As global technological advancement accelerates, the demand for energy continues to rise, making material conservation a critical goal for nations worldwide. In the foundry industry, traditional casting methods often consume significant labor and financial resources, whereas investment casting excels in achieving near-net or net-shape manufacturing with minimal machining allowances. The integration of computer simulation with practical casting processes has become increasingly prevalent in investment casting, enabling the production of components that are both scientifically validated and suitable for industrial applications. The 304 stainless steel ball valve is widely used in fluid control systems, such as pipeline control valves, where it must withstand high pressures. Consequently, stringent quality standards are imposed on its production, with defects like cracks, cold shuts, shrinkage porosity, and shrinkage holes being unacceptable. Under conventional investment casting practices, the production of stainless steel ball valves often relies on empirical methods, frequently leading to shrinkage defects at the corners between the valve body and flanges, resulting in low yield rates. To mitigate shrinkage issues, improve product yield, and reduce development costs, redesigning the gating system in the investment casting process presents a viable solution.
In recent studies on investment casting, researchers have explored various optimizations. For instance, centrifugal investment casting of high-Nb titanium aluminide exhaust valves was investigated through simulation and experimentation, confirming that a shell filling time of 3.8 s, a powder-to-liquid ratio of 1.5:1, and a stirring time of 30 minutes yield castings with acceptable microstructures. Similarly, shrinkage problems in turbine casing brackets were addressed by modifying venting channels, adjusting pouring positions and durations, and optimizing solidification sequences, ultimately meeting ASTM E192 standards. Additionally, the investment casting process for automotive transmission control rods was improved by altering the gating system, effectively eliminating shrinkage defects in 45 steel components. Numerical simulations of large-section steel main bearing seats using ProCAST software revealed that horizontal pouring and the use of external chills significantly reduce shrinkage in critical areas. Research on 17-4PH stainless steel impellers for centrifugal pumps via AnyCasting software identified complex internal structures and uneven thickness as causes of shrinkage and gas holes, with defects resolved by adding and sizing risers alongside parameter adjustments. For 431 stainless steel fishing net hooks, analysis of solidification processes pinpointed causes of shrinkage, cold shuts, pinholes, and cracks, leading to the adoption of a 90-degree tree-shaped mold and comparison of circular versus hexagonal runner sections to eliminate defects at hook tips. However, studies specifically optimizing shrinkage defects in 304 stainless steel hydraulic valves remain limited. To enhance overall casting performance and reduce scrap rates, this work simulates the traditional investment casting process for a 304 stainless steel ball valve, identifies defect causes, and through gating system redesign and orthogonal experiments, determines an optimal parameter set to minimize shrinkage at valve body-flange corners. Production trials confirm the elimination of these defects.
The stainless steel ball valve features envelope dimensions of 186 mm × 186 mm × 120 mm, with an average flange thickness of 12.5 mm. The material is 304 stainless steel, whose primary chemical composition is summarized in Table 1.
| C | Cr | Mn | Mo | Ni | S | Si |
|---|---|---|---|---|---|---|
| 0.08 | 18 | 1.5 | 0.5 | 8 | 0.03 | 1 |
In the traditional investment casting approach, a three-dimensional model of the ball valve was created in SolidWorks and imported into ProCAST software’s Mesh module for preprocessing. Mesh quality was checked with a global element size set to 4, generating 116,162 surface meshes. After adding a 6 mm shell, 666,312 volume meshes were produced. The investment casting material was defined as 304 stainless steel with the composition in Table 1, while the shell material was zircon sand. Process parameters were set in the Cast module: pouring temperature of 1550°C, shell preheating temperature of 1150°C, gravity filling, and pouring speed of 1.5 kg/s. The interface heat transfer coefficient between stainless steel and zircon sand was selected as COINC type, calculated using the following model:
$$ Q = \text{Flux} + h(T – T_{\alpha}) + \sigma \varepsilon (T^4 – T_{\alpha}^4) $$
where \( Q \) is the interface heat transfer coefficient, \( \text{Flux} \) is the heat flux rate, \( T \) is the casting surface temperature, \( T_{\alpha} \) is the ambient temperature, \( \varepsilon \) is the material emissivity, \( h \) is the convective heat transfer coefficient, and \( \sigma \) is the Stefan-Boltzmann constant. Based on calculations and literature, \( Q = 500 \, \text{W/(m}^2 \cdot \text{K)} \) was used.
