In the field of advanced manufacturing, the investment casting process is renowned for its ability to produce complex geometries with high dimensional accuracy and superior surface finish. This makes it indispensable in aerospace, automotive, and other high-performance industries. However, traditional development of the investment casting process relies heavily on iterative trial-and-error experiments, leading to high costs and prolonged timelines. To address this, numerical simulation tools like ProCAST have emerged as powerful aids, enabling predictive analysis of filling, solidification, and defect formation. In this study, I focus on optimizing the investment casting process for a support component, leveraging ProCAST to evaluate temperature fields and defect distributions. Based on initial trial results, I propose three pouring system designs, simulate their performance, and validate the optimal scheme through actual production. The goal is to shift from empirical methods to a simulation-driven approach, reducing reliance on physical trials while enhancing casting quality.
The support component in question features a hollow structure with overall dimensions of 186 mm × 308 mm × 318 mm and an average wall thickness of approximately 4.5 mm. Its intricate design necessitates a carefully controlled investment casting process to avoid internal defects such as shrinkage porosity and hot tears. The material used is nickel-based superalloy K444, chosen for its excellent high-temperature properties. The chemical composition of K444 is summarized in Table 1, which is critical for setting accurate material parameters in simulations. Initial trial production employed a simple pouring system with a pouring temperature of 1420°C and a mold preheat temperature of 980°C, held for 5 hours. To promote directional solidification from bottom to top, the base plate was air-cooled while other sections were wrapped with 12 mm thick insulating wool. Despite these measures, post-casting inspections using X-ray and fluorescent testing revealed significant penetrating shrinkage defects in the upper-middle side walls. These defects indicated that the intended solidification sequence was not achieved; instead, simultaneous nucleation and growth likely created isolated liquid pools within the mushy zone, leading to unsupplemented shrinkage. This outcome underscored the need for a more robust optimization of the investment casting process.
| C | Cr | Co | W | Mo | Ti | Nb | Hf | B | Zr | Ni |
|---|---|---|---|---|---|---|---|---|---|---|
| 0.07 | 15.40 | 10.82 | 5.58 | 2.07 | 4.62 | 0.24 | 0.37 | 0.075 | 0.05 | Bal. |
To improve the investment casting process, I devised three optimized pouring system designs, designated as “Rectangular,” “V-shaped,” and “Linear” schemes. Each design aims to enhance temperature gradients and feeding capabilities. The Rectangular scheme features a surrounding gating system with nine ingates per side, insulated with 6 mm wool, while the pouring cup and horizontal runners use 12 mm wool. The V-shaped scheme reduces ingates to four per side, omits bottom runners, and increases ingate size and insulation to 12 mm to bolster feeding. The Linear scheme further refines temperature gradients by arranging five staggered ingates crosswise on both sides, with all runners insulated by 12 mm wool. These designs were modeled in ProCAST, with simulation parameters detailed in Table 2. Key settings include a pouring temperature of 1420°C, mold preheat of 980°C, pouring time of 4 seconds, and heat transfer coefficients tailored to different cooling conditions. The fraction solid criterion, defined as the volume fraction of solid within a region, was used to track solidification progression and feeding channel availability. The solidification time, indicating when each region fully solidifies, helps assess directional solidification. For defect prediction, the Niyama criterion, widely applied in investment casting process simulations, was employed to identify shrinkage porosity. The Niyama value $N$ is calculated as:
$$N = \frac{G}{\sqrt{\dot{T}}}$$
where $G$ is the temperature gradient (K/m) and $\dot{T}$ is the cooling rate (K/s). Regions with $N$ below a threshold (e.g., 0.01) are prone to shrinkage defects. This formula is integral to optimizing the investment casting process for defect minimization.
