The relentless pursuit of performance and efficiency in aerospace, maritime, and high-end sporting equipment drives the adoption of advanced materials and manufacturing techniques. Titanium alloys, renowned for their exceptional strength-to-weight ratio, superior corrosion resistance, and biocompatibility, stand at the forefront of this evolution. Among the various shaping processes, the investment casting process offers unparalleled advantages for producing complex, net-shape titanium components with excellent surface finish. This near-net-shape capability is crucial for minimizing costly machining of expensive titanium alloys. However, the unique physical properties of titanium—its low density, high reactivity, and poor fluidity and feeding characteristics under vacuum casting conditions—pose significant challenges. These challenges are acutely manifested in components with extreme geometry, such as long, slender rods with severe sectional thickness variations. Defects like shrinkage porosity, hot tears, and high residual stress are common, often necessitating expensive post-processing like Hot Isostatic Pressing (HIP). This study details a first-person engineering approach to overcoming these challenges for a specific rod-shaped titanium alloy casting, leveraging numerical simulation as the core tool to design, analyze, and optimize the investment casting process, thereby achieving sound castings without the need for HIP.
The subject component, designated for a high-performance application, is a prime example of a geometry that tests the limits of conventional titanium casting wisdom. The core structure is an elongated rod with a length-to-width ratio of approximately 9.3:1. While one end features relatively uniform wall thickness, the critical area lies at the junction between the rod and a flanking feature, where a drastic change in cross-section creates a severe thickness ratio of 10.5:1. The technical specifications mandate ZTC4 alloy (a Chinese equivalent to Ti-6Al-4V), with all surfaces except one machined hole required to be as-cast to a dimensional tolerance of GB/T 6414-2017 CT6. The foremost technical hurdle was to devise an investment casting process that could guarantee the complete absence of shrinkage defects and maintain low residual stress precisely at these critical locations: the machined hole periphery and the root of the thick-section junction, all while forgoing the safety net of HIP treatment.
Traditional approaches to the investment casting process for such geometries often rely on directional solidification principles, aiming to create a steep thermal gradient. An initial process scheme (Scheme 1) was designed based on this theory. The pattern assembly was oriented vertically, with the thicker end placed at the top. A top-gating system was employed, where the ingate also acted as a feeding riser. To counteract potential distortion, stabilizing ribs were added to the thin ends of the rod. The gating system was designed with an open ratio (Sprue:Runner:Ingate) of 1:1.02:2.06 to facilitate rapid filling and establish a strong thermal gradient. Concurrently, an alternative philosophy was explored. Scheme 2 was designed based on the theory of equilibrium solidification, which emphasizes shorter feeding paths and more uniform thermal distribution to reduce stress. Here, the rod was positioned horizontally. Multiple ingates were placed laterally at both ends of the rod component to shorten the metal flow and feeding distance. The gating ratio was adjusted to 1:1.02:1.5. The core challenge was to predict which investment casting process philosophy would yield a defect-free part. This is where numerical simulation transitioned from a supportive tool to the central decision-making platform.
A commercial casting simulation software was employed to model the filling, solidification, and stress development for both process schemes. The material properties of ZTC4 alloy and the ceramic shell (approximately 12mm thick, preheated to 40°C) were defined. The vacuum casting conditions were simulated to account for the limited metallostatic pressure. The analysis focused on two key outputs: the prediction of shrinkage porosity (Niyama criterion) and the distribution of residual stress.
The simulation of Scheme 1 revealed significant issues inherent to the directional approach for this geometry. The temperature field did show a vertical gradient, but the long, thin shape and the presence of stabilizing ribs created complex thermal nodes. As predicted by the simulation, macro-shrinkage cavities formed at two locations: the severe thickness-change junction and within the stabilizing rib at the thin end of the rod. The stress analysis was even more concerning. High residual stress concentrations, exceeding 300 MPa, were localized at the base of the stabilizing ribs and the thick-section junction. The stress concentration factor $K_t$ at these notches can be conceptually related to the geometry by equations like:
$$ \sigma_{max} = K_t \cdot \sigma_{nom} $$
where $\sigma_{max}$ is the peak stress, and $\sigma_{nom}$ is the nominal stress. The combination of a shrinkage cavity (acting as a crack initiator) and high tensile residual stress at the same location creates a high-risk scenario for in-service failure. The post-casting removal of the stabilizing ribs would also lead to significant stress redistribution and part distortion, compromising dimensional accuracy.
