In the field of precision casting, the production of thin-walled shell castings presents significant challenges due to their complex geometries, stringent dimensional accuracy requirements, and the need for high mechanical properties. Shell castings are widely used in protective components for vehicles and aerospace applications, where they must withstand impact loads while maintaining structural integrity. Traditional casting methods often fall short in meeting these demands, leading to defects such as shrinkage porosity and inadequate filling. Therefore, gypsum mold investment casting has emerged as a preferred technique for manufacturing thin-walled shell castings, offering advantages like excellent surface finish, dimensional stability, and the ability to produce intricate details. In this study, we focus on the process design and optimization for a thin-walled shell casting using gypsum mold investment casting, supported by numerical simulation to predict and mitigate defects.
The shell casting under investigation features a curved, dome-like structure with an internal concave surface that adds to its complexity. The shell castings have a minimum radius of 239 mm, a maximum radius of 259.5 mm, and an average wall thickness of 4 mm, classifying them as thin-walled components. Such geometry necessitates a careful approach to gating and feeding system design to ensure proper metal flow and solidification. Initially, we designed a bottom-gating system to promote平稳的充型 and reduce turbulence, which is critical for minimizing gas entrapment and oxide inclusions. The gating system was positioned on one side of the shell castings, with multiple ingates distributed to facilitate uniform filling. The dimensions of the gating elements were as follows: a tapered sprue (d1=18 mm, d2=16 mm, height=250 mm), a trapezoidal runner (a=40 mm, b=35 mm, height=33 mm, length=45 mm), cylindrical ingates (diameter=9.6 mm, length=4 mm), and a conical pouring cup (d1=18 mm, d2=36 mm, height=18 mm). This initial design aimed to provide adequate feeding but required validation through simulation.
To analyze the casting process, we employed numerical simulation using ViewCast software, which allows for detailed modeling of fluid flow, heat transfer, and solidification phenomena. The 3D model of the shell castings, including the gating system, was converted to an STL file and imported into the software. The mesh was refined to approximately 2 million elements to ensure accuracy. The material for the shell castings was ZL101A aluminum alloy, with a pouring temperature of 705°C and an initial mold temperature of 220°C. The thermophysical properties of both the casting material and the gypsum mold are critical for simulation accuracy; these parameters are summarized in Table 1. The thermal conductivity (λ) and specific heat capacity (c) vary with temperature, influencing the cooling rates and defect formation in shell castings.
| Material | Temperature (°C) | Thermal Conductivity, λ (W·m⁻¹·K⁻¹) | Specific Heat Capacity, c (J·kg⁻¹·K⁻¹) |
|---|---|---|---|
| ZL101A (Shell Castings) | 100 | 154.9 | 879 |
| 200 | 163.3 | 921 | |
| 300 | 167.5 | 1005 | |
| 400 | 167.5 | 1100 | |
| Gypsum Mold | 100 | 0.72 | 1100 |
| 200 | 0.60 | 1000 | |
| 300 | 0.50 | 900 | |
| 400 | 0.50 | 1000 |
The simulation of the initial design revealed insights into the filling and solidification behavior of the shell castings. During filling, the metal liquid entered the mold cavity through the ingates at t=1.20 s, with complete filling achieved by t=2.55 s. The process was平稳, without significant turbulence, as shown in the velocity contours. However, the solidification simulation indicated potential issues: thin-walled sections solidified first, followed by thicker regions, leading to inadequate feeding in the upper right corner of the shell castings where mass accumulates. This resulted in predicted shrinkage porosity and cavities, as visualized in the defect prediction plots. The solidification time for various regions can be approximated using Chvorinov’s rule, expressed as:
$$ t_s = C \left( \frac{V}{A} \right)^n $$
where \( t_s \) is the solidification time, \( V \) is the volume, \( A \) is the surface area, \( C \) is a constant dependent on mold material and casting conditions, and \( n \) is an exponent typically around 2. For shell castings, the modulus \( \frac{V}{A} \) is higher in thick sections, leading to longer solidification times and increased risk of shrinkage if not properly fed. In our initial design, the gating system’s location limited its effectiveness in feeding these isolated hot spots.
To address these defects, we optimized the gating system for the shell castings. The key changes included relocating the pouring position from the side to the center of the shell castings and maintaining a multi-point ingate distribution to improve thermal balance. This redesign aimed to create a more directional solidification pattern, with the gating system acting as a thermal reservoir to feed the thicker sections. The optimized gating system dimensions were similar, but the central placement altered the fluid dynamics and heat transfer. We then simulated this optimized scheme to evaluate its performance. The filling process showed improved uniformity, with complete filling at t=2.6 s, and the solidification sequence indicated that the thick sections now solidified concurrently with thinner areas, reducing the isolation of hot spots. The defect prediction for the optimized shell castings showed a significant reduction in shrinkage porosity, with defects primarily confined to the gating system itself, which is acceptable as it will be removed during post-processing.
