As a researcher focused on advanced manufacturing techniques for aerospace applications, I have extensively investigated the challenges associated with producing large, complex, thin-walled structural parts. The turbine rear frame is a critical component in modern aero-engines, and its strut, which integrates load-bearing lugs, a concave float plate, and airfoil-shaped vanes, presents significant difficulties for traditional manufacturing. This study details my comprehensive approach, using numerical simulation as a core tool, to develop and optimize a reliable investment casting process for this high-integrity component. The investment casting process, also known as lost-wax casting, is particularly suited for such complex geometries but requires meticulous control to prevent defects.
The component in question is made from K4169 superalloy and features a challenging geometry with thick sections (up to 31 mm at the lugs) transitioning to thin walls (as low as 2.5 mm on the float plate) and intricate vane structures. Traditional trial-and-error methods for developing the investment casting process are time-consuming and costly. Therefore, I employed ProCAST simulation software to virtually prototype the casting, analyze filling patterns, pressure evolution within the mold cavity, and predict shrinkage porosity and cavity defects. This proactive simulation-based approach is integral to modernizing the investment casting process.
The foundation of any accurate simulation lies in correctly defining the physical model and boundary conditions. For this investment casting process analysis, I created a detailed finite element model that included the wax pattern assembly (part, gating, risers) and the ceramic shell. A dual-node technique was applied at the metal-shell interface to accurately capture the thermal discontinuity. The governing equations for fluid flow, heat transfer, and stress were solved iteratively. To specifically analyze the risk of air entrapment or “gas suffocation,” a significant concern in the investment casting process for parts with deep concave regions, I activated the gas entrapment and pressure setting modules. This allows the software to calculate the pressure difference between the evolving metal pressure and the ambient reference, visualizing pressure evolution within the cavity. The key parameters defining this virtual investment casting process are summarized in the table below.
| Parameter | Value / Specification |
|---|---|
| Alloy Material | K4169 Superalloy |
| Shell Material | Mullite-based Ceramic |
| Cooling Method | Vacuum Cooling |
| Shell Thickness | 8 mm (average) |
| Pouring Temperature | 1450 °C |
| Shell Preheat Temperature | 800 °C |
| Furnace Chamber Dimensions | Φ2 m × 2.4 m |
| Interface Heat Transfer Coefficient | 500 W·m⁻²·K⁻¹ |
| Filling Percentage (for analysis) | 100% |
Analysis of the Initial Investment Casting Process Design
The initial gating system was designed with two primary ingates attached to the thick load-bearing lugs, expecting the metal to flow downward into the float plate and then up into the vane sections. Simulation of this initial investment casting process revealed several critical issues.
Filling Pattern and Cavity Pressure Evolution: The filling sequence was unstable. By 40% fill, the metal had rapidly filled over 90% of the vanes and lugs, but a large portion of the concave float plate remained empty. The pressure in the unfilled regions of the vanes and lugs exceeded 0.22 MPa, indicating severe gas entrapment. The metal velocity at the junction between the vane and float plate was excessively high, around 2.2 m/s, promoting turbulence. This turbulent flow in an investment casting process disrupts the thermal gradient and can lead to defects. The pressure continued to build in the last unfilled region of the float plate, exceeding 0.27 MPa at 60% fill, confirming a major “gas suffocation” issue that would likely result in misruns or cold shuts.
Prediction of Solidification Defects: The Niyama criterion and shrinkage porosity modules in ProCAST were used to predict defects. The simulation clearly showed areas of high risk for shrinkage porosity and cavities, particularly in the central and ribbed sections of the airfoil vanes and in two locations on the float plate. The defect formation can be modeled conceptually by considering the thermal gradient \(G\) and the cooling rate \(R\). Porosity risk is often inversely related to the \(G/\sqrt{R}\) ratio. In areas where solidification isolates liquid pockets, the continuous feeding required in the investment casting process is interrupted, leading to shrinkage. This can be represented as:
$$ V_{shrinkage} \propto \beta \cdot (V_{casting} – V_{solid}) $$
where \(V_{shrinkage}\) is the volume of shrinkage defect, \(\beta\) is the alloy’s volumetric shrinkage coefficient, \(V_{casting}\) is the casting volume, and \(V_{solid}\) is the volume of solid metal. The isolated hot spots in the vanes and float plate, caused by poor thermal management in the initial investment casting process design, became sites for this shrinkage.
