In the field of advanced manufacturing, titanium alloys are widely utilized in aerospace, marine, and energy sectors due to their high strength-to-weight ratio and excellent corrosion resistance. The trend towards lightweight and integrated components has driven the development of large, complex, thin-walled titanium castings. However, these structures are prone to metal casting defects such as shrinkage porosity and voids, which compromise mechanical properties, fatigue life, and reliability. Centrifugal investment casting combines the precision of investment molding with the enhanced filling and feeding capabilities of rotational forces, making it a key method for producing large titanium alloy casings. This study focuses on the simulation and analysis of metal casting defects in a gas turbine intermediate casing, employing numerical modeling to optimize the process and improve product quality.
We begin by examining the structural characteristics of the casing, which features a maximum outer diameter of 1350 mm, a height of 1000 mm, and wall thicknesses ranging from 10 mm to 100 mm. The casing includes multiple annular flanges, mounting bosses, and internal ribs, contributing to its complexity. Such geometries exacerbate challenges in achieving complete filling and uniform solidification, leading to metal casting defects like shrinkage porosity. To address this, we designed an open-bottom gating system with a spiral configuration, where the cross-sectional areas of the sprue, runner, and ingate follow a ratio of 1:1.5:2.3. This design promotes steady melt flow and reduces turbulence during filling. Additionally, risers were placed at thick sections and transition zones to facilitate feeding. The gating system was modeled and simulated using ProCAST software, with a mesh of over 10 million elements to capture detailed thermal and fluid dynamics.
The simulation parameters included a ZTC4 titanium alloy with a pouring temperature of 1700°C and a mold preheat temperature of 550°C. Centrifugal casting was performed at a low rotational speed of 80 rpm to balance filling and defect formation. The governing equations for fluid flow and heat transfer during casting are critical for predicting metal casting defects. The Navier-Stokes equations describe the melt flow:
$$ \rho \left( \frac{\partial \mathbf{v}}{\partial t} + \mathbf{v} \cdot \nabla \mathbf{v} \right) = -\nabla p + \mu \nabla^2 \mathbf{v} + \rho \mathbf{g} + \mathbf{F}_{\text{centrifugal}} $$
where \( \rho \) is density, \( \mathbf{v} \) is velocity, \( p \) is pressure, \( \mu \) is dynamic viscosity, \( \mathbf{g} \) is gravity, and \( \mathbf{F}_{\text{centrifugal}} \) represents centrifugal forces. Heat transfer during solidification is modeled using the energy equation:
$$ \rho c_p \frac{\partial T}{\partial t} = \nabla \cdot (k \nabla T) + \rho L \frac{\partial f_s}{\partial t} $$
Here, \( T \) is temperature, \( c_p \) is specific heat, \( k \) is thermal conductivity, \( L \) is latent heat, and \( f_s \) is the solid fraction. The evolution of \( f_s \) is described by the Scheil equation for non-equilibrium solidification:
$$ f_s = 1 – \left( \frac{T_m – T}{T_m – T_l} \right)^{1/(k_p – 1)} $$
where \( T_m \) is the melting point, \( T_l \) is the liquidus temperature, and \( k_p \) is the partition coefficient. These equations help identify regions prone to metal casting defects by tracking temperature gradients and solidification rates.
During the filling simulation, the melt entered the sprue and distributed evenly through the spiral runner, achieving a parabolic flow front that minimized air entrapment. The casing filled completely in approximately 10 seconds, with the outer ring filling first, followed by the inner sections via the ribs. This orderly filling reduced the risk of cold shuts and misruns, common precursors to metal casting defects. The solidification analysis revealed that areas with high thermal mass, such as thick flanges and transition corners, exhibited delayed solidification. For instance, the solid fraction distribution showed that after 21 seconds, the bulk of the casting had a solid fraction of around 40%, while after 91 seconds, unresolved hot spots persisted at the bottom ingate connections and upper riser zones. These regions are susceptible to shrinkage porosity due to inadequate feeding and slow heat dissipation.
To quantify the susceptibility to metal casting defects, we applied the Niyama criterion, defined as:
$$ Ny = \frac{G}{\sqrt{\dot{T}}} $$
where \( G \) is the temperature gradient and \( \dot{T} \) is the cooling rate. A threshold value of 0.9 was used to predict shrinkage porosity. The results indicated that 13.68% of the casting volume was affected by such defects, primarily at thick mounting bosses, flow path transitions, and the bottom gating areas. The table below summarizes the defect distribution based on simulation data:
| Location | Defect Type | Volume Fraction (%) | Primary Cause |
|---|---|---|---|
| Thick Flanges | Shrinkage Porosity | 5.2 | Insufficient Feeding |
| Flow Path Transitions | Shrinkage Voids | 4.1 | Heat Accumulation |
| Bottom Ingate Connections | Macroporosity | 3.8 | High Thermal Influence |
| Riser Bases | Microporosity | 0.58 | Localized Solidification |
The centrifugal force field alters the feeding dynamics compared to gravity casting. The effective feeding pressure \( P_{\text{eff}} \) in centrifugal casting can be expressed as:
$$ P_{\text{eff}} = \rho \omega^2 r \Delta r + \rho g h $$
where \( \omega \) is angular velocity, \( r \) is radius, \( \Delta r \) is the radial distance, and \( h \) is height. At low speeds like 80 rpm, gravitational effects remain significant, leading to complex pressure distributions that contribute to metal casting defects in isolated liquid pools. Simulation results showed that the final solidification zones occurred near the top risers and bottom gates, highlighting the need for optimized gating and riser design.
Experimental validation was conducted by producing actual castings using the optimized gating system. X-ray inspection revealed shrinkage defects at 22 locations, of which 18 aligned with simulation predictions, yielding an accuracy of 81.8%. This confirms the reliability of numerical models in identifying metal casting defects. The images from X-ray analysis clearly showed porosity in thick sections and transition areas, underscoring the impact of local solidification conditions.
Based on these findings, we propose several improvements to mitigate metal casting defects. First, increasing the modulus of risers and applying tapered subsidies with angles of 7° to 9° can enhance directional solidification. The feeding efficiency \( \eta \) of a riser can be estimated as:
$$ \eta = \frac{V_{\text{feed}}}{V_{\text{riser}}} = 1 – \frac{f_s}{\rho} $$
where \( V_{\text{feed}} \) is the volume available for feeding and \( V_{\text{riser}} \) is the riser volume. Second, using annular risers on upper flanges and conical risers on thick bosses improves metal supply. Third, controlling mold shell thickness in critical areas, such as thinning the coating at transition corners, accelerates cooling. Embedding chills or cooling channels in stubborn hot spots further reduces solidification time. The table below outlines key improvement measures and their expected impacts:
| Improvement Measure | Application Area | Expected Reduction in Defects (%) |
|---|---|---|
| Increased Riser Modulus | Thick Flanges | 30-40 |
| Tapered Subsidies | Transition Zones | 20-30 |
| Shell Thickness Control | Bottom Gates | 15-25 |
| Embedded Chills | Hot Spots | 10-20 |
In conclusion, this study demonstrates the effectiveness of centrifugal investment casting combined with numerical simulation in addressing metal casting defects in large titanium alloy casings. The open-bottom spiral gating system ensured stable filling, while solidification analysis identified critical defect-prone areas. Validation through X-ray inspection confirmed high prediction accuracy. The proposed measures, such as optimized riser design and localized cooling, provide a foundation for high-quality mass production. Future work will focus on refining the simulation models and exploring advanced materials to further reduce metal casting defects in complex components.