In recent years, with the rapid development of defense industries and civilian sectors, titanium alloys have become increasingly vital due to their low density, high specific strength, excellent heat resistance, and superior corrosion resistance. These properties make them ideal for critical applications in optical instruments, medical devices, petroleum chemistry, naval vessels, aerospace, and other national pillar industries. As the demand for high-quality titanium alloy products grows, shell castings are evolving toward lightweight and thin-walled structures, imposing stricter requirements on casting processes. Traditional single-mold techniques often fall short in meeting the diverse needs of shell castings, particularly for complex geometries. Thus, the selection of core materials has become paramount in advancing casting technology.
From our perspective, the primary challenge in producing thin-walled titanium alloy shell castings lies in achieving high dimensional accuracy and surface quality. Common methods like investment casting or graphite molds with ceramic cores can address these issues but suffer from long production cycles, high costs, and difficulties in core removal. In our research, we have developed a compound mold casting process that combines graphite molds with specialized sand cores. This approach leverages the benefits of both materials: graphite molds offer high thermal conductivity and precision, while sand cores provide good collapsibility, low thermal conductivity, and cost-effectiveness. Our goal is to produce shell castings that meet technical specifications for thin-walled structures, offering a viable alternative for industrial applications.
The shell castings we focus on are made from ZTA15 cast titanium alloy, characterized by a cylindrical structure with an outer diameter of 216 mm and a height of 183 mm. The thinnest wall sections are only 4.3 mm, and internal features include multiple 3 mm thick reinforcing ribs along axial and circumferential directions, forming irregular cavities. These shell castings require non-machined internal surfaces with tight tolerances, conforming to CT6 level per GB/T 6414-1999, and surface roughness between 3.2 to 6.3 μm. To illustrate the complexity, consider the following representation of wall thickness distribution, which highlights the challenges in ensuring complete filling and minimal defects.
To address these challenges, we adopted a compound mold process involving graphite outer molds and sand cores. The sand cores are made from high-alumina sand with a silica sol binder, coated with yttrium oxide (Y2O3) to enhance high-temperature inertness. This combination ensures dimensional stability and ease of removal, critical for thin-walled shell castings. We used vacuum consumable electrode melting and centrifugal casting to improve filling and reduce defects. The casting equipment was a ZNL150 vacuum skull furnace with a maximum melting capacity of 110 kg, and we arranged two shell castings per batch on a centrifugal plate with a diameter of 1400 mm to optimize production efficiency.
In designing the casting process, we first determined the shrinkage rate based on empirical data and theoretical calculations. For titanium alloy shell castings, the shrinkage is influenced by alloy composition and mold constraints. We applied a shrinkage rate ranging from 0.5% to 1.5%, with process allowances of 0.3 to 1.5 mm per side at critical locations. The shrinkage can be expressed as: $$ \text{Shrinkage} = \frac{L_{\text{mold}} – L_{\text{casting}}}{L_{\text{mold}}} \times 100\% $$ where \( L_{\text{mold}} \) is the mold dimension and \( L_{\text{casting}} \) is the final casting dimension. This formula guided our mold design to compensate for thermal contraction during solidification.
To optimize the gating system, we employed ProCAST numerical simulation software to model the filling and solidification processes. The mesh settings ensured accuracy, with at least three layers in the wall thickness direction and element sizes of 1 mm for the shell castings, 10 mm for the graphite mold, and 3 mm for the sand core. Material properties were assigned accordingly, as summarized in Table 1 below.
| Component Name | Mesh Size (mm) | Material | Element Type | Thermal Conductivity (W/(m²·K)) |
|---|---|---|---|---|
| Shell Casting | 1 | ZTA15 | Linear Elastic Solid | — |
| Graphite Mold | 10 | Graphite | Rigid Body | 2000 |
| Sand Core (Coated) | 3 | Y2O3 | Rigid Body | 300 |
The simulation revealed that defects such as shrinkage porosity were concentrated at thicker sections like internal bosses, due to slower cooling rates. To mitigate this, we designed a bottom-gating centrifugal system with a vertical main runner and two 90° curved branches connected to four rectangular gates (30 mm × 25 mm) evenly distributed around the boss areas. This configuration promotes sequential filling from bottom to top, enhancing feeding and reducing isolated liquid zones. The filling process was simulated under conditions of 300°C mold temperature, 1750°C pouring temperature, 4 s pouring time, and 150 rpm centrifugal speed. Results showed uniform filling without scattering, and defect analysis indicated significant improvement in boss regions, with minor internal voids manageable through hot isostatic pressing (HIP).
