In the production of railway gearbox components, ductile iron castings are widely used due to their excellent mechanical properties and cost-effectiveness. This article details the development and optimization of a casting process for a split gearbox used in metro systems, focusing on overcoming challenges such as shrinkage porosity, slag inclusion, and structural inhomogeneity. The gearbox consists of upper and lower halves, with the lower half being more complex and thus serving as the primary focus for process design. Through rigorous simulation and experimental validation, we have identified an efficient method that ensures high-quality ductile iron castings with minimal defects.
The split gearbox castings are characterized by thin walls and varying thicknesses, which pose significant challenges during solidification. The upper half measures 788 mm × 340 mm × 287 mm, while the lower half is 992 mm × 465 mm × 287 mm, with a main wall thickness of 12 mm and axle bore walls of 50 mm. The flange at the joint surface has a thickness of 48 mm, and each casting weighs approximately 121 kg. The material specification is EN-GJS-400-15, which requires high ductility and strength. Non-destructive testing (NDT) standards must be met, including magnetic particle, penetrant, radiographic, and ultrasonic inspections, as per EN 1369, DIN EN 1371-1, EN 12681, and EN 12680-3, respectively. Key areas demand stricter criteria, such as UT0-1 and RT0-3 levels, while other regions allow for slightly higher tolerances. Achieving these standards in ductile iron castings necessitates precise control over the casting process to prevent defects like shrinkage and slag inclusions.
Process analysis begins with understanding the solidification behavior of ductile iron castings. Ductile iron exhibits a mushy solidification mode, leading to a tendency for dispersed microshrinkage if not properly managed. The modulus method is commonly employed to design feeding systems, and we utilized MAGMA software for simulation to predict shrinkage porosity and thermal modulus distribution. The simulation results indicated that the thermal modulus ranged from 0.33 cm to 1.86 cm across the casting, which is insufficient for self-feeding through graphitic expansion alone. Thus, external feeding via risers is essential. The shrinkage distribution simulation revealed critical areas prone to defects, particularly at junctions and thick sections, necessitating a optimized riser design and chilling techniques.
The solidification time for ductile iron castings can be estimated using Chvorinov’s rule: $$ t = B \left( \frac{V}{A} \right)^2 $$ where \( t \) is the solidification time, \( B \) is a constant dependent on the material and mold conditions, \( V \) is the volume of the casting, and \( A \) is the surface area. For ductile iron castings with thin walls, this equation highlights the importance of controlling the cooling rate to avoid shrinkage. Additionally, the feeding distance \( L \) for risers can be approximated by: $$ L = k \cdot M $$ where \( k \) is a constant and \( M \) is the modulus. In our case, the simulations helped optimize these parameters to ensure sequential solidification towards the risers.
We designed three distinct casting process schemes for the lower gearbox half, each with varying pouring positions, gating systems, and riser configurations. The objective was to achieve high yield, reduce sand core usage, and simplify operations while meeting quality standards. Below is a comparative table summarizing the key aspects of each scheme:
| Scheme | Pouring Position | Gating System | Number of Cores | Riser Type | Yield (%) | Sand-to-Iron Ratio (%) | Chill Usage (%) |
|---|---|---|---|---|---|---|---|
| Scheme 1 | Vertical Top Side | Top Gating | 6 | Multiple Open Risers | 53.3 | 6.7 | 6.9 |
| Scheme 2 | Inclined Vertical Side | Middle Gating | 4 | Multiple Side Risers | 71.6 | 8.4 | 26.9 |
| Scheme 3 | Horizontal Bottom Side | Bottom Gating | 2 | Side Heat/Dark Risers | 69.6 | 6.4 | 20.2 |
In Scheme 1, we employed a vertical top-side pouring approach with multiple risers for feeding. The sand cores were 3D-printed in six segments to facilitate placement of chills and cleaning. However, this method involved complex assembly and resulted in a lower yield due to excessive gating and riser mass. The gating ratio was set at 1:2.55:1.59 for sprue:runner:ingate, with a pouring time of 26 seconds. Despite simulation showing acceptable results, practical trials revealed difficulties in core assembly and higher scrap rates, making it less suitable for mass production of ductile iron castings.
Scheme 2 utilized an inclined vertical side pouring position, which aimed to improve feeding efficiency. The cores were 3D-printed and assembled horizontally before being positioned vertically in the mold. This scheme achieved a higher yield of 71.6%, but the sand-to-iron ratio increased to 8.4%, indicating higher sand consumption. Chill usage was significant at 26.9%, which added complexity to the process. The gating ratio was 1:1.94:1.63, with a pouring time of 38 seconds. Although simulation predicted adequate feeding, the operational complexity and potential for misalignment during core assembly posed risks for consistent quality in ductile iron castings.