Simulation results under traditional investment casting conditions revealed significant shrinkage defects, with a maximum shrinkage porosity rate of 13.2%. These defects were concentrated at the corners between the valve body and flanges, attributed to uneven wall thickness and poor heat dissipation in concave areas, leading to non-sequential solidification. After filling, molten metal in the cavity begins to solidify, with corners cooling faster due to greater air exposure, while slower-cooling areas solidify later. Liquid metal feeding from later-solidifying to earlier-solidifying regions is hindered at corners distant from ingates, causing shrinkage voids. This aligns with practical defects observed in investment casting production, validating the simulation.
The gating system design critically influences casting quality in investment casting. Typically comprising a pouring cup, sprue, runner, and ingates, its optimization can substantially reduce defects. Given the ball valve’s structure, adjusting the gating system’s layout and dimensions effectively addresses shrinkage. The pouring position—where the casting is oriented in the mold—is crucial in investment casting design, impacting quality, dimensional accuracy, and molding processes. Considering the ball valve’s geometry, placing ingates on both ends of the valve body prevents shrinkage at corners, ensuring critical machined surfaces face downward or sideways. This design employs top gating. The sprue was designed with a length of 100 mm and diameter of 24 mm, with a cross-sectional area 1.4 times the total ingate area to facilitate gravity filling and reduce shrinkage. A buffer was placed 30 mm from the ingates at the runner base. Both sprue and runner have circular cross-sections. Ingate cross-sectional shapes include rectangular, circular, annular, fan-shaped, and slot types. Given the eight threaded holes on the flange faces, a fan-shaped ingate is suitable. Ingate length should be as short as possible for easy cutting, typically 10–15 mm; here, it was set to 12 mm. Two ingates were placed on opposite end faces. The redesigned gating system structure is shown in Figure 4 (referenced conceptually, without explicit numbering as per instructions).
After redesigning the gating system, the model was re-imported into ProCAST for simulation. Mesh density was set to 4, producing 147,072 surface meshes and 785,303 volume meshes. Parameters were initially set as in the traditional scheme, with gravity filling direction optimized to -z for best filling and minimal shrinkage. Material properties and interface coefficients remained unchanged. Results showed improved shrinkage distribution, with a maximum shrinkage porosity rate of 3.75%. Solidification analysis indicated mostly sequential solidification without isolated liquid zones. While gating system optimization reduced shrinkage significantly, minor porosity persisted, prompting further parameter adjustment.
To maximize defect reduction, process parameters were optimized via orthogonal experiments. Based on traditional parameters (pouring temperature 1550°C, pouring speed 1.5 kg/s, shell preheating temperature 1150°C), three factors were considered: pouring temperature, pouring speed, and shell preheating temperature. A three-level orthogonal array L9(3³) was used, assuming no interactions, with the response variable being the maximum shrinkage porosity rate. Factors and levels are listed in Table 2.
| Level | A: Pouring Temperature (°C) | B: Pouring Speed (kg/s) | C: Shell Preheating Temperature (°C) |
|---|---|---|---|
| 1 | 1520 | 1.0 | 1120 |
| 2 | 1550 | 1.5 | 1150 |
| 3 | 1580 | 2.0 | 1180 |
The orthogonal experimental design and simulation results are presented in Table 3.
