In the pursuit of advanced aerospace propulsion systems, the demand for high-temperature, lightweight structural materials is paramount. Titanium-aluminide intermetallic alloys, notably TiAl (e.g., Ti-48Al-2Cr-2Nb) and Ti2AlNb (e.g., Ti-22Al-25Nb) alloys, have emerged as critical candidates due to their excellent high-temperature strength, specific strength, and oxidation resistance. These properties make them ideal for components like turbine blades, combustor liners, and casings, enabling significant weight reduction and enhanced performance. Among various manufacturing routes, precision investment casting stands out as a key technology for achieving the near-net-shape fabrication of complex components from these challenging alloys. However, the widespread application is hindered by the inherent characteristics of Ti-Al alloys, such as poor melt fluidity, high reactivity with mold materials, significant solidification shrinkage, and susceptibility to defects like misruns, shrinkage porosity, and hot tearing in complex geometries.
This work presents a systematic investigation into the casting performance of TiAl and Ti2AlNb alloys during precision investment casting. The core objective was to establish fundamental process-structure-property relationships by evaluating the influence of key foundry parameters—specifically shell preheating temperature and component geometry—on critical casting outcomes: fillability and dimensional contraction. We deconstructed a typical complex aerospace component into a set of eight characteristic structural elements (CSEs) representing common features like thin walls, sections with varying thickness, sharp corners, ribs, holes, fillets, curved surfaces, and rings. By studying these elements, we aim to provide actionable data for optimizing the precision investment casting process for Ti-Al alloy components.
The alloys studied are representative of their classes. Ti-48Al-2Cr-2Nb (TiAl) is a nearly lamellar alloy with a melting point of approximately 1480°C, a relatively narrow freezing range, and a density of about 3.9 g/cm³. Ti-22Al-25Nb (Ti2AlNb), based on the orthorhombic phase, has a higher melting point (~1740°C), a wider solidification interval, and a density of 5.2 g/cm³. Their primary attributes are contrasted in the table below.
| Property | Ti-48Al-2Cr-2Nb (TiAl) | Ti-22Al-25Nb (Ti2AlNb) |
|---|---|---|
| Melting Point (°C) | ~1480 | ~1740 |
| Density (g/cm³) | ~3.9 | ~5.2 |
| Room Temperature Strength (MPa) | 400 – 700 | 900 – 1400 |
| Target Service Temperature (°C) | 650 – 750 | 600 – 650 |
| Primary Manufacturing Challenge | Poor room-temperature ductility & machinability | High melting point & reactivity |
The precision investment casting process followed a standard sequence: fabrication of wax patterns via 3D printing, assembly onto a runner system, iterative ceramic shell building using yttria-based face coats and mullite-based backup coats, dewaxing, high-temperature firing, and finally preheating to the target temperature before pouring. A bottom-gated gravity pouring system was employed using a skull furnace. The critical variable under study was the shell preheating temperature, which was set at 400°C, 600°C, and 800°C for separate trials. The geometry variable was addressed by designing the eight CSEs with varying dimensions, such as plates with 2, 4, and 6 mm thickness.
The primary metrics for evaluation were filling ratio and linear shrinkage. The filling ratio $F_r$ for a feature is defined as the ratio of the actual filled length $L_f$ to the intended design length $L_d$ of the wax pattern:
$$F_r = \frac{L_f}{L_d} \times 100\%$$
A ratio of 100% indicates complete filling. Linear shrinkage $S_l$ was calculated in both the length and thickness directions by comparing the average dimension of the cast component $D_c$ to the dimension of the wax pattern $D_w$:
$$S_l = \frac{D_w – D_c}{D_w} \times 100\%$$
Residual stress simulations for ring-shaped CSEs were also conducted using ProCAST software to understand cracking tendencies.
The Critical Role of Shell Preheating Temperature
Shell preheat temperature is a master variable in precision investment casting, profoundly affecting thermal gradients, solidification kinetics, and metal-mold interaction. Our experiments revealed its decisive impact on the fillability of both TiAl and Ti2AlNb alloys.
