The production of complex, thin-walled housings such as gearboxes presents a significant manufacturing challenge. Traditional methods like fabricated welding and assembly are often complex, costly, and extend production cycles. Sand casting, while viable, typically results in lower dimensional accuracy and poorer surface finish. To overcome these limitations, we have successfully adopted and refined the investment casting process for two distinct types of drill rig gearbox housings, achieving reliable batch production. This article details our first-hand technical methodology, from pattern assembly to process design, featuring practical calculations and process parameters.
The primary challenge in investment casting such components lies in their large internal cavities with relatively small external openings. This geometry complicates both pattern making and core design. We will analyze the formation of these box-type parts by applying theories of die design, wax pattern manufacture, and casting solidification. Our approach introduces a rational precision casting process, along with derived formulas for calculating ingate dimensions, risers, and external chills, all validated through practical application.
1. Problem Statement and Component Analysis
The successful application of investment casting hinges on a thorough understanding of the component’s geometry. The two housings under discussion, designated Type A and Type B, share common characteristics but differ in scale and specific challenges.
- Type A Gearbox Housing: Material is ZG270-500. The part weighs approximately 18.5 kg with overall dimensions of 268 mm x 180 mm x 126 mm. Wall thickness varies significantly, from a minimum of 12 mm to a maximum of 45-50 mm at boss locations.
- Type B Gearbox Housing: Also made from ZG270-500, this larger housing weighs about 11.5 kg. Its external dimensions are 180 mm x 163 mm x 116 mm, with wall thicknesses ranging from 10 mm to 40 mm.
The core difficulty is creating the large internal cavity. For investment casting, this internal form must be defined by a soluble wax pattern. The small access points make it impossible to create the pattern as a single piece using conventional injection methods. Therefore, innovative pattern-making strategies are essential.
2. Wax Pattern Manufacturing Strategies
Given the structural constraints, we employed different wax pattern fabrication techniques for the two housings, moving beyond standard single-injection practices in investment casting.
2.1. Type A Housing: Split-Pattern Assembly
For the Type A housing, the internal geometry necessitated creating the wax pattern in two halves. These halves were then assembled into a complete pattern. A simple wax weld or acetone bonding was impractical due to the large joining area and the need for dimensional stability during handling.
Our solution was to use metal alignment pins. We fabricated pins from 1.5 mm diameter iron wire, shaped as shown in the sketch below, with a sharp end and a flattened head.
The assembly procedure was as follows:
- The two wax pattern halves were aligned on a fixed core rod.
- Approximately 6 to 8 pre-heated alignment pins (to about 60°C) were evenly pressed into the joint seam using pliers.
- The seam and the areas where the pins penetrated were smoothed over using a hot wax knife or a brush dipped in wax liquid.
This method ensured accurate dimensional alignment and rapid assembly. Crucially, these iron pins are typically flushed out during the dewaxing process. Any remnants that do not wash out will melt upon contact with the molten steel during pouring, posing no risk to the final casting quality.
2.2. Type B Housing: Combination of Movable Cores and Loose Pieces
The Type B housing’s design allowed for a slightly larger access opening, enabling a different approach. We utilized a combination of a withdrawable core and loose pieces to form the internal cavity. The main challenge here was stabilizing the small, heavy loose pieces within the mold to prevent movement and ensure dimensional accuracy.
Our innovative fix involved machining four small process holes on the lower sides of the loose pieces. We then passed 1.2 mm diameter steel rods through the outer walls of the die, directly into these holes in the loose pieces. This effectively transferred the weight and secured the loose pieces directly to the robust die walls, not just to other fragile core sections. This provided excellent stability, easy mold disassembly, and high pattern accuracy. After wax injection, the holes were plugged with wax rods of the identical diameter and sealed with liquid wax.
3. Process Design: Shrinkage, Feeding, and Chilling
The process design for investment casting of large, complex housings cannot rely on standard rules. Shrinkage allowances must be determined judiciously, and feeding systems must account for dispersed hot spots.
3.1. Determination of Shrinkage Allowance
Shrinkage is not uniform across the entire housing due to varying constraints from the ceramic shell and internal geometry. We determined allowances on a feature-by-feature basis:
- Main Body & Internal Cavity: Subject to restraint from shell expansion and metal contraction, a shrinkage allowance of 1.8% to 2.0% was applied.