| Feature | Scheme 1 (Directional) | Scheme 2 (Equilibrium) |
|---|---|---|
| Orientation | Vertical | Horizontal |
| Gating Type | Top gating, single ingate as riser | Side gating, multiple ingates |
| Solidification Principle | Sequential, long thermal gradient | Equilibrium, shorter feeding paths |
| Simulated Major Shrinkage | At thick junction AND thin-end rib | At thick junction AND ingate hot spot |
| Simulated Stress Concentration | Very High (>300 MPa at ribs) | High (>267 MPa at ingate junction) |
| Key Risk | Defect+Stress combo at critical zone; distortion | Defect at ingate; stress at gate junction |
The simulation of Scheme 2 presented a different set of outcomes. The filling sequence was more balanced, and the solidification fronts progressed more uniformly from multiple points. Shrinkage porosity was predicted at the thick-section junction and, notably, at a hot spot formed where the lateral ingate met the thin end of the rod. The stress field showed a concentration at this same ingate-part junction, with values around 267 MPa. While the stress magnitude was slightly lower than in Scheme 1, the gradient was still significant. Crucially, the defect at the thick junction was still present, meaning the core requirement was not met. However, the simulation provided critical insight: the equilibrium solidification principle showed promise by reducing overall stress and isolating one major defect to a non-critical area (the ingate junction). The path forward was to optimize this scheme based on the simulation learnings.
The numerical simulation results provided a clear mandate for optimization focused on the investment casting process for equilibrium solidification. Three critical modifications were made to Scheme 2:
- Elimination of Stabilizing Ribs: The ribs were identified as unnecessary stress concentrators and creators of thermal nodes. Their removal simplified the pattern and reduced the risk of shrinkage and stress.
- Casting the Machined Hole: Contrary to initial intuition, the simulation suggested that attempting to cast a solid section (to be later machined into the hole) created a feeding problem. Casting the 10mm hole as a cored feature resulted in more uniform wall thickness around it, promoting directional solidification toward the adjacent ingate and eliminating the shrinkage risk in that critical area.
- Strategic Ingate Repositioning: The ingates were relocated for optimal thermal management. The side ingates were placed symmetrically from the runner to ensure simultaneous filling. Most importantly, a new top ingate was directly attached to the problem area—the severe thickness-change junction. This ingate would act as a dedicated feeder, directly supplying liquid metal to the last region to solidify, thereby actively preventing shrinkage formation there.
The optimized process scheme was subjected to a final simulation. The results confirmed the efficacy of the changes. The filling pattern was smooth and simultaneous for all cavities in the cluster. The solidification sequence showed the thick junction solidifying last, fed directly by its dedicated ingate. The shrinkage porosity prediction showed that the only remaining macro-shrinkage was successfully relocated to the top of this new ingate itself—a location that would be removed during finishing. The critical zones (hole periphery and thick junction root) were predicted to be sound. The residual stress distribution was significantly improved, with the maximum von Mises stress in the part body reduced and gradients smoothed. The stress differential at the thick junction was now below 90 MPa, dramatically lowering the risk of stress-corrosion cracking or distortion. This final, simulation-validated investment casting process was approved for production.
The production was executed strictly according to the simulation parameters. Medium-temperature wax patterns were produced and assembled into the optimized tree configuration. The ceramic shell was built using a silica sol binder system to 8.5 layers, achieving the simulated thickness. The shells were dewaxed, fired, and cast in a vacuum arc skull melting furnace with a mold temperature of approximately 40°C, using ZTC4 alloy. After standard shell removal and finishing, the castings underwent non-destructive evaluation. Radiographic inspection confirmed the simulation’s prediction: no internal shrinkage defects were detected in the critical functional areas of the rods. Fluorescent Penetrant Inspection (FPI) revealed no surface-breaking cracks or hot tears, indicating a benign residual stress state as forecasted. Dimensional analysis confirmed compliance with the CT6 tolerance grade. A batch of 100 castings was produced using this optimized investment casting process. The first-pass yield from radiographic inspection was 98%, with only two units rejected for isolated non-shrinkage inclusions. The FPI yield was 100%. This high yield, achieved without HIP, validated the simulation-driven process optimization and resulted in substantial cost savings.
This case study underscores the transformative role of numerical simulation in modern investment casting process development, particularly for challenging titanium alloys. The following key conclusions and generalized principles can be derived:
- For slender titanium castings with high aspect ratios (>9:1), an investment casting process designed on equilibrium solidification principles—featuring shorter flow paths, multiple feeding points, and strategic gate placement—is superior to one based purely on directional solidification. This approach better manages the inherent poor feeding characteristics of titanium.
- Simulation is critical for identifying and eliminating unnecessary geometric features that introduce thermal mass. Features like stabilizing ribs, often added empirically to prevent distortion, can become primary sites for shrinkage and stress concentration. The investment casting process should aim for geometric simplicity, using simulation to predict and control distortion instead.
- Direct feeding is paramount for isolated heavy sections. In titanium investment casting, the feeding distance is short. The most reliable method to eliminate shrinkage in a heavy boss or junction is to attach an ingate directly to it, effectively turning the gate into a feeding riser, even if it increases clean-up cost.
- A holistic simulation workflow encompassing filling, solidification, and stress analysis is essential. It provides a complete picture of defect formation risks and post-casting behavior, enabling the design of an investment casting process that achieves both internal soundness and dimensional stability.
The successful production of this complex rod casting demonstrates that a rigorous, simulation-led approach to the investment casting process can overcome the traditional limitations of titanium casting, enabling the reliable and cost-effective manufacture of high-integrity components for the most demanding applications.