| Issue Category | Observation from Simulation | Potential Physical Defect |
|---|---|---|
| Filling Dynamics | High, turbulent metal velocity (~2.2 m/s) at critical junctions. | Erosion of shell, oxide inclusion, turbulence-driven porosity. |
| Cavity Pressure | Localized pressure >0.27 MPa in unfilled float plate region. | Gas entrapment (“suffocation”), leading to misruns. |
| Thermal/Solidification | Isolated hot spots in vane centers and float plate. | Macro- and micro-shrinkage porosity, cavities. |
| Feeding Path | Long, restricted path from lugs to vane tips. | Inadequate feeding to compensate for solidification shrinkage. |
Optimization of the Investment Casting Process
Based on the simulation findings, I implemented a multi-faceted optimization strategy for the investment casting process. The goal was to achieve a tranquil, progressive fill, eliminate gas entrapment, and establish a directional solidification pattern toward functional risers.
1. Mold and Gating System Redesign: The most critical change was modifying the exhaust system. The single vent on the float plate was replaced with three strategically placed vents to ensure no isolated air pockets could form. This is a vital principle in optimizing any investment casting process for parts with deep draws. The entire assembly was rotated approximately 15° within the flask, aligning the vent plane optimally for gas escape. The flask size was increased to 440 mm × 360 mm × 600 mm to accommodate these changes and improve the overall thermal mass of the mold.
2. Process Parameter Adjustment: The pouring time was increased by 50% compared to the initial design. This reduces the metal velocity, minimizing turbulence. The relationship between flow rate \(Q\), cross-sectional area of the gate \(A_g\), and velocity \(v\) is given by \(Q = A_g \cdot v\). By increasing the fill time \(t\) (where \(Q = V_{casting}/t\)), we effectively reduce \(v\), promoting laminar flow.
$$ v_{avg} = \frac{V_{casting}}{A_g \cdot t} $$
3. Thermal Management Enhancement: To control the solidification sequence, exothermic padding (20 mm thick iron sand) was applied to the central and ribbed sections of the vane. This padding acts as a chill, accelerating cooling in these thick, problematic areas and helping to establish a thermal gradient that promotes feeding from the designed risers attached to the lugs. The optimized gating and feeding system for the investment casting process is summarized below.
| Optimization Area | Specific Action | Intended Effect |
|---|---|---|
| Exhaust & Venting | Added two additional vents on float plate (total 3). | Eliminate gas entrapment, ensure continuous pressure relief. |
| Orientation | Rotated pattern 15° within flask. | Improve natural flow path and venting efficiency. |
| Pouring Parameters | Increased total pouring time by 50%. | Reduce metal velocity, minimize turbulence. |
| Solidification Control | Applied exothermic pads to vane center and ribs. | Transform isolated hot spots into directionally solidified zones. |
| Feeding System | Maintained top risers on lugs but improved thermal path. | Ensure liquid feed metal is available until critical sections solidify. |
Simulation Results of the Optimized Investment Casting Process
Running the numerical simulation with the optimized model yielded dramatically improved results for the investment casting process.
Stable Filling and Pressure Management: The metal now filled the component progressively and smoothly. At 20% fill, the metal front was calm, and the cavity pressure everywhere was below 0.07 MPa—a normal value. By 40% fill, the vanes were completely filled without any pressure build-up, and the average metal velocity in critical areas was reduced to about 27% of the initial process’s velocity. The three vents functioned effectively, preventing any pressure from exceeding 0.7 MPa until the very end of filling, where pressure slightly increased only in the vent channels themselves, which is inconsequential. This confirms the elimination of the “gas suffocation” phenomenon in the optimized investment casting process.