| Experiment No. | A (°C) | B (kg/s) | C (°C) | Maximum Shrinkage Porosity Rate (%) |
|---|---|---|---|---|
| 1 | 1520 | 1.0 | 1120 | 3.10 |
| 2 | 1520 | 1.5 | 1150 | 3.88 |
| 3 | 1520 | 2.0 | 1180 | 3.74 |
| 4 | 1550 | 1.0 | 1150 | 2.29 |
| 5 | 1550 | 1.5 | 1180 | 3.52 |
| 6 | 1550 | 2.0 | 1120 | 4.51 |
| 7 | 1580 | 1.0 | 1180 | 2.44 |
| 8 | 1580 | 1.5 | 1120 | 4.27 |
| 9 | 1580 | 2.0 | 1150 | 3.93 |
Range analysis was conducted to assess factor influence. The mean maximum shrinkage porosity rate for each factor level was calculated, and ranges (R) were derived as the difference between maximum and minimum means. Results: \( R_A = 0.1363 \), \( R_B = 1.4563 \), \( R_C = 0.7954 \). Thus, factor influence order is \( R_B > R_C > R_A \), indicating pouring speed has the greatest effect on shrinkage, followed by shell preheating temperature, and pouring temperature the least. Considering that in investment casting, excessively high shell preheating temperatures should be avoided to prevent shrinkage, the optimal parameter combination was determined as A2B1C2: pouring temperature 1550°C, pouring speed 1.0 kg/s, shell preheating temperature 1150°C.
Variance analysis was performed to confirm significance, as shown in Table 4. The F-values indicate that pouring speed has the most significant effect, shell preheating temperature is next, and pouring temperature is least significant, corroborating the range analysis and optimal combination.
| Factor | Sum of Squares | Degrees of Freedom | Mean Square | F-value |
|---|---|---|---|---|
| Pouring Temperature (A) | 0.0104 | 2 | 0.0052 | 13.96 |
| Pouring Speed (B) | 1.0637 | 2 | 0.5319 | 119.30 |
| Shell Preheating Temperature (C) | 0.3422 | 2 | 0.1711 | 38.41 |
| Error | 0.0356 | 2 | 0.0178 | – |
| Total | 1.4519 | 8 | – | – |
With the optimal parameters set, numerical simulation was rerun using ProCAST. The optimized investment casting process resulted in a maximum shrinkage porosity rate of 2.29%, a substantial improvement over the traditional 13.2%. Production trials under these optimized conditions yielded 304 stainless steel ball valve castings with no shrinkage defects at valve body-flange corners, confirming the effectiveness of the approach. This success underscores the value of combining gating system redesign with systematic parameter optimization in investment casting.
The solidification process in investment casting can be further analyzed using thermal models. The rate of solidification is governed by heat transfer equations, such as the Fourier heat conduction law:
$$ q = -k \nabla T $$
where \( q \) is the heat flux, \( k \) is the thermal conductivity, and \( \nabla T \) is the temperature gradient. For sequential solidification to prevent shrinkage, the temperature gradient must be controlled, often expressed as:
$$ G = \frac{dT}{dx} $$
where \( G \) is the temperature gradient. A higher \( G \) promotes directional solidification, reducing shrinkage risk. In investment casting, the thermal modulus \( M \) is also considered:
$$ M = \frac{V}{A} $$
where \( V \) is volume and \( A \) is surface area. Regions with lower \( M \) solidify faster, influencing feeding paths. Optimizing gating and parameters ensures favorable \( G \) and \( M \) distributions.
Moreover, the Niyama criterion is often used to predict shrinkage porosity in castings, given by:
$$ N_y = \frac{G}{\sqrt{\dot{T}}} $$
where \( \dot{T} \) is the cooling rate. A higher \( N_y \) value indicates lower shrinkage susceptibility. In investment casting simulations, this criterion helps identify defect-prone areas.
In conclusion, this study demonstrates a comprehensive approach to optimizing investment casting for 304 stainless steel ball valves. By leveraging numerical simulation with ProCAST, traditional defects were identified and addressed through gating system redesign and orthogonal experiment-based parameter optimization. The optimal investment casting parameters—pouring temperature 1550°C, pouring speed 1.0 kg/s, and shell preheating temperature 1150°C—reduced maximum shrinkage porosity to 2.29%, eliminating corner defects in production. This methodology highlights the efficacy of simulation-driven design in investment casting, contributing to improved yield, cost savings, and enhanced product quality in precision manufacturing. Future work could explore advanced investment casting techniques, such as vacuum-assisted pouring or additive manufacturing for mold shells, to further refine process efficiency and defect control in stainless steel components.