In my experience as an engineering professional involved in the manufacturing of complex turbine components, I have encountered numerous challenges related to structural intricacies, particularly when transitioning from traditional methods like stamping and welding to advanced investment casting processes. The turbine in question exhibits a significant disparity in wall thickness between its shell and blades, with the shell being substantially thicker and the blades remarkably thin at approximately 1 mm. Furthermore, these 21 curved blades are integrally connected to a semicircular ring, also with a thin wall. The overall dimensions include an outer shell diameter of 270 mm, a ring outer diameter of 230 mm, and a height of 62 mm, with a final casting weight of 2.3 kg. Initially, a stamped and welded assembly was attempted, but this approach proved labor-intensive, time-consuming, inefficient, and ultimately failed to meet the design performance and operational parameter requirements. Consequently, the decision was made to adopt the investment casting process, a method renowned for producing high-integrity, complex geometries. This shift necessitated an intense focus on detailed process control across every stage to ensure success, especially under non-vacuum melting and pouring conditions.
The core difficulty in this investment casting process lay in replicating the intricate internal cavity featuring the numerous thin, curved blades fused to the semicircular ring. My strategy involved decomposing the process into meticulously controlled steps, each requiring specific parameters and adjustments. The following sections detail my firsthand account of implementing and refining these steps to achieve a viable production methodology.
The first critical step was the design of the injection molds for the wax patterns. Given the undercuts and complex geometry of the internal blades and ring, creating a single wax pattern via a one-piece mold was impossible due to insurmountable ejection problems. Therefore, I designed two separate molds. One mold was dedicated to forming the wax pattern of the main shell structure, while a second, more intricate mold was created to produce the wax assembly of the 21 blades along with their connecting semicircular ring. This separation was fundamental to the investment casting process, allowing for the creation of accurate patterns that could later be assembled.
With the mold design settled, attention turned to the wax material itself. Consistent, high-quality wax is paramount in the investment casting process to prevent defects in the final ceramic shell and, consequently, the metal casting. I instituted a rigorous three-stage filtration system for the low-temperature wax material reclaimed after the dewaxing stage. The wax liquid, after being melted out in a hot water bath, undergoes its first filtration as it flows from the dewaxing tank into the processing tank. Following chemical treatment (typically involving acid addition for purification), it passes through a second filter before entering a settling tank for impurity precipitation. Finally, a third filtration occurs just before the wax is poured into molds to form large ingots. To further enhance wax uniformity, I introduced a wax planning step. Instead of using the ingots directly, they are mounted on a planning machine that shaves them into thin, consistent flakes. These flakes are then melted and stirred to create a wax paste that is exceptionally homogeneous and free from particulate matter, which is crucial for injecting the fine details of the blade patterns.
The control of environmental conditions during pattern making is another subtle but critical factor. I maintained the pattern injection room at a temperature no higher than 25°C (77°F). Upon ejection from their respective molds, the blade-ring wax assembly and the shell wax pattern were not quenched in water—a common practice that can induce stress or distortion. Instead, they were placed on a flat surface and allowed to cool naturally for a minimum of two hours. This slow, uniform cooling minimizes warpage. After this period, the patterns were inspected and trimmed, and within three hours of ejection, they were assembled.
The assembly method for the wax patterns represented a significant departure from traditional soldering with a hot iron. I opted for a specialized adhesive wax. The key to success was precise temperature control. When heated to the typical 60°C (140°F), the adhesive wax was too viscous, leading to uneven application and unsightly “wax buildup” at the joints. Through experimentation, I found that heating the adhesive wax to 70°C (158°F) provided optimal fluidity. The semicircular ring pattern was briefly dipped into the molten adhesive for no more than two seconds. It was then withdrawn, and the adhesive layer was brushed gently with a soft brush to ensure even distribution. A crucial pause of 5 to 7 seconds followed, allowing the wax to slightly “set” and become tacky. Finally, the ring pattern was carefully and steadily positioned inside the shell pattern, creating a complete, bonded wax replica of the final turbine. This method ensured a clean, strong joint without thermal damage to the delicate blades.
The design of the gating and feeding system is arguably the heart of any successful investment casting process, especially for a thin-walled component like this turbine. Its primary functions are to enable complete mold filling, feed solidification shrinkage to prevent porosity, and allow for the escape of air and burnt wax residues. I explored and tested three distinct design philosophies for this turbine’s investment casting process.