Elimination of Predicted Shrinkage Defects: The most significant outcome was the virtual elimination of shrinkage porosity and cavities from the main body of the strut. The solidification simulation showed a clear directional progression from the thin vane walls and float plate toward the thicker lugs and finally to the top risers. The exothermic pads successfully modified the thermal field, preventing the formation of isolated liquid pools. The feeding paths were open throughout solidification. The Niyama criterion values across the part were now within a safe range. The final solidification time \(t_f\) for different sections can be approximated by Chvorinov’s rule:
$$ t_f = B \cdot \left( \frac{V}{A} \right)^n $$
where \(B\) is the mold constant, \(V\) is volume, \(A\) is surface area, and \(n\) is an exponent (typically ~2). The optimization ensured that the modulus \((V/A)\) was lowest for the vane tips and highest for the risers, creating the desired solidification gradient essential for a sound investment casting process.
| Performance Metric | Initial Process | Optimized Process | Improvement |
|---|---|---|---|
| Max Metal Velocity (Critical Junction) | ~2.2 m/s | ~0.6 m/s | ~73% Reduction |
| Max Cavity Pressure (in part) | >0.27 MPa | <0.7 MPa (in vents only) | Elimination of “suffocation” |
| Predicted Shrinkage in Vanes | Severe (Center & Ribs) | None | Complete Elimination |
| Predicted Shrinkage in Float Plate | Present (2 locations) | None | Complete Elimination |
| Filling Stability | Unstable, Turbulent | Stable, Laminar | Dramatic Improvement |
Industrial Validation and Production
The ultimate test of any simulated investment casting process is its performance in real-world production. Based on the confidence gained from the numerical results, I proceeded with the physical implementation of the optimized investment casting process. A batch of five prototype struts was cast using the standard熔模铸造 (investment casting) production chain: wax pattern assembly of the optimized design, ceramic shell building, dewaxing, firing, vacuum pouring of K4169 alloy, and post-cast processing.
All five prototypes were subjected to non-destructive testing using digital radiography (DR) in accordance with the stringent ASTM E192 standard. The DR images confirmed the absence of internal defects such as gas pores, inclusions, and shrinkage cavities within the critical sections of the strut—precisely as predicted by the simulation. The component quality fully met the specification requirements. Encouraged by this success, a subsequent production batch of 40 struts was manufactured using the same optimized investment casting process parameters. All 40 parts passed the same rigorous inspection criteria, demonstrating the robustness and repeatability of the optimized investment casting process. This successful translation from virtual simulation to physical production underscores the immense value of numerical modeling in modernizing and de-risking the investment casting process for critical aerospace components.
Conclusion
In this study, I have demonstrated a systematic, simulation-driven methodology for developing a high-yield investment casting process for a complex turbine rear frame strut. The initial process design, while conceptually straightforward, harbored significant flaws leading to turbulent filling, gas entrapment, and shrinkage defects. Through detailed numerical simulation using ProCAST, these issues were not only identified but also understood in terms of their root causes—excessive velocity, poor venting, and unfavorable solidification sequences. The optimization strategy directly addressed these root causes: enhancing venting, reducing pouring speed, and manipulating the thermal field with exothermic materials. The resulting optimized investment casting process showed, in simulation, a stable fill, negligible gas pressure issues, and a defect-free solidification pattern. This virtual success was conclusively validated by actual castings that met the highest quality standards. This work reaffirms that numerical simulation is an indispensable tool for the advancement of the investment casting process, enabling first-pass success, reducing development time and cost, and ensuring the reliability of safety-critical components in demanding applications like aerospace propulsion. The principles applied here—focusing on controlled filling, active venting, and directional solidification—are universally applicable to enhancing the robustness of the investment casting process for a wide range of complex geometries.