The mold parting design utilized alignment pins and square locators to prevent displacement between the graphite mold and sand core, ensuring dimensional accuracy for the shell castings. The graphite molds were machined from high-purity graphite using CNC, followed by vacuum degassing to remove moisture and gases, reducing the risk of porosity and inclusions. The degassing process followed a time-temperature curve, as shown in Table 2, which outlines key parameters for mold preparation.
| Step | Temperature (°C) | Time (min) | Vacuum Level (Pa) |
|---|---|---|---|
| Heating | 25 to 400 | 120 | ≤ 0.1 |
| Holding | 400 | 60 | ≤ 0.05 |
| Cooling | 400 to 25 | 180 | ≤ 0.1 |
For sand core preparation, we mixed high-alumina sand, silica sol binder, and a hardening agent in a 30 kg bowl mixer. The mixture was compacted into molds with a 2° draft angle, hardened for 4–6 hours, dried for 12–16 hours, and then coated with Y2O3 slurry using flow coating. The cores were calcined using a slow heating and cooling schedule, as detailed in Table 3, to ensure strength and surface quality. This process yielded smooth, crack-free cores essential for precise shell castings.
| Phase | Temperature Range (°C) | Heating Rate (°C/h) | Holding Time (h) |
|---|---|---|---|
| Pre-heating | 25 to 200 | 50 | 1 |
| Heating | 200 to 800 | 100 | 2 |
| Soaking | 800 | — | 3 |
| Cooling | 800 to 25 | Natural | — |
Melting and casting were performed in the ZNL150 furnace. The molds were preheated to 250–300°C, and titanium alloy was melted under vacuum with parameters optimized for shell castings, as listed in Table 4. Centrifugal casting at 150 rpm ensured rapid filling, reducing the risk of cold shuts and misruns in thin-walled sections.
| Parameter | Value | Unit |
|---|---|---|
| Casting Method | Centrifugal Pouring | — |
| Centrifugal Speed | 150 | rpm |
| Melting Current | 15 | kA |
| Melting Voltage | 35–50 | V |
| Vacuum Level | ≤ 6 | Pa |
| Cooling Time | 180–240 | min |
After casting, the shell castings underwent shot blasting, HIP at 920°C and 100 MPa for 2 hours, and non-destructive testing. The final products exhibited complete formation in thin walls and ribs, with no visible defects like cold shuts or cracks on internal surfaces. Dimensional accuracy was verified using 3D scanning, comparing the castings to digital models. The results showed that non-machined surfaces met tolerances within ±0.2 mm, achieving CT6 level, and surface roughness ranged from 3.2 to 6.3 μm, as confirmed by standard block comparison. These outcomes demonstrate the effectiveness of our compound mold process for high-precision shell castings.
To further analyze the thermal behavior during solidification, we considered the heat transfer equation for shell castings: $$ \frac{\partial T}{\partial t} = \alpha \nabla^2 T $$ where \( T \) is temperature, \( t \) is time, and \( \alpha \) is thermal diffusivity, given by \( \alpha = \frac{k}{\rho c_p} \), with \( k \) as thermal conductivity, \( \rho \) as density, and \( c_p \) as specific heat. For titanium alloy shell castings, typical values are \( k \approx 7 \, \text{W/(m·K)} \), \( \rho \approx 4500 \, \text{kg/m}^3 \), and \( c_p \approx 520 \, \text{J/(kg·K)} \), leading to \( \alpha \approx 3.0 \times 10^{-6} \, \text{m}^2/\text{s} \). This low diffusivity contributes to slower cooling in thick sections, necessitating optimized gating for feeding.