- Bosses and Flanges on Flat Surfaces: Less constrained, these features used a standard allowance of 1.5% to 1.6%.
- Long Mounting Feet (Type A): These features experience complex stresses during solidification, tending to contract inwards while being pulled outwards. An intermediate allowance of 1.6% was specified for the distance between the feet.
3.2. Gating, Riser, and Chill Design
Feeding is complicated by the presence of multiple isolated heavy sections (bearing bosses, mounting feet). Using risers for every hot spot is impractical. We adopted a combined strategy of top gating through risers, controlled solidification using external chills, and leveraging sequential solidification for self-feeding between adjacent sections.
The system is designed as a top-pouring arrangement where molten metal enters directly through the risers, which also act as the main feeders. Formulas derived from our production experience are as follows:
(1) Ingate and Riser Calculation
The dimensions are based on the hot spot or “hot circle” diameter ($D_{hot}$).
Ingate cross-sectional area ($A_{ingate}$):
$$A_{ingate} = (1.3 \text{ to } 1.4) \times \frac{\pi D_{hot}^2}{4}$$
Ingate height ($H_{ingate}$): 12 to 14 mm.
Ingate width ($W_{ingate}$): $W_{ingate} = (0.6 \text{ to } 1.0) \times W_c$, where $W_c$ is the width of the casting at the ingate junction.
Ingate length ($L_{ingate}$): $L_{ingate} = 6 \text{ to } 8$ mm.
Riser cross-sectional area ($A_{riser}$):
$$A_{riser} = (1.0 \text{ to } 1.2) \times A_{ingate}$$
Riser width ($W_{riser}$): $W_{riser} = A_{riser} / L_{riser}$ (where $L_{riser}$ is the riser length, often set along the casting edge).
Riser height ($H_{riser}$):
$$H_{riser} = (1.5 \text{ to } 1.8) \times (W_{riser} + L_{riser}) / 2$$
(2) External Chill Calculation
External chills are used to accelerate solidification at specific heavy sections, promoting directional solidification towards the riser or creating a simultaneous solidification condition with a thinner adjacent wall.
- Chill Mass ($M_{chill}$): The required mass of the chill is related to the volume of the casting section it is cooling.
$$M_{chill} = k \times \rho_{steel} \times V_c \times (M_c / (M_c + M_a))$$
Where:
$k$ is an empirical factor (typically 0.5-0.7),
$\rho_{steel}$ is the density of steel,
$V_c$ is the volume of the casting section being chilled,
$M_c$ is the modulus of the chilled section,
$M_a$ is the modulus of the adjacent casting section. - Chill Contact Area ($A_{chill}$): The area of the chill in contact with the casting should be sufficient to extract heat effectively.
$$A_{chill} = (1.2 \text{ to } 1.5) \times \frac{\pi D_{hot}^2}{4}$$
Where $D_{hot}$ is the diameter of the hot spot.
(3) Casting Self-Feeding Principle
Due to the large surface area of the housing, molten metal experiences significant temperature drop as it flows from the top gates to the lower sections. When it reaches the lower thick sections, its temperature is near the liquidus, minimizing initial solidification contraction. Subsequently, hotter metal from above arrives, creating a thermal gradient. This can lead to a near-simultaneous solidification at the junction between thick and thin walls, effectively allowing the thicker section to be “fed” by the faster-solidifying thin wall, mitigating shrinkage porosity.
4. Shell Building Process
The shell-building stage is critical for large, flat surfaces prone to distortion or shell cracking. We implemented four key process controls.
4.1. Structural Support with Sand Backups
The assembled wax pattern tree, including heavy housings, is too massive to be supported solely by the gating system. We place a specially made sand backup (or “sling”) under the housing cavity. This sling is made from sodium silicate-bonded sand, contoured to match the housing’s bottom surface and coated with ceramic slurry. A 6 mm steel rod is embedded inside with ends protruding 12-20 mm for wiring. The entire assembly is wired securely through the gating wax, with the weight ultimately suspended from a reinforced lifting ring at the top of the cluster.
4.2. Slurry Formulation and Viscosity Control
Large flat areas are susceptible to slurry drain, leading to sand inclusion and poor surface finish. We use a primary slurry with a higher powder-to-binder ratio. To prevent sand penetration defects (veining), the first layer employs a coarser stucco sand.