The first design employed a large spherical feeder (riser) attached via three substantial feed gates. This was combined with centrifugal pouring. The theory was that the thick gates would aid in dewaxing, venting, and providing feed metal to the shell body, while centrifugal force would enhance mold filling. The feeder’s size was determined based on the principle of the feeding zone radius, ensuring it could effectively feed the relevant section of the casting.
The second design retained the spherical feeder but used a four-arm internal gate system (a “four-fork” ingate) under centrifugal pouring. This aimed to distribute the metal more evenly into the turbine cavity from multiple points.
The third, and ultimately most successful, design featured an integrated, annular pouring system. A continuous ring-shaped gate was attached to the top periphery of the turbine shell pattern. Above this ring, five vertical channels conveyed metal from a circular runner, which itself was fed by a central pouring cup. This entire assembly was designed for centrifugal casting. The integrated ring gate provided uniform metal entry around the entire circumference of the thin-shelled turbine, dramatically improving fill integrity compared to the previous designs. The superiority of this integrated gating system for fill completeness was conclusively proven during trials.
To quantify the centrifugal casting parameters, I relied on fundamental formulas. The centrifugal force ($F_c$) and the gravity multiplier ($G$) are critical for determining the required rotational speed. The formulas are:
$$F_c = m \omega^2 r$$
$$G = \frac{\omega^2 r}{g}$$
Where:
– $m$ is the mass of the metal,
– $\omega$ is the angular velocity (in radians per second),
– $r$ is the effective radius of rotation (from the axis to the casting’s center of mass),
– $g$ is the acceleration due to gravity (9.81 m/s²).
For practical purposes, we control the rotational speed ($N$) in revolutions per minute (RPM), related to $\omega$ by $\omega = \frac{2\pi N}{60}$. Based on calculations and empirical adjustment for this specific turbine geometry and metal fluidity, the optimal rotational speed was set at 293 RPM. The pouring time was strictly controlled between 5 and 8 seconds, with rotation ceasing immediately as the metal level approached the neck of the feeder to prevent over-spinning and turbulence.
The ceramic shell building process, a cornerstone of the investment casting process, required meticulous layering. My initial trial process involved seven layers with specific parameters for slurry viscosity, stucco sand grain size, and drying conditions. The following table summarizes the initial trial shell-building schedule:
| Layer | Slurry Type & Dip Time | Stucco Sand (Grit Size) | Drying Time & Conditions | Additional Steps |
|---|---|---|---|---|
| 1 (Prime) | Mullite slurry, 35 sec | 80-100 mesh | 10 h @ 23°C, 65% RH | None |
| 2 | Mullite slurry, 22 sec | 60-80 mesh | 12 h @ 23°C, 65% RH | None |
| 3 | Mullite slurry, 15 sec | 60-80 mesh | 12 h @ 23°C, 50% RH | Air blowing |
| 4 | Mullite slurry, 14 sec | 30-60 mesh | 12 h @ 23°C, 50% RH | Air blowing |
| 5 | Mullite slurry, 14 sec | 16-30 mesh | 12 h @ 23°C, 50% RH | Air blowing |
| 6 | Mullite slurry, 14 sec | 16-30 mesh | 12 h @ 23°C, 50% RH | Air blowing |
| 7 (Seal) | Mullite slurry, 14 sec | None | 16 h @ 23°C, 50% RH | Air blowing |
While this produced a strong shell, post-casting shell removal from the intricate blade cavity was extremely difficult. To mitigate this, I modified the process before the third coating. After the second layer, I manually “invested” the internal blade cavity by pouring and packing fine 60-80 mesh mullite sand into it. I then sealed the openings with a thick paste made from mullite slurry and powder. This created a friable core within the cavity. After this, the standard coating sequence (layers 3 through 7) proceeded. After firing and casting, this internal investment could be easily broken out, greatly simplifying final shell removal and cleaning without damaging the delicate blades. This adaptation significantly improved the practicality of the investment casting process for this component.
Dewaxing was performed using a standard steam autoclave or hot water method. Immediately after dewaxing, the shells were taken out and rinsed thoroughly with boiling water at least twice. This step is vital to remove any residual wax or loose particles from the intricate internal surfaces, which could otherwise cause gas defects or inclusions during pouring.