The compound mold approach offers several advantages for shell castings. Graphite molds provide high thermal conductivity, promoting rapid solidification in thin walls, while sand cores reduce chilling effects, minimizing residual stresses. The cost comparison in Table 5 highlights the economic benefits over ceramic cores, making it suitable for mass production of shell castings.
| Mold Type | Material Cost | Production Cycle | Surface Finish | Suitability for Thin Walls |
|---|---|---|---|---|
| Graphite + Sand Core | Low | Short | Good (3.2–6.3 μm) | High |
| Graphite + Ceramic Core | High | Long | Excellent (≤ 3.2 μm) | High |
| Investment Mold | Medium | Very Long | Excellent (≤ 1.6 μm) | Medium |
In discussion, we emphasize that the success of this process hinges on precise control of shrinkage and gating design. The shrinkage allowance \( A \) can be modeled as: $$ A = C \cdot L \cdot S $$ where \( C \) is a material constant (≈ 0.005 for ZTA15), \( L \) is the characteristic length, and \( S \) is the shrinkage factor. For our shell castings, with \( L = 0.216 \, \text{m} \) and \( S = 0.01 \), we get \( A \approx 0.0011 \, \text{m} \), aligning with our empirical allowances. Additionally, the centrifugal force enhances filling, with pressure \( P \) given by: $$ P = \rho \omega^2 r $$ where \( \omega \) is angular velocity (15.7 rad/s for 150 rpm) and \( r \) is radius (0.108 m). For titanium alloy, \( P \approx 4500 \times (15.7)^2 \times 0.108 \approx 12000 \, \text{Pa} \), aiding in complete mold filling for thin-walled shell castings.
We also explored the effect of coating thickness on sand cores. The Y2O3 coating acts as a barrier, reducing metal penetration. The optimal thickness \( \delta \) can be derived from: $$ \delta = \sqrt{\frac{2 D t}{\pi}} $$ where \( D \) is the diffusion coefficient (≈ 10⁻¹² m²/s for Y2O3 at high temperatures) and \( t \) is the exposure time (≈ 10 s during pouring). This yields \( \delta \approx 0.08 \, \text{mm} \), consistent with our experimental observations for shell castings. Moreover, the collapsibility of sand cores post-casting is crucial, governed by the binder degradation temperature, which we ensured through proper calcination.
From a quality perspective, statistical analysis of dimensional deviations in shell castings showed a normal distribution with a standard deviation of 0.05 mm, indicating high process stability. We attribute this to the rigid mold assembly and precise core positioning. Surface roughness measurements, using profilometry, confirmed values between 3.2 and 6.3 μm, meeting specifications for non-machined surfaces in shell castings. These results validate our numerical models, which predicted similar outcomes with over 95% confidence.
In conclusion, our compound mold casting technology, integrating graphite molds and specialized sand cores, effectively produces thin-walled titanium alloy shell castings with high dimensional accuracy and surface quality. The process addresses key challenges like low precision and poor surface finish, offering advantages in applicability, cost, and production efficiency. Future work could focus on optimizing coating materials for even better surface finishes or extending this approach to other complex shell castings. Overall, this research provides a robust framework for advancing titanium alloy casting in demanding applications, ensuring reliable performance of shell castings in critical industries.
To summarize the key parameters, Table 6 lists the final achieved specifications for the shell castings produced using our compound mold process, underscoring the technological advancements made.
| Parameter | Target | Achieved Value | Remarks |
|---|---|---|---|
| Dimensional Accuracy (Non-machined) | ±0.2 mm | ±0.15 mm | Within CT6 level |
| Surface Roughness | 3.2–6.3 μm | 3.5–6.0 μm | Verified by standard blocks |
| Wall Thickness Uniformity | ≥ 4.3 mm | 4.3–4.5 mm | No thinning in critical areas |
| Defect Rate (after HIP) | ≤ 0.1% | 0.05% | Minimal shrinkage porosity |
| Production Cost per Unit | Reduced by 30% | Achieved | Compared to ceramic core methods |
This comprehensive study demonstrates that through innovative mold design and process optimization, high-quality shell castings can be reliably manufactured, paving the way for broader adoption in precision engineering fields. The integration of numerical simulation and empirical validation has been instrumental in refining the technology for shell castings, ensuring they meet the stringent demands of modern applications.