For the TiAl alloy, increasing the preheat temperature from 400°C to 800°C resulted in a dramatic improvement in filling ratio, particularly for thin sections. For instance, a 2-mm thick plate (Pl2) saw its fill ratio jump from approximately 33.7% at 400°C to 45.6% at 600°C and finally to near-complete filling at 800°C. This trend was consistent across features like ribs (Fo) and variable sections (Va). The superior fluidity at 800°C can be attributed to the reduced thermal gradient between the molten metal and the mold, slowing the initial solidification rate and allowing the viscous TiAl melt to flow further into the cavity before freezing. The table below summarizes the filling performance for selected TiAl CSEs.
| CSE Type (Thickness) | Fill Ratio @ 400°C (%) | Fill Ratio @ 600°C (%) | Fill Ratio @ 800°C (%) |
|---|---|---|---|
| Plate (2 mm) | 33.7 | 45.6 | >98 |
| Plate (4 mm) | 88.3 | 94.6 | >98 |
| Rib (4 mm) | < 50 | >95 | >99 |
| Variable Section | ~85 | ~97 | >99 |
The Ti2AlNb alloy exhibited a similar positive correlation between preheat temperature and fillability, though the absolute fill ratios and sensitivity differed. At 400°C, its fill ratios were generally lower than those of TiAl for equivalent features, likely due to its higher melting point and potentially greater superheat requirement. However, at 600°C, the fillability of Ti2AlNb improved markedly, reaching levels comparable to or even exceeding TiAl for features like variable sections, ribs, and holes. At 800°C, most Ti2AlNb CSEs achieved near-perfect filling (>99%). The relationship can be conceptually modeled by considering the effective fluidity length $L_f$, which is a function of preheat temperature $T_p$, melt superheat $\Delta T$, and a geometric factor $k_g$ related to section thickness:
$$L_f \propto k_g \cdot (\Delta T + \alpha \cdot T_p)$$
where $\alpha$ is a coefficient representing the effectiveness of shell preheat in delaying the solidification front. This simplified model highlights how increased $T_p$ directly extends the flow length.
Geometry and Wall Thickness: Enabling Flow and Feeding
Beyond preheat temperature, the geometry of the feature itself is a fundamental determinant of success in precision investment casting. A universal finding across both alloys was the beneficial effect of increasing wall thickness on fillability. For example, TiAl plates with thicknesses of 2, 4, and 6 mm showed progressively higher fill ratios at a constant preheat temperature. This is because a thicker section has a higher modulus, leading to a slower local cooling rate. The reduced thermal gradient along the thickness direction (smaller $\frac{dT}{dx}$) allows the thermal center to remain liquid longer, facilitating both improved flow to the end of the section and better interdendritic feeding to counteract shrinkage. Features like variable sections, which naturally transition from thick to thin, demonstrated excellent fillability as they promote directional solidification from the thin (fast cooling) end back toward the thick (hot spot) section and the feeder.
The contrast was evident in features like sharp corners (Co) and thin ribs (Fo2), which consistently presented the greatest challenge for complete filling, especially at lower preheat temperatures. These features act as thermal chills, rapidly extracting heat and creating isolated pockets that are difficult for the melt to reach before a solid skin forms.
Dimensional Contraction: Balancing Fill and Accuracy
While high preheat temperature enhances filling, it also influences the solidification contraction behavior and final dimensions. Linear shrinkage was measured in both the longitudinal (flow direction) and transverse (thickness) directions. A key finding was that the optimal preheat temperature for dimensional accuracy did not necessarily coincide with that for best fillability.
For the TiAl alloy, longitudinal shrinkage at 400°C was high and inconsistent, particularly for poorly filled features. At 800°C, shrinkage stabilized but was generally between 2-3%. Interestingly, at 600°C, the shrinkage values were often the lowest and most consistent, typically below 2.5% in the longitudinal direction. Shrinkage in the thickness direction was generally higher, ranging from 2% to over 4%, with 600°C again often yielding the most favorable (lowest) values.
