In this work, we investigated the manufacturing of a typical thin-walled aluminum alloy casting, a ventilation port component, by combining high precision investment casting with counter-gravity casting technology. The component is made of E357.0 alloy, with a uniform wall thickness of 2 mm and overall dimensions of 240 mm × 140 mm × 100 mm. The entire casting is left unmachined, and the metallurgical quality must meet the Grade C requirements of ASTM E155. Based on our initial analysis, conventional gravity casting is insufficient to produce a fully shaped casting due to the extreme thinness of the walls. Therefore, we adopted the counter-gravity casting method, specifically the vacuum-assisted filling and pressurized solidification technique, integrated with high precision investment casting to achieve the required quality.
The principle of vacuum-assisted filling and pressurized solidification is based on a two-chamber system. The mold is placed in the upper chamber and sealed, while the lower chamber is initially open to the atmosphere. The upper chamber is evacuated to a predetermined vacuum level, causing the molten aluminum to rise smoothly into the mold cavity. After filling, the lower chamber is sealed and pressurized, applying a controlled pressure on the melt surface to ensure effective feeding and solidification under pressure. This method significantly reduces the gas back-pressure in the mold cavity during filling, improving the molten metal’s ability to flow into thin sections and minimizing gas entrapment. It is particularly well-suited for producing thin-walled castings where traditional gravity pouring often results in incomplete filling or oxide inclusions.
Process Design
We designed the gating system for the ventilation port casting using a fused silica-ethyl silicate composite shell mold. The shell was preheated to 350 °C before pouring. The key parameters influencing the casting quality were identified as the molten metal filling velocity and the pouring temperature. The filling velocity was controlled by setting the vacuum level and the evacuation time in the automated control system. The relationship between vacuum pressure and liquid metal rise can be approximated by the hydrostatic equation:
$$
\Delta P = \rho g h
$$
where $\Delta P$ is the pressure difference (in Pa), $\rho$ is the density of the molten aluminum alloy (approximately 2400 kg/m³ for E357.0 at pouring temperature), $g$ is the gravitational acceleration (9.81 m/s²), and $h$ is the height of the metal column (in meters). For our casting, the total height of the cavity is about 100 mm, so the required pressure difference to overcome gravity is about 2354 Pa. However, additional pressure is needed to overcome frictional losses and surface tension, especially in thin sections.
We conducted a series of experiments with six combinations of filling velocity and pouring temperature, as shown in Table 1. The filling velocity was determined by dividing the height of the mold cavity by the measured evacuation time. For example, at a vacuum level of 50 kPa, the theoretical rise height per kPa is approximately 40 mm, so the real-time velocity could be estimated.
| Experiment No. | Filling Velocity (mm/s) | Pouring Temperature (°C) | Vacuum Level (kPa) |
|---|---|---|---|
| 1 | 63 | 725 | 30 |
| 2 | 63 | 735 | 30 |
| 3 | 77 | 725 | 36 |
| 4 | 77 | 735 | 36 |
| 5 | 100 | 725 | 47 |
| 6 | 100 | 735 | 47 |
The relationship between vacuum pressure and theoretical filling velocity can be expressed using Bernoulli’s equation for inviscid flow, neglecting friction:
$$
v = \sqrt{\frac{2\Delta P}{\rho}}
$$
For $\Delta P = 47$ kPa, $\rho = 2400$ kg/m³, the theoretical maximum velocity would be about 6.3 m/s, but due to flow resistance in the thin channels, the actual measured velocities are much lower (on the order of 0.1 m/s). This simplification, however, helps to understand the influence of vacuum on filling capacity. The Reynolds number during filling can be estimated as:
$$
Re = \frac{\rho v D_h}{\mu}
$$
where $D_h$ is the hydraulic diameter of the thin section (2 mm) and $\mu$ is the dynamic viscosity of molten aluminum (about 1.2 mPa·s at 725 °C). For a velocity of 0.1 m/s, $Re \approx 400$, indicating laminar flow. This is beneficial for avoiding turbulence and gas entrapment in high precision investment casting.
Experimental Results
The experimental outcomes for each parameter set are summarized in Table 2. We evaluated the casting completeness and internal soundness using X-ray radiography and fluorescent penetrant inspection.
| Experiment No. | Filling Velocity (mm/s) | Pouring Temperature (°C) | Result |
|---|---|---|---|
| 1 | 63 | 725 | Incomplete filling (short run) |
| 2 | 63 | 735 | Incomplete filling (short run) |
| 3 | 77 | 725 | Incomplete filling |
| 4 | 77 | 735 | Complete filling, but porosity at hot spots |
| 5 | 100 | 725 | Complete filling, defect-free (satisfied specification) |
| 6 | 100 | 735 | Complete filling, but porosity at hot spots |
From the results, it is evident that a filling velocity of 63 mm/s was insufficient to achieve complete filling regardless of temperature. Increasing the velocity to 77 mm/s improved filling at 735 °C but still resulted in porosity in the thicker sections (corners). The best outcome was obtained with Experiment No. 5: a filling velocity of 100 mm/s and a pouring temperature of 725 °C. This combination produced a fully formed casting with no internal defects, meeting all acceptance criteria. In contrast, Experiment No. 6 (100 mm/s, 735 °C) exhibited porosity at the hot spots, indicating that excessively high pouring temperature promotes shrinkage porosity in the final solidification regions.
