom, and the cover opening faced upward. This allowed the internal sand core to vent gases upward during filling and solidification, drastically reducing the gas porosity risk. The gating system was redesigned as a semi-open system with the cross-sectional area ratio of sprue:runner:ingate = 1:2:1.5. Two ingates were positioned symmetrically at the edges of the rectangular base, promoting uniform filling from the bottom upward. The choke area remained at 2.54 cm² at the sprue bottom. To accelerate filling and minimize temperature drop, the pouring time was reduced from 3.5 s to 2.5 s, corresponding to a 30% increase in average flow rate. The flow rate in the ingates can be estimated by:
$$ Q = \frac{V}{t} $$
where \(V\) is the total casting volume (approximately \(1.2 \times 10^{-3} \, \text{m}^3\)) and \(t\) is the pouring time. The increased flow velocity helped maintain sufficient metal temperature, especially in thin sections.
A critical improvement in the sand casting foundry design was the strategic placement of chills. Five steel chills were embedded in the sand mold at the locations of the rib roots and the bottom center of the cover. The chills locally increased the cooling rate, promoting directional solidification from the bottom and rib regions toward the risers. The thermal diffusivity of the chill material (steel) is an order of magnitude higher than that of the resin-bonded sand, effectively extracting heat from the hot spots. The solidified fraction evolution was monitored in AnyCasting. The solidification time was reduced by approximately 40% in the previously problematic areas. The risers were redesigned to have a greater volume and a neck diameter sufficiently large to remain liquid until the end of solidification. Two risers were placed on the circular top face, positioned near the thickest sections. The riser feeding distance can be evaluated using the modulus method:
$$ M = \frac{V}{A} $$
where \(V\) is the volume of the casting region and \(A\) is its cooling surface area. For the optimized sand casting foundry, the modulus of the riser necks was set 20% larger than the modulus of the region they feed.
| Region | Initial defect volume (mm³) | Optimized defect volume (mm³) |
|---|---|---|
| Rib roots (upper half) | 12.4 | 0.3 |
| Rib roots (lower half) | 8.7 | 0.1 |
| Bottom central area | 5.2 | 0.0 |
| Circular top (near riser) | 1.1 | 0.0 (riser itself contains all defects) |
| Total in casting body | 27.4 | 0.4 |
The optimized sand casting foundry simulation clearly demonstrated a near-complete elimination of internal defects. The flow pattern became laminar and stable, with the metal rising uniformly from the bottom. The gas tracking module indicated that the core gases were effectively expelled through the open top. The solidification front advanced from the chilled regions toward the risers, confirming directional solidification. The Niyama values in all previously defective zones increased to above 1.5, indicating a safe margin against shrinkage porosity. The risers contained all the residual porosity, which is easily removed during fettling. The temperature distribution during solidification followed a near-ideal gradient, which can be described by the following empirical relationship for directional solidification:
$$ \frac{\partial T}{\partial z} > \frac{\rho L}{k} \cdot \frac{\partial f_s}{\partial t} $$
where \(\rho\) is the density of the alloy, \(L\) is the latent heat, \(k\) is the thermal conductivity, and \(f_s\) is the solid fraction. The optimized sand casting foundry conditions satisfied this inequality throughout the casting body.
Furthermore, the filling simulation showed that the new gating system reduced the maximum flow velocity from 1.8 m/s to 0.9 m/s, minimizing turbulence and avoiding air entrapment. The use of a ceramic foam filter at the sprue base effectively trapped slag and oxides, ensuring clean metal delivery. The entire pouring time of 2.5 seconds still allowed the metal to reach the highest points without cold shuts. The thermal analysis revealed that the temperature of the last-filled regions (the risers) was still above 650°C, sufficient for proper feeding.
Following the successful numerical simulation, a production trial was conducted using the optimized sand casting foundry parameters. Twenty covers were cast in a batch using the same resin-bonded sand system. Visual inspection showed no surface defects such as cold shuts, misruns, or gas blowholes. The dimensional accuracy was verified using a coordinate measuring machine; all dimensions fell within the specified tolerances of ±0.5 mm. Leak testing was performed by pressurizing the internal cavity to 20 kPa and submerging the cover in water. No bubbles were observed for any of the twenty castings, confirming 100% leak-tightness. Metallographic examination of cross-sections taken from the rib roots and the bottom plate revealed a fully dense microstructure with fine equiaxed grains and no visible microporosity. The secondary dendrite arm spacing (SDAS) was measured to be approximately 30 μm, indicating a relatively rapid solidification rate, especially in the chilled areas. The mechanical properties were also evaluated: the ultimate tensile strength reached 280 MPa, and the elongation was 3.5%, meeting the requirements for the aerospace application.
| Quality metric | Initial production (typical) | Optimized production (measured) |
|---|---|---|
| Visual rejection rate (%) | 15 | 0 |
| Leak test failure rate (%) | 20 | 0 |
| Average porosity area fraction at ribs (%) | 1.2 | <0.1 |
| Dimensional conformity | 95% within tolerance | 100% within tolerance |
| SDAS at bottom (μm) | 45 | 28 |
The success of this study underscores the power of numerical simulation as an integral tool in sand casting foundry development. By systematically analyzing the flow and solidification behavior, we were able to identify the root causes of defects and propose cost-effective modifications without resorting to trial-and-error. The sand casting foundry process is inherently complex due to the interaction of mold filling, phase change, and heat transfer. However, with modern simulation software like AnyCasting, the engineer can visualize the process in detail and optimize parameters such as gating design, chill placement, riser geometry, and pouring conditions. The key lessons learned from this project include: (1) casting orientation should favor natural gas venting and directional solidification; (2) chills are highly effective in controlling hot spots in ribbed thin-wall castings; (3) reducing pouring time improves feeding and reduces cold shuts; (4) multiple risers are necessary for large flat surfaces. All these principles are universally applicable to any sand casting foundry operation dealing with aluminum alloys.
In conclusion, the numerical simulation and process optimization of the sand casting foundry for the aluminum alloy cover have led to a robust and reliable manufacturing process. The optimized design not only eliminated internal defects but also improved productivity and reduced cost by minimizing scrap. The methodology described here can be readily adapted to similar complex castings. Future work may focus on further refining the chill placement using topology optimization algorithms and incorporating real-time sensing during sand casting foundry production to dynamically adjust parameters. Nevertheless, the current results demonstrate that a careful combination of simulation, engineering judgment, and practical validation can achieve zero-defect casting in a sand casting foundry environment.
The author acknowledges the support of the university laboratory and the industrial partner for providing the casting facilities and materials. The simulation software AnyCasting was used under an academic license. The actual production trials were conducted in the partner’s sand casting foundry workshop. The data and analysis presented herein represent the culmination of a systematic study undertaken to advance the understanding of defect formation in sand casting foundry processes and to provide a practical solution for high-quality aluminum alloy component manufacturing.