As the lightest metallic structural material, magnesium alloys offer excellent specific strength and stiffness, good castability and machinability, and ease of recycling, making them ideal for lightweight applications in aerospace and other fields. With abundant magnesium resources, China accounts for over 50% of global primary magnesium production, supported by a vast research community and deep-processing industrial system. In recent years, driven by global emphasis on energy conservation and emission reduction, along with the urgent need for further weight reduction in new energy vehicles, magnesium alloys are rapidly expanding their lightweight application scenarios. Among these, Mg-Al series magnesium alloys, particularly ZM5 (ZMgAl8Zn), are widely used due to their balanced mechanical properties, absence of rare earth or zirconium elements, and cost-effectiveness. ZM5 alloys exhibit good atmospheric corrosion resistance and are commonly employed in aircraft engine components, pods, and other aerospace structural castings. However, during sand casting of large thin-wall ZM5 casting parts, the developed dendritic solidification morphology often leads to shrinkage porosity or even shrinkage cavities on the surface, severely reducing yield and compromising safety.
In this study, we address frequent shrinkage defects at the rounded corners of an ultra-large magnesium alloy aircraft pod thin-wall casting part. We first analyze the defect causes from perspectives of casting structure, alloy material, and casting process. A solution is proposed to locally add and optimize ingates in the existing pouring system, enhancing riser feeding capability for defect-prone areas while inhibiting oxide inclusion defects. Using finite element simulation (ProCAST), we conduct casting process analysis for the optimized design, including flow and temperature fields, solid fraction, and temperature contour slices, confirming feasibility. Finally, casting trials validate the approach. Results demonstrate that the optimized process effectively eliminates shrinkage defects at rounded corners of ultra-large magnesium alloy thin-wall casting parts. This research provides valuable insights for casting process design of similar ultra-large magnesium alloy thin-wall casting parts.
Casting simulation technology (CAE) enables simulation of mold filling and solidification, predicting defects like shrinkage porosity, inclusions, and cracks, thereby reducing trial iterations, shortening development cycles, and lowering costs. Software such as ProCAST, Magma, AnyCasting, and Flow3D are widely used. In China, institutions like Tsinghua University and Huazhong University of Science and Technology have made significant contributions. This paper focuses on solving local shrinkage defects in an ultra-large ZM5 magnesium alloy aircraft pod thin-wall casting part through defect cause analysis, gating system optimization, ProCAST simulation, and experimental verification.
The casting part under study is a rear pod lower segment, a critical structure for an aircraft refueling pod. Its surface area is 39179.67 cm², nearly five times the minimum surface area (8000 cm²) for ultra-large castings per specification HB7780, classifying it as ultra-large. This sand-cast part is a rectangular box-like structure with an open bottom, prone to casting stress and deformation, hence process strengthening ribs are added. As an ultra-large thin-wall casting part, with wall thicknesses as low as 7 mm on top slopes and side planes, prolonged filling and rapid cooling during casting easily induce shrinkage porosity and cavities.
The material is ZM5 magnesium alloy, with chemical composition as shown in Table 1. Aluminum improves room-temperature strength, hardness, heat resistance, and corrosion resistance; zinc mitigates hot shortness and enhances corrosion resistance; manganese increases impact resistance, fatigue life, and improves heat treatment and corrosion properties.
| Element | Al | Zn | Mn | Mg |
|---|---|---|---|---|
| Content | 7.5–9.0 | 0.2–0.8 | 0.15–0.50 | Balance |
Based on the Mg-Al phase diagram, ZM5 alloy solidification involves liquid (L), α-Mg, and β-Mg17Al12 phases, with minor Al11Mn4 and Al8Mn5 phases. At 437°C, a eutectic reaction L → α + β occurs. From the liquidus temperature (~620°C), solidification proceeds through (L + α) two-phase region, α-Mg solid solution single-phase region, and (α + β) two-phase region. For thin-wall casting parts, non-equilibrium rapid solidification leads to dendritic growth of primary α-Mg in the (L + α) region, followed by eutectic reaction in Al-rich interdendritic liquid forming (α + β) divorced eutectic. The dendritic network at hot spots like rounded corners hinders melt flow and feeding, and high solidification shrinkage of ZM5 alloy promotes shrinkage defects. Pouring temperature is typically 90–150°C above solidus; considering high thermal conductivity of ZM5 and resin sand, an initial pouring temperature of 745°C is selected.
The shrinkage defects occur at rounded corners where the top plane intersects vertically with side planes (radius R=30 mm). Conventional solutions like increasing riser cross-section or using insulating risers were ineffective. Cause analysis includes:
- Casting Structure: The small radius at intersections acts as a heat accumulation zone, prone to shrinkage due to poor heat dissipation.
- Material Properties: ZM5 alloy’s dendritic solidification characteristic promotes shrinkage porosity.
- Casting Process: The original two-layer horizontal gating system with bottom filling (Figure 5a in reference) avoids direct flow but results in long flow paths (2400–3004 mm), causing rapid cooling and inadequate feeding at rounded corners.
To address this, we optimize the gating system by adding ingates on the two existing horizontal runners (Figure 5b in reference), raising upper melt temperature, enhancing vertical temperature gradient for directional solidification, and improving riser feeding. To avoid mold flash and inclusion defects due to limited mold space (20 mm sand thickness), we innovatively use upright horn-shaped ingates instead of conventional flat ones, with the horn tip extending into the riser above the rounded corner. This directs melt into the riser upper part, increasing riser temperature and enhancing feeding for the rounded corner area.