Analysis and Discussion
The ventilation port casting has a uniform wall thickness of 2 mm, with the only significant hot spots located at the fillets and corners where the geometry changes direction. Solidification modeling reveals that these regions have a higher local modulus, leading to slower cooling. The critical condition for porosity formation can be related to the feeding distance and the local thermal gradient. The Niyama criterion often used for porosity prediction is:
$$
N_y = \frac{G}{\sqrt{R}}
$$
where $G$ is the temperature gradient (K/mm) and $R$ is the cooling rate (K/s). When $N_y$ falls below a threshold value (typically around 1 K$^{0.5}$·mm$^{-1.5}$ for aluminum alloys), shrinkage porosity is likely. In our process, increasing the pouring temperature from 725 °C to 735 °C reduces the thermal gradient because the entire mold is heated more uniformly, causing $N_y$ to decrease at the hot spots, thus promoting porosity. On the other hand, increasing the filling velocity from 77 mm/s to 100 mm/s enhances the mold filling completeness by providing a higher pressure head and reducing the time available for premature solidification. However, too high a filling velocity can also cause turbulent flow, but in the laminar regime (Re ~ 400), the flow remains stable.
The role of vacuum pressure in high precision investment casting is critical. The pressure differential $\Delta P$ drives the melt upward. The required $\Delta P$ to fill a thin channel of length $L$ and thickness $t$ can be estimated by the Hagen-Poiseuille equation for laminar flow in a narrow slit:
$$
\Delta P = \frac{12 \mu L v}{t^2}
$$
For $L = 100$ mm, $t = 2$ mm, $v = 100$ mm/s, $\mu = 1.2 \times 10^{-3}$ Pa·s, we obtain $\Delta P \approx 360$ Pa. This is much smaller than the hydrostatic pressure (~2350 Pa) and the applied vacuum (47 kPa), indicating that the flow resistance in the thin section is negligible compared to the gravitational head. Therefore, the main limitation for filling is not the viscous loss but rather the surface tension and the ability to maintain a stable liquid front. The Bond number, comparing gravity to surface tension, is defined as:
$$
Bo = \frac{\rho g t^2}{\sigma}
$$
where $\sigma$ is the surface tension of molten aluminum (about 0.86 N/m). For $t = 2$ mm, $Bo \approx 0.11$, indicating that surface tension dominates; thus, careful control of filling velocity is needed to avoid capillary blockage. A too-slow filling velocity (e.g., 63 mm/s) allows the melt to freeze prematurely at the leading edge due to heat loss, while a sufficiently high velocity (100 mm/s) ensures that the melt front advances before solidification.
To further analyze the optimal condition, we performed a thermal simulation (assuming a mold preheat of 350 °C and a casting temperature of 725 °C). The dimensionless Fourier number for heat conduction during filling is:
$$
Fo = \frac{\alpha t_f}{d^2}
$$
where $\alpha$ is the thermal diffusivity of the shell mold (approximately $5 \times 10^{-7}$ m²/s), $t_f$ is the filling time (about 1 second for a height of 100 mm at 100 mm/s), and $d$ is the shell thickness (average 6 mm). $Fo \approx 0.014$, which is very small, indicating that heat loss during filling is minimal. Therefore, mold preheating is effective in maintaining the fluidity.
Table 3 summarizes the calculated parameters for the key experiments, demonstrating why Experiment No. 5 succeeded.
| Parameter | Experiment No. 1 | Experiment No. 4 | Experiment No. 5 | Experiment No. 6 |
|---|---|---|---|---|
| Filling velocity (mm/s) | 63 | 77 | 100 | 100 |
| Pouring temperature (°C) | 725 | 735 | 725 | 735 |
| Filling time (s) | 1.59 | 1.30 | 1.00 | 1.00 |
| Calculated $\Delta P$ flow resistance (Pa) | 227 | 277 | 360 | 360 |
| Estimated solidification time at hot spot (s) | 12.5 | 13.2 | 12.5 | 13.8 |
| Niyama criterion at hot spot (K$^{0.5}$·mm$^{-1.5}$) | 1.2 | 0.9 | 1.3 | 0.8 |
| Result | Incomplete filling | Porosity | Defect-free | Porosity |
The Niyama criterion values corroborate the experimental results: values below 1.0 indicate porosity, while values above 1.2 suggest sound castings. Experiment No. 5 achieved a criterion of 1.3, well above the threshold, whereas Experiment No. 6 fell to 0.8 due to the higher pouring temperature. Notably, the filling completeness is primarily governed by the flow velocity, not the Niyama criterion. For Experiments 1-3, the filling time was too long (low velocity) relative to the solidification time of the thin sections, causing premature freezing and incomplete fill.
We also investigated the effect of shell mold permeability on the vacuum process. The pressure drop across the mold wall was calculated using Darcy’s law:
$$
\Delta P_{mold} = \frac{\mu_g Q L_m}{k A}
$$
where $\mu_g$ is the viscosity of gas (air) at mold temperature, $Q$ is the volumetric flow rate of gas through the mold, $L_m$ is the mold thickness, $k$ is the permeability, and $A$ is the area. For the composite shell used (silica sol-ethyl silicate), the permeability is about $10^{-11}$ m². The resulting pressure drop is negligible compared to the applied vacuum, confirming that the mold design is adequate for high precision investment casting.
Conclusions
Through systematic experimentation, we demonstrated that high precision investment casting combined with counter-gravity casting (vacuum-assisted filling and pressurized solidification) can reliably produce thin-walled aluminum alloy castings such as the ventilation port component. The critical process parameters are the molten metal filling velocity and the pouring temperature. Our optimized conditions are a filling velocity of 100 mm/s and a pouring temperature of 725 °C, which yield castings with complete filling and no shrinkage porosity. These findings are applicable to other thin-walled geometries requiring high precision investment casting, where the avoidance of surface tension effects and thermal gradients is essential.
Future work will focus on scaling the process to larger thin-walled components and incorporating advanced simulation tools to predict the optimal parameter windows for varying geometries. The success of this study underscores the versatility of high precision investment casting in aerospace and automotive applications where weight reduction and dimensional accuracy are paramount.