The thermal treatment of the ceramic shell, or firing, was conducted in two distinct stages—a refinement I found essential for shell strength and cleanliness. First, a pre-firing stage was conducted at 950°C (1742°F). This burns out any last traces of wax and begins the sintering process of the ceramic. After cooling, these pre-fired shells underwent a second internal wash with water to dislodge any ash residues from the complex internal passages. Second, the shells were placed inside cylindrical steel containers for a packing fire. The shells were surrounded by coarse sand within the container, and the top surface of the sand was brushed with a thin layer of silica sol binder. This packing provides mechanical support during handling and firing, minimizes shell distortion, and creates a micro-environment that promotes uniform heating. The final high-temperature firing was then conducted, with the furnace temperature set between 1100°C and 1150°C (2012°F – 2102°F). The shells were held at this temperature for a minimum of 30 minutes to ensure complete sintering and thermal stability before receiving the molten metal. The high firing temperature is crucial for achieving adequate shell strength to withstand the metallostatic and centrifugal pressures during the investment casting process.
The melting and pouring operations, though conducted under non-vacuum conditions, demanded strict discipline. The material was ZG310-570 (a cast carbon steel). Melting was carried out in a medium-frequency induction furnace. After the charge was fully molten, the temperature was raised to 1560-1570°C (2840-2858°F). At this point, preheated ferro-manganese (0.20% by weight) and ferro-silicon (0.10%) were added as pre-deoxidizers. The slag was removed, and a covering flux was applied. Following further slag removal, a final deoxidation was performed using 0.03% pure aluminum. The metal was then allowed to calm for a brief period before final slag skimming. The target tap temperature was 1610-1620°C (2930-2948°F). To ensure precise control over the small amount of metal needed per shell (approximately matching the 2.3 kg casting weight plus gating), I used specially made 10 kg capacity teapot-style ladles. These ladles, with their pour spout, allowed for a clean, controlled stream. Crucially, both the ladles and the shells (in their packing boxes) were preheated in the same furnace to avoid thermal shock. One full ladle was used to pour one shell cluster. Any leftover metal in the ladle after pouring was immediately returned to the induction furnace to prevent temperature drop and contamination for the next heat. Coordination between the furnace operator, the ladle handler, and the centrifugal machine operator was emphasized as a key success factor in this phase of the investment casting process.
The solidification and cooling phase is where internal stresses are generated. To prevent hot tearing or distortion in the thin blades, a controlled slow cool was imperative. The packed shells, containing the freshly poured castings, were not broken open immediately. They were left in their insulated steel containers to cool slowly within the residual sand packing. The opening time was strictly regulated based on ambient temperature. For example, at a room temperature of 30°C (86°F), the casting was not removed from the container until at least 3 hours after pouring. In colder conditions, this time was extended further. Quenching or rapid cooling with water was strictly forbidden.
Finally, the cleaning and finishing operations were conducted with utmost care to preserve the integrity of the fragile blades. The ceramic shell was removed manually and with gentle pneumatic tools where necessary. The gating system was cut off using abrasive wheels or saws, taking care not to snag or impact the blades. The castings were then shot blasted using fine steel media with a particle size not exceeding 0.3 mm to clean the surface without eroding the thin sections. Every handling step emphasized gentle placement and organized stacking to prevent mechanical damage.
In reflection, the successful production of this demanding turbine component was not the result of any single revolutionary technique, but rather the cumulative effect of rigorous, detailed control across every single step of the investment casting process. From wax formulation and pattern assembly through shell building, thermal processing, controlled melting, precision centrifugal pouring, and gentle post-casting handling, each phase presented opportunities for improvement and refinement. The key lesson was that for complex, thin-walled investment castings, especially those produced under non-vacuum conditions, process robustness is built on a foundation of countless small, well-defined, and consistently executed details. The investment casting process, when subjected to such comprehensive scrutiny and control, proved fully capable of meeting stringent design and performance requirements that alternative manufacturing methods could not satisfy. The evolution of the gating system, in particular, highlights how empirical testing within a well-controlled framework leads to optimized solutions for filling integrity. This entire experience underscores the fact that in precision manufacturing, mastery lies in the meticulous orchestration of the entire investment casting process sequence.