For simulation verification, we model the optimized casting part, gating system, and chills using CATIA. Finite element meshing is performed with minimum size of 3 mm in thin-wall regions, ensuring at least 2 mesh layers per section, totaling 104,352 elements. A virtual mold box is applied to reduce mesh count. Simulation parameters are listed in Table 2, and sand mold thermophysical properties in Table 3.
| Parameter | Value |
|---|---|
| Material (Casting) | ZM5 |
| Material (Mold) | Resin Sand |
| Material (Chill) | Cast Iron |
| Pouring Temperature | 745°C |
| Filling Velocity | 1 m/s |
| Gravity (Direction, Magnitude) | -Y, 9.8 m/s² |
| Heat Transfer Coefficient (Casting-Mold) | 300 W/(m²·°C) |
| Heat Transfer Coefficient (Casting-Chill) | 800 W/(m²·°C) |
| Heat Transfer Coefficient (Mold-Chill) | 300 W/(m²·°C) |
| Initial Temperature (Melt) | 745°C |
| Initial Temperature (Mold) | 25°C |
| Initial Temperature (Chill) | 25°C |
| Temperature (°C) | Thermal Conductivity (W/(m·°C)) | Specific Heat Capacity (J/(kg·°C)) |
|---|---|---|
| 25 | 0.73 | 680 |
| 200 | 0.64 | 905 |
| 400 | 0.59 | 1020 |
| 600 | 0.59 | 1098 |
| 800 | 0.64 | 1150 |
Coupled flow and temperature field simulations show smooth filling. At 17 s, all horizontal runners are filled, and melt enters through bottom ingates at ends. At 30 s, over 85% of the casting part is filled, with melt converging between large and small windows, potentially causing oxide inclusions. At 34 s, casting filling completes, and riser filling finishes at 59 s. Solidification analysis indicates that at 99 s, the large plane begins solidifying around 596.5°C. At 309 s, the casting part largely solidifies (~438°C), while the riser starts solidifying (~596°C). At 1399 s, the casting part is almost fully solidified, and by 3229 s, complete solidification occurs. The optimized system achieves directional solidification from bottom to top, with the casting part solidifying before the riser.
Solid fraction analysis reveals no isolated liquid regions during solidification. Temperature contour slices at defect-prone rounded corners show riser temperature significantly higher than the corner temperature, confirming improved feeding. Shrinkage porosity prediction with a critical volume fraction of 1% indicates no shrinkage defects, validating the optimization.
The thermal behavior during solidification can be described using the heat conduction 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 $Q$ is heat source term. For casting parts, $Q$ may include latent heat release during phase change, which for ZM5 alloy can be modeled as:
$$ Q = L \frac{\partial f_s}{\partial t} $$
with $L$ being latent heat and $f_s$ solid fraction. The solid fraction evolution can be approximated using Scheil’s equation for non-equilibrium solidification:
$$ f_s = 1 – \left( \frac{T_m – T}{T_m – T_l} \right)^{\frac{1}{1-k}} $$
where $T_m$ is melting point, $T_l$ is liquidus temperature, and $k$ is partition coefficient.
To further quantify feeding efficiency, we consider the pressure drop in the gating system using Darcy’s law for fluid flow in porous media (dendritic mesh):
$$ \mathbf{u} = -\frac{K}{\mu} \nabla P $$
where $\mathbf{u}$ is velocity, $K$ is permeability, $\mu$ is viscosity, and $P$ is pressure. Permeability decreases as solid fraction increases, hindering feeding. The optimized ingates enhance pressure gradient towards the riser, improving feeding.
Casting trials with the optimized process produce five casting parts, all free of shrinkage defects at rounded corners, increasing yield from 39% to 80%. This demonstrates the effectiveness of the solution for ultra-large magnesium alloy thin-wall casting parts.
In summary, we analyze shrinkage defects in an ultra-large aircraft pod magnesium alloy casting part from structural, material, and process perspectives. The original bottom-filling gating system lacks adequate temperature gradient for directional solidification. By adding ingates on horizontal runners and using upright horn-shaped ingates, we enhance riser feeding capability. ProCAST simulations confirm improved temperature gradients and elimination of shrinkage. Experimental validation shows significant yield improvement. This work offers practical guidance for designing and optimizing casting processes for similar ultra-large thin-wall casting parts.
For future work, other factors affecting casting quality could be explored, such as mold material properties, cooling rate control, and alloy modification. Additionally, advanced simulation techniques incorporating microstructure prediction could further optimize process parameters. The methodology here can be extended to other magnesium alloy systems or larger casting parts in aerospace and automotive industries.
To generalize the optimization approach, we propose a formula for evaluating feeding efficiency in such casting parts:
$$ \eta_f = \frac{\int_{t_0}^{t_f} (T_r – T_c) \, dt}{\Delta T_{ref} \cdot t_f} $$
where $\eta_f$ is feeding efficiency index, $T_r$ is riser temperature, $T_c$ is critical region temperature (e.g., rounded corner), $\Delta T_{ref}$ is reference temperature difference, and $t_f$ is solidification time. Higher $\eta_f$ indicates better feeding. For the optimized design, $\eta_f$ increases significantly.
In conclusion, the integration of defect analysis, gating system innovation, and casting simulation effectively solves shrinkage issues in ultra-large magnesium alloy thin-wall casting parts, highlighting the importance of CAE in modern foundry practice.