The effectiveness of the optimization can be further analyzed through mathematical modeling of heat transfer. The temperature distribution in the shell castings during solidification can be described by the heat conduction equation:
$$ \frac{\partial T}{\partial t} = \alpha \nabla^2 T $$
where \( T \) is temperature, \( t \) is time, and \( \alpha \) is the thermal diffusivity, given by \( \alpha = \frac{\lambda}{\rho c} \), with \( \rho \) being density. For shell castings, the low thermal conductivity of the gypsum mold (as seen in Table 1) results in slower heat extraction, which must be managed through gating design. By centralizing the gating, we enhanced the thermal gradient, promoting directional solidification toward the feeder. Additionally, the feeding efficiency can be quantified using the feeding distance formula for thin-walled castings:
$$ L_f = k \sqrt{t_s} $$
where \( L_f \) is the feeding distance, \( k \) is a material constant, and \( t_s \) is the local solidification time. In the optimized design, the reduced \( t_s \) in thick sections due to better thermal connectivity increased \( L_f \), thereby minimizing shrinkage in shell castings.
Following the simulation validation, we proceeded to practical production of the shell castings using the optimized process. The gypsum slurry preparation is crucial for mold quality; the composition used in our production is detailed in Table 2. This配方 ensures adequate strength, permeability, and collapsibility for the shell castings. The slurry was poured around wax patterns, dried, and dewaxed before pouring the ZL101A alloy at 705°C. The resulting shell castings were inspected visually and via X-ray radiography, confirming the absence of internal defects such as shrinkage pores or cracks. The surface quality met the requirements of GB/T 9438-2013 Class II castings, demonstrating the success of the optimization.
| Component | Particle Size (mm) | Weight Percentage (wB%) |
|---|---|---|
| Gypsum | – | 28-32 |
| Quartz Powder | 0.075-0.053 | 9.0-11 |
| Quartz Sand | 0.053-≤0.053 | 5.0-8.0 |
| Bauxite | 0.43-0.20 | 31-35 |
| Bauxite Sand | <0.053 | 11-16 |
| Coal Gangue | 0.43-0.20 | 4.0-6.0 |
| Diatomite | 0.21-0.11 | 2.0-4.0 |
| Water | – | 28-32 |
To evaluate the mechanical performance of the produced shell castings, we subjected them to T6 heat treatment (solution treatment and aging) to enhance strength and hardness. Multiple specimens were extracted from the shell castings and tested for tensile strength, elongation, and hardness. The results, presented in Table 3, show that all metrics exceed the required thresholds (tensile strength ≥275 MPa, elongation ≥2%, hardness ≥80 HBW), validating the suitability of the optimized process for high-quality shell castings. The consistency in properties across specimens underscores the reliability of the gypsum mold investment casting method for thin-walled applications.
| Specimen | Tensile Strength (MPa) | Elongation (%) | Hardness (HBW) |
|---|---|---|---|
| 1 | 326 | 5.5 | 99.5 |
| 2 | 324 | 4.0 | 89.2 |
| 3 | 305 | 4.5 | 104.0 |
The success of this optimization hinges on a deep understanding of the solidification dynamics in shell castings. We can model the shrinkage defect formation using the criterion function for porosity prediction, often based on the Niyama criterion:
$$ G / \sqrt{\dot{T}} \leq C_{ny} $$
where \( G \) is the temperature gradient, \( \dot{T} \) is the cooling rate, and \( C_{ny} \) is a threshold constant. In regions where this criterion is violated, such as the thick sections in the initial design, shrinkage defects are likely. By improving \( G \) through central gating, we reduced the risk for the shell castings. Furthermore, the fluid flow during filling can be analyzed using the Navier-Stokes equations:
$$ \rho \left( \frac{\partial \mathbf{v}}{\partial t} + \mathbf{v} \cdot \nabla \mathbf{v} \right) = -\nabla p + \mu \nabla^2 \mathbf{v} + \rho \mathbf{g} $$
where \( \mathbf{v} \) is velocity, \( p \) is pressure, \( \mu \) is dynamic viscosity, and \( \mathbf{g} \) is gravity. For shell castings, the bottom-gating design minimizes velocity fluctuations, reducing oxide formation. The optimization further stabilized flow by centralizing the inlet, as evidenced in the simulation results.
In discussion, the role of numerical simulation in process design for shell castings cannot be overstated. It allows for virtual testing of multiple scenarios without costly trial runs. Our use of ViewCast enabled precise identification of defect-prone zones, guiding targeted optimizations. Additionally, the gypsum mold’s properties, such as low thermal conductivity, contribute to slower cooling, which can be beneficial for feeding but requires careful control to avoid defects. The multi-point ingate distribution in shell castings ensures even metal distribution, reducing thermal gradients and stress concentrations. This is particularly important for thin-walled shell castings, where rapid solidification can lead to incomplete filling or cold shuts.
From a broader perspective, the methodology applied here—combining simulation-driven design with practical validation—offers a robust framework for manufacturing complex shell castings. Future work could explore advanced materials for gypsum molds to enhance thermal properties or integrate machine learning for automated optimization of gating designs. The continuous improvement in simulation accuracy will further reduce defects in shell castings, expanding their applications in demanding industries.
In conclusion, through a systematic approach involving initial process design, numerical simulation, and optimization, we successfully produced high-quality thin-walled shell castings using gypsum mold investment casting. The relocation of the pouring position to the center and the use of multi-point ingates significantly reduced shrinkage defects, as confirmed by both simulation and practical production. The mechanical properties of the shell castings met all requirements, demonstrating the effectiveness of the optimized process. This study underscores the value of simulation tools in casting process development and provides a reliable method for manufacturing precision shell castings with complex geometries. The insights gained can be applied to other thin-walled casting projects, enhancing overall efficiency and quality in the foundry industry.