Scheme 3, which we ultimately adopted, features a horizontal parting line with bottom-side pouring and side heat/dark risers. This approach minimizes turbulence and oxidation of magnesium, a key element in ductile iron that can lead to slag formation. The gating system has a ratio of 1:1.6:1.53, and the pouring time is reduced to 18 seconds, ensuring a smooth fill. The sand cores are 3D-printed in only two pieces, simplifying assembly and reducing costs. Chills are strategically placed at the extremities to direct solidification towards the risers, creating a modulus gradient of \( M_{\text{riser}}:M_{\text{neck}}:M_{\text{casting}}:M_{\text{end}} = (1.35-1.55):1.1:1:(0.5-0.6) \). This promotes sequential solidification and effectively prevents shrinkage defects in ductile iron castings.
The MAGMA simulations for Scheme 3 confirmed a low ingate velocity of 0.6-0.9 m/s, which reduces Mg oxidation and slag inclusion. The shrinkage porosity analysis showed that critical areas met UT and RT standards, with no significant defects. The thermal modulus distribution ensured that risers provided adequate feeding, and the use of chills enhanced directional solidification. This makes Scheme 3 highly suitable for high-volume production of ductile iron castings, as it balances yield, simplicity, and quality.
Melting and treatment processes are crucial for achieving the desired microstructure and mechanical properties in ductile iron castings. We controlled the chemical composition within strict limits: carbon (C) at 3.6-3.7%, silicon (Si) at 2.55-2.65%, manganese (Mn) below 0.2%, phosphorus (P) below 0.03%, sulfur (S) at 0.008-0.012%, and magnesium (Mg) at 0.04-0.05%. The high carbon equivalent ensures good fluidity and graphitization, while low manganese and phosphorus minimize carbide formation and brittleness.
Nodularization was performed using the sandwich method, where 1.1% light rare earth magnesium ferrosilicon alloy and 1.1% barium-silicon inoculant are buried in the pouring ladle, covered with steel scraps to prevent premature reaction. The treatment reaction can be represented by the equation: $$ \text{Mg} + \text{S} \rightarrow \text{MgS} $$ which highlights the desulfurization process critical for forming spheroidal graphite. Post-inoculation with 0.15% sulfur-oxygen inoculant during pouring further enhances graphite nucleation. The pouring temperature was maintained at 1380 ± 10°C to ensure complete filling and reduce cold shuts.
The solidification of ductile iron castings involves complex phase transformations. The graphite spheroid count is a key indicator of quality, and it can be influenced by the cooling rate and inoculation. We aim for a graphite count of 160-320 nodules/mm² to promote a ferritic matrix, as required by EN-GJS-400-15. The base microstructure should primarily consist of ferrite and pearlite, with minimal carbides. The hardness can be estimated using the relationship: $$ \text{HB} = 100 + 0.44 \times \text{Pearlite\%} $$ where HB is the Brinell hardness. For our castings, the target hardness range is 130-210 HB, which corresponds to a pearlite volume fraction of approximately 35%.
Experimental validation of the three schemes was conducted through actual production trials. Scheme 3 demonstrated superior performance, with first-pass qualification and consistent results in NDT. The following table compares the verification outcomes:
| Scheme | Yield (%) | Sand-to-Iron Ratio (%) | Chill Usage (%) | Operational Complexity | NDT Results | Remarks |
|---|---|---|---|---|---|---|
| Scheme 1 | 53.3 | 6.7 | 6.9 | High | Qualified after 3rd batch | Complex core assembly |
| Scheme 2 | 71.6 | 8.4 | 26.9 | Moderate | Not validated | Risks in vertical positioning |
| Scheme 3 | 69.6 | 6.4 | 20.2 | Low | First-pass qualification | Ideal for mass production |
In Scheme 3, the ductile iron castings exhibited excellent mechanical properties. Hardness tests on the joint surface (25 mm thickness) showed values within 130-210 HB, and metallographic analysis revealed a ferritic matrix with graphite nodule counts exceeding 160/mm². The microstructure was uniform, with no evidence of shrinkage or slag inclusions in critical areas. Radiographic and ultrasonic inspections confirmed UT0-1 and RT0-3 levels in key zones, fully complying with technical requirements. This consistency makes Scheme 3 the preferred method for producing high-integrity ductile iron castings for railway applications.
In conclusion, the development of an optimized casting process for metro split gearbox ductile iron castings has been successfully achieved through systematic simulation and experimentation. Scheme 3, with its horizontal parting, bottom-side pouring, and strategic use of risers and chills, offers high yield, simplicity, and reliability. The MAGMA software played a vital role in predicting and mitigating defects, ensuring that the ductile iron castings meet stringent NDT standards. This approach not only enhances production efficiency but also guarantees the long-term performance of ductile iron castings in demanding railway environments. Future work could focus on further optimizing the gating design and exploring advanced inoculation techniques to push the boundaries of ductile iron casting technology.