In this study, we explore the casting process for large-scale ductile iron castings used in sand suction pumps, which are critical for dredging and land reclamation projects. These pumps operate in harsh environments with high sediment concentrations, requiring robust components free from defects like shrinkage and porosity. The pump body, weighing approximately 46,600 kg and measuring 5,800 mm × 3,480 mm × 2,200 mm, presents significant challenges due to its irregular shape, thick sections up to 355 mm, and stringent quality requirements. The material specified is QT500-7 ductile iron, which demands precise control over metallurgical processes to achieve desired mechanical properties and microstructure. Our approach integrates expanded polystyrene (EPS) mold casting, also known as lost foam casting, with advanced gating and feeding systems to produce high-integrity ductile iron castings. Throughout this article, we emphasize the methodologies applied to optimize the production of ductile iron castings, ensuring reliability in demanding applications.
The EPS mold technique was selected for its ability to replicate complex geometries without the need for draft angles or complex parting lines. This method involves creating a foam pattern that mirrors the final ductile iron casting, incorporating machining allowances and shrinkage factors. The pattern was segmented into modules for precision fabrication using CNC machining and wire cutting, followed by assembly to ensure dimensional accuracy. However, challenges such as poor surface finish due to low sand compaction during molding were addressed by applying two layers of graphite coating, each 0.2 mm thick, to enhance surface quality and prevent sand inclusion. The gating system was designed as a two-tier bottom-up stepped configuration to facilitate controlled filling and minimize turbulence. Key calculations for metal head pressure and pouring time were derived using established formulas. The metal head pressure \( H_p \) is given by:
$$ H_p = H_{\text{box}} + H_{\text{cup}} – \frac{C}{2} $$
where \( H_{\text{box}} \) is the mold box height, \( H_{\text{cup}} \) is the pouring cup height, and \( C \) is the casting height. The pouring time \( t \) was calculated based on the total poured weight \( G_L \):
$$ t = \sqrt{G_L} $$
For the choke area \( A_{\text{choke}} \), we used:
$$ A_{\text{choke}} = \frac{G_L}{\mu \rho \sqrt{2 g H_p}} $$
with \( \mu = 0.4 \) (flow coefficient), \( \rho = 7,300 \, \text{kg/m}^3 \) (density of molten iron), and \( g = 9.8 \, \text{m/s}^2 \). The gating ratios were set as \( A_{\text{runner}} = 1.2 \times A_{\text{sprue}} \) and \( A_{\text{ingate}} = 0.8 \times A_{\text{sprue}} \), optimized for three-ladle pouring to ensure uniform filling. To address solidification issues in thick sections, external and internal chills were strategically placed. For instance, external chills were applied around the flange areas to prevent shrinkage porosity in bolt holes, while formed internal chills were used in regions exceeding 300 mm thickness to reduce solidification time below 50 minutes, thereby maintaining spheroidal graphite formation. The feeding system included six Ø180 mm insulated sleeve risers and two Ø140 mm conventional risers to compensate for shrinkage and facilitate slag trapping and venting. Vent channels were incorporated at high points to expel gases during pouring, ensuring defect-free ductile iron castings.
In the molding process, we employed furan resin sand with a resin content of 1.2%. The mold was built in four sections to allow inspection of core fits during assembly. Cores were produced using six core boxes, reinforced with welded or cast chaplets for stability. The largest core was specially designed to consolidate two intermediate cores into one, simplifying assembly. Core washes based on zircon flour were applied at a Baume degree of 65–70 to improve surface finish. To counteract mold wall movement, bottom and intermediate parting surfaces were reinforced with wooden and welded barriers, backed by sodium silicate sand and resin sand. Additionally, steel plates were welded along the width sides to prevent mold expansion. This meticulous control in molding is crucial for achieving dimensional accuracy and surface quality in ductile iron castings.
The spheroidization and inoculation treatments are pivotal in determining the performance of ductile iron castings. We utilized a DY-7F heavy rare earth spheroidizer at 1.05% addition, combined with CALBALLOY inoculant at 0.4% and YFY-1A efficient inoculant at 0.15%. A multi-stage inoculation process was adopted to enhance nodule count and prevent fading. Pre-treatment with 0.4% preconditioner was applied before tapping to increase nucleation sites, improving fluidity and internal quality. During tapping, 0.4% inoculant was added by projection, followed by stream inoculation with 0.15% inoculant during pouring. The base iron composition was controlled to 3.4–3.5% C, 1.5–1.6% Si, 0.35–0.45% Mn, P ≤ 0.04%, and S ≤ 0.03%, with post-treatment targets of 3.2–3.4% C, 2.3–2.5% Si, 0.35–0.45% Mn, P ≤ 0.04%, S ≤ 0.02%, 0.65–0.75% Cu, and 0.035–0.055% residual Mg. These parameters ensure high spheroidization rates and mechanical properties in ductile iron castings.