Primary Slurry Parameters:
- Powder-to-Liquid Ratio: 1.1 : 1 to 1.2 : 1
- Binder: Sodium Silicate (M ≈ 3.2-3.3)
- Viscosity: Controlled seasonally.
- Spring/Autumn: 30-35 seconds (Ford Cup #4).
- Summer: 35-40 seconds.
- Winter: 40-45 seconds.
- First Layer Stucco: 50/70 mesh refined silica sand.
4.3. Reinforcement with Mesh and Wires
To prevent shell distortion on large external planes, a welded steel mesh frame is embedded into the shell during the application of the 4th or 5th slurry layer. For internal large flat surfaces, 4-gauge (approx. 5 mm) iron wires are fixed across the cavity during shell building to provide internal support.
4.4. Detailed Process Parameters
The following tables summarize our standardized shell-building process for these housings, a result of extensive optimization in our investment casting practice.
| Slurry Layer | Binder (Sodium Silicate, M=3.2-3.3) | Refined Silica Flour (part) | Chamotte Powder (part) | Viscosity (Ford Cup #4, sec) | Additives (Wt.%) | Remarks |
|---|---|---|---|---|---|---|
| Primary (1st & 2nd) | 1.0 | 1.1 – 1.2 | – | 30-35 | Penetrant JFC: 0.3-0.5 n-Octanol: 0.1-0.3 |
High powder content for surface finish. |
| Back-up (3rd onward) | 1.0 | – | 0.8 – 1.0 | 35-40 | – | Viscosity of 3rd layer ~35 sec. Add flour to thicken from 4th layer. |
| Slurry Layer | Stucco Sand Mesh | Drying Time (min) | Hardening Agent (NH4Cl, 20-30%) | Hardening Time (min) | Post-Hardening Drying (min) |
|---|---|---|---|---|---|
| 1st (Primary) | 50/70 | 10-20 | Immersion | 3-5 | 45-60 |
| 2nd (Transition) | 30/50 | 10-20 | Immersion | 5-7 | 45-60 |
| 3rd (Back-up) | 20/40 | 20-30 | Spray | 10-12 | 30-45 |
| 4th – 6th (Back-up) | 16/30 | 30-40 | Spray | 15-20 | 30-45 |
5. Shell Firing and Casting
The final steps are tailored to the designed feeding and chilling strategy to ensure sound castings.
5.1. Shell Firing and Support
Prior to firing, internal ceramic supports (made from the same backup sand) are placed inside the shell cavity beneath the ingates/risers. These are fixed with sodium silicate to the shell bottom. This prevents shell deformation or cracking due to the hydrostatic pressure of the molten metal during pour.
5.2. “Cold Shell” Casting Practice
To maximize the chilling effect and enhance the self-feeding mechanism, we employ a “cold shell” casting technique. The shell is fired without backup investment in a furnace at approximately 900°C for 1 to 2 hours. It is then removed from the furnace and allowed to cool in ambient air to about 100-150°C. The cooled shell is then placed in a flask and firmly packed with dry sand. This practice ensures the shell has high strength and permeability while being at a low enough temperature to rapidly extract heat from the metal.
5.3. Pouring Parameters
Because the shell is cold, the superheat of the molten metal must be increased slightly to ensure proper fluidity to fill the thin sections.
- Metal Pouring Temperature: 1580°C – 1620°C (for carbon steel).
- Metal Pouring Temperature: 1560°C – 1580°C.
- Pouring Speed: 3-5 seconds per mold, aiming for a rapid fill to minimize temperature loss.
6. Results and Conclusion
The implementation of this integrated investment casting process has proven highly successful. The strategic use of chills and the controlled thermal gradient prevent shrinkage defects in the heavy sections, while the adequate pouring pressure and temperature ensure complete filling of thin walls and bosses. The resulting castings exhibit high dimensional accuracy and excellent surface finish, surpassing the capabilities of sand casting for such components.
Key to this success is the holistic view of the process: innovative pattern-making to overcome geometric constraints, differentiated shrinkage allowances, calculated feeding and chilling based on solidification principles, and a shell-building and casting practice designed to support the entire strategy. This approach has enabled us to achieve a consistent yield rate of approximately 85% for these complex gearbox housings, demonstrating the robustness and viability of the tailored investment casting process for large, intricate box-type components.