The Ti2AlNb alloy displayed a distinct advantage in terms of lower and more predictable shrinkage. At the optimal process window of 600°C preheat, the longitudinal linear shrinkage for most Ti2AlNb CSEs was below 1.5%, which is notably lower than that of TiAl cast under similar conditions. This can be expressed as:
$$S_{l,\ Ti2AlNb}(600^\circ C) < S_{l,\ TiAl}(600^\circ C)$$
This difference may be related to the alloys’ distinct solidification paths, phase transformation sequences, and resultant thermal contraction coefficients. The more consistent shrinkage of Ti2AlNb simplifies pattern design and compensation in precision investment casting.
| Alloy / Preheat Temp. | Avg. Long. Shrinkage @ 400°C (%) | Avg. Long. Shrinkage @ 600°C (%) | Avg. Long. Shrinkage @ 800°C (%) |
|---|---|---|---|
| TiAl | High & Variable | ~1.8 – 2.5 | ~2.0 – 3.0 |
| Ti2AlNb | High & Variable | < 1.5 | ~1.5 – 2.0 |
Defect Formation: The Cracking Tendency and Process Window
A major defect of concern in precision investment casting of intermetallics is hot tearing or cracking, often initiated by residual stresses during cooling. Simulation and experimental analysis of ring-shaped CSEs provided critical insights. Residual stress simulations for Ti2AlNb rings at 600°C preheat showed that von Mises stress concentration was highest at the connection point between the ring and the feeder gate, especially for thinner rings (2mm and 4mm thick). The stress magnitude $\sigma_{vm}$ decreased significantly with increasing ring thickness $t$:
$$\sigma_{vm} \propto \frac{1}{t^n}$$
where $n$ is a positive exponent. Experimentally, TiAl alloy rings cast at 600°C were prone to cracking at this stress-concentrated gate junction. Ti2AlNb rings, under the same conditions, showed significantly better integrity and were largely crack-free. This highlights TiAl’s greater susceptibility to thermal stress cracking due to its lower ductility and fracture toughness at elevated temperatures compared to Ti2AlNb.
The choice of 800°C preheat, while excellent for fillability, introduces a significant trade-off: increased chemical reactivity between the highly reactive Ti-Al melt and the ceramic shell. This often leads to surface contamination and the formation of subsurface porosity or inclusions, degrading the mechanical properties. Therefore, 600°C emerges as a balanced and more suitable shell preheating temperature for the precision investment casting of these alloys, particularly for Ti2AlNb and for thicker sections of TiAl. This temperature provides sufficient thermal energy to ensure good fillability (especially for sections >4mm), minimizes harmful metal-mold reactions, promotes lower and more predictable shrinkage, and helps manage residual stress levels.
Synthesis and Conclusions
This comprehensive study on the precision investment casting of TiAl and Ti2AlNb alloys leads to several key conclusions that can guide process optimization:
1. Shell preheat temperature is the primary driver for fillability. Increasing preheat temperature from 400°C to 800°C systematically improves the filling ratio of both alloys by reducing the initial cooling rate and extending melt fluidity. The relationship can be qualitatively described by a fluidity length model dependent on preheat temperature.
2. Section thickness is a critical geometric factor. Increased wall thickness consistently enhances fillability by reducing local cooling rates. Features designed to promote directional solidification (like tapered sections) yield superior casting soundness compared to uniform thin walls or isolated thick sections.
3. An optimal process window exists near 600°C preheat. While 800°C ensures the best filling, 600°C offers a superior balance for precision investment casting. At this temperature:
- Fillability is adequate for most features with thicknesses ≥4mm.
- Linear shrinkage is minimized and most predictable, with Ti2AlNb exhibiting lower shrinkage (<1.5%) than TiAl.
- The risk of severe metal-mold reaction defect formation is reduced compared to 800°C.
4. Alloy-specific behaviors dictate defect susceptibility. Ti2AlNb alloys demonstrate advantages in lower shrinkage and better resistance to stress-related cracking under identical casting conditions. TiAl alloys, though castable at lower temperatures, require careful design (avoiding sharp notches, ensuring generous fillets) and controlled cooling to mitigate their propensity for hot tearing, particularly at stress concentration points like gate junctions.
5. Data-driven design is enabled. The quantitative data on fill ratios and shrinkage for specific feature types (plates, ribs, holes, etc.) generated in this study provide a foundational database for the design of wax patterns and process parameters for complex Ti-Al alloy components via precision investment casting.
In summary, successful precision investment casting of titanium aluminide components requires a holistic approach that carefully balances thermal management (via shell preheat), geometrical design (favoring progressive solidification), and alloy selection based on final property requirements. The findings underscore that Ti2AlNb alloys are particularly well-suited for this manufacturing route, offering a compelling combination of castability, dimensional accuracy, and mechanical performance for demanding high-temperature aerospace applications.