Melting and pouring were conducted using three ladles to manage the large volume, with a pouring temperature range of 1,330–1,350 °C. Fast, continuous pouring was maintained to avoid cold shuts and inclusions. Prior to pouring, four 10-ton weights were placed on the mold, supported by steel tubes, to counter buoyancy forces. The chemical composition was closely monitored, as summarized in the table below, which compares target and actual ranges for key elements in ductile iron castings.
| Element | Target Range (%) | Actual Range (%) |
|---|---|---|
| Carbon (C) | 3.2–3.4 | 3.3–3.4 |
| Silicon (Si) | 2.3–2.5 | 2.4–2.5 |
| Manganese (Mn) | 0.35–0.45 | 0.38–0.42 |
| Phosphorus (P) | ≤ 0.04 | 0.03–0.04 |
| Sulfur (S) | ≤ 0.02 | 0.01–0.02 |
| Copper (Cu) | 0.65–0.75 | 0.68–0.72 |
| Residual Mg | 0.035–0.055 | 0.04–0.05 |
After pouring, the casting was allowed to cool slowly in the mold for 240 hours to promote stress relief and simulate an annealing effect, leveraging the insulating properties of resin sand. The shakeout temperature was kept below 300 °C, with prior loosening of the mold over two days to prevent thermal cracking. The resulting ductile iron castings were then separated from the mold, and the riser systems were removed. The produced pump body casting, assembled with its cover, is shown below, demonstrating the successful application of our process for ductile iron castings.
Mechanical properties and microstructural analysis were performed on attached test lugs to validate the quality of the ductile iron castings. The table below summarizes the results, which meet the requirements of QT500-7, with tensile strength exceeding 420 MPa, elongation above 5%, and spheroidization rates over 80%. The microstructure, examined at 100× magnification, revealed fine graphite nodules and a pearlitic matrix, confirming effective spheroidization and inoculation in these ductile iron castings.
| Sample | Tensile Strength (MPa) | Yield Strength (MPa) | Elongation (%) | Hardness (HB) | Graphite Nodule Size (mm) | Spheroidization Rate (%) | Pearlite Content (%) |
|---|---|---|---|---|---|---|---|
| 1 | 460 | 325 | 11.0 | 175 | 4.92 | 92.06 | 54.91 |
| 2 | 490 | 330 | 7.0 | 177 | 4.45 | 92.00 | 60.71 |
| 3 | 485 | 330 | 7.5 | 179 | 5.21 | 90.54 | 57.89 |
The microstructural integrity of the ductile iron castings was further assessed using quantitative metrics. The nodule count and matrix composition are critical for performance in abrasive environments. We calculated the solidification modulus \( M \) for critical sections to optimize chill design, where \( M = \frac{V}{A} \), with \( V \) as volume and \( A \) as cooling surface area. For sections with \( M < 9 \, \text{cm} \), external chills were sufficient, while thicker regions required internal chills to maintain a spheroidization rate above 80%. The effectiveness of the feeding system was evaluated using the feeding efficiency \( \eta \), defined as:
$$ \eta = \frac{V_{\text{riser}}}{V_{\text{shrinkage}}} \times 100\% $$
where \( V_{\text{riser}} \) is the riser volume and \( V_{\text{shrinkage}} \) is the estimated shrinkage volume. Our design achieved \( \eta > 90\% \), ensuring sound ductile iron castings without shrinkage defects. Additionally, the cooling curve analysis during solidification confirmed that the temperature gradient supported directional solidification, minimizing mid-section porosity in these ductile iron castings.
In conclusion, our comprehensive process design for the sand suction pump body demonstrates the feasibility of producing large, high-quality ductile iron castings using EPS mold technology. By integrating advanced gating, chilling, and feeding systems, along with controlled spheroidization and inoculation, we achieved ductile iron castings with superior mechanical properties and microstructural consistency. This approach not only addresses common defects like shrinkage and porosity but also provides a reference for similar heavy-section ductile iron castings in industrial applications. Future work could focus on optimizing the EPS mold coating and automating the pouring process to further enhance the efficiency and reliability of producing ductile iron castings.