| Parameter | Value |
|---|---|
| Pouring Temperature | 1420°C |
| Mold Preheat Temperature | 980°C |
| Pouring Time | 4 s |
| Heat Transfer Coefficient (Mold-Casting) | 300 W/(m²·K) |
| Heat Transfer Coefficient (12 mm Insulation-Air) | 0.2 W/(m²·K) |
| Heat Transfer Coefficient (6 mm Insulation-Air) | 1 W/(m²·K) |
| Heat Transfer Coefficient (Air Cooling-Air) | 10 W/(m²·K) |
| Ambient Temperature | 20°C |
Simulation results for the three pouring schemes revealed distinct solidification behaviors. The fraction solid over time, plotted in Figure 1 (simulated data), shows that the Rectangular scheme cooled rapidly due to thinner insulation on ingates, reaching 60-66.7% solid fraction at the ingates by the end of solidification, which may impair feeding. In contrast, the V-shaped and Linear schemes maintained lower ingate solid fractions (33-40%), suggesting better feeding capacity. However, the Linear scheme risked early closure of feeding channels within the casting, as seen in localized high solid fractions. The solidification time distributions, summarized in Table 3, indicate that the V-shaped scheme achieved the most desirable bottom-up solidification, with the base solidifying first and the top last. This is critical for a successful investment casting process, as it ensures sequential feeding and reduces defect formation.
| Scheme | Solidification Time Range (s) | Directional Solidification Assessment |
|---|---|---|
| Rectangular | ~376–800 | Moderate, with faster cooling at ingates |
| V-shaped | ~337–800 | Excellent, clear bottom-up progression |
| Linear | ~340–800 | Good, but potential channel blockage |
Defect analysis using the Niyama criterion highlighted significant differences among the schemes. For the Rectangular scheme, shrinkage porosity exceeding 1% was predicted in the upper side walls, correlating with inadequate feeding from ingates. The Linear scheme showed similar defects due to premature isolation of liquid regions. In contrast, the V-shaped scheme exhibited minimal defects within the casting, with porosity concentrated mainly in the gating system, indicating a robust investment casting process. The defect distributions are quantified in Table 4, based on simulated porosity percentages. This analysis underscores the importance of balancing insulation and ingate design to maintain open feeding channels throughout solidification. The heat transfer during solidification can be described by the Fourier equation:
$$\frac{\partial T}{\partial t} = \alpha \nabla^2 T$$
where $T$ is temperature, $t$ is time, and $\alpha$ is thermal diffusivity. Optimizing the investment casting process involves manipulating boundary conditions, such as insulation thickness, to control $\alpha$ and achieve desired temperature gradients.
| Scheme | Porosity in Casting (%) | Porosity in Gating System (%) | Overall Defect Severity |
|---|---|---|---|
| Rectangular | >1 (in side walls) | <1 | High |
| V-shaped | <1 | >1 | Low |
| Linear | >1 (in isolated zones) | <1 | Moderate |
Based on the simulation outcomes, I selected the V-shaped pouring scheme for production validation. A small-scale trial was conducted using identical parameters: pouring at 1420°C, mold preheat at 980°C, and insulation as per the design. The resulting castings displayed smooth surfaces without visible flaws like excess metal or shortages. X-ray inspection confirmed the absence of internal shrinkage defects, validating the ProCAST predictions. This successful trial demonstrates that simulation-driven optimization can effectively replace extensive physical trials, reducing costs and time in the investment casting process. The key factors contributing to this success include enhanced temperature gradients from the V-shaped layout and sufficient feeding from adequately insulated ingates. To further generalize these findings, I derived a feeding efficiency metric $F_e$ for the investment casting process:
$$F_e = \frac{A_i \cdot t_f}{V_c \cdot \Delta T}$$
where $A_i$ is the total ingate area, $t_f$ is the feeding time, $V_c$ is the casting volume, and $\Delta T$ is the temperature difference between casting and ingates. Higher $F_e$ values indicate better feeding, as seen in the V-shaped scheme with larger ingates and prolonged liquid availability.
In conclusion, this study highlights the transformative role of numerical simulation in advancing the investment casting process. By employing ProCAST, I analyzed three pouring system designs and identified the V-shaped scheme as optimal for achieving directional solidification and minimizing defects. The simulation results aligned closely with actual production, underscoring the reliability of tools like ProCAST for predicting solidification behavior and defect formation. Key takeaways include the importance of maintaining open feeding channels through strategic insulation and ingate placement, as well as the value of criteria such as fraction solid and Niyama in guiding optimization. Future work could explore dynamic cooling controls or advanced alloy simulations to further refine the investment casting process. Ultimately, integrating simulation into casting design not only enhances quality but also promotes sustainable manufacturing by reducing material waste and energy consumption. This approach paves the way for more efficient and reliable production of complex components across industries.