In my years of foundry engineering practice, I have encountered many challenging components, but few have demanded as much attention as the hydraulic press injection base produced from ductile iron castings. This base is a critical structural element in large oil-hydraulic presses that are used for rubber injection molding and vacuum forming. During operation, the press applies forces ranging from 200 t to 1 000 t, so the base must exhibit high strength, excellent toughness, uniform hardness in the inner bore, good wear resistance, and a dense, leak-free microstructure. Because of the extreme differences in wall thickness and the complexity of the internal geometry, the casting is highly prone to shrinkage cavities, shrinkage porosity, gas holes, and slag inclusions. In this article, I will explain in detail how I designed and implemented a successful casting process for these ductile iron castings, including the pouring position, gating system, risers, chills, venting, melting practice, and defect-prevention measures.
The component is essentially a cylinder-like structure with mounting flanges, guide-pillar bores, side lugs, and an oil-injection port. The gross weight of the casting is 1 380 kg, and the maximum envelope dimensions are 1 200 mm × 880 mm × 780 mm. The nominal wall thickness is 90 mm, but some ribs and guiding sections are only 30 mm thick. The oil-cylinder bore must be machined to a high precision and surface finish, and the internal hydraulic oil circuit demands excellent density. Each casting is subjected to a 25 MPa pressure test for 30 min without any leakage. The specification prohibits shrinkage porosity, gas holes, slag inclusions, and cracks. These requirements are severe for ductile iron castings because the solidification behavior of spheroidal graphite iron involves both liquid contraction and graphitic expansion, which must be carefully balanced to avoid internal defects.
Before presenting the process details, I should emphasize that the chemical composition and mechanical properties are tightly controlled. The final cast iron must have the following composition: carbon from 3.6% to 3.7%, silicon from 2.4% to 2.5%, manganese from 0.4% to 0.5%, phosphorus no more than 0.06%, sulfur no more than 0.02%, residual magnesium from 0.03% to 0.05%, and residual rare earth from 0.01% to 0.02%. The required mechanical properties for the ductile iron castings are: ultimate tensile strength at least 500 MPa, yield strength at least 320 MPa, elongation at least 7%, and Brinell hardness between 170 and 230 HB. The nodularity must be grade 1 to 3, and the graphite size should be class 5 to 7 according to the relevant standards. These specifications are extremely strict, so every step of the process must be optimized.
I selected furan resin sand for both molding and core-making because it provides excellent dimensional accuracy, good collapsibility, and low gas evolution. The molding layout uses a side-bottom gating system with foam ceramic filters, combined with safety risers and overflow risers. The oil-cylinder sand core is a large, heavy core that must be reinforced with a steel pipe core rod. Four side-lug cores, four guide-pillar-bore cores, and four self-penetrating cores (often called “automatic cores” in Chinese foundry terminology) complete the core package. To facilitate pattern withdrawal, ten loose pieces are required for the reinforcing ribs and the oil-injection port. The oil-cylinder top and internal cavity surfaces are equipped with shaped chills to accelerate cooling and promote a dense structure in these thick sections. The oil-injection port pad also receives shaped chills to suppress shrinkage. This combination of chills, risers, and controlled solidification is the cornerstone of producing sound ductile iron castings without expensive external feeders.
Selection of Pouring Position
Two possible pouring positions were considered: oil-cylinder opening upward and oil-cylinder opening downward. I compared both arrangements carefully, and their advantages and disadvantages are summarized in Table 1.
| Pouring position | Advantages | Disadvantages |
|---|---|---|
| Oil-cylinder opening upward | Allows directional solidification; convenient mold closing; chills can be placed on top surfaces. | The metal flow may be less tranquil; risk of oxidation and slag entrapment if the sprue is poorly designed; increased machining allowance on the upper surface. |
| Oil-cylinder opening downward | Metal fills smoothly and quietly; easy to vent; good surface quality on the important bore with less machining allowance; promotes balanced solidification. | Difficult to achieve true directional solidification; slag may accumulate on the top surface (which is the less important face); more complicated mold assembly. |
Table 1 clearly shows a trade-off between the two options. After careful deliberation, I chose the oil-cylinder opening upward. Although the downward position seems better for smooth filling, the upward position allows the use of directly placed safety risers on the heavy cylinder top and is much easier for mold closing. With the help of chills and a properly designed gating system, the upward position can still achieve balanced solidification and produce high-quality ductile iron castings.
Because the casting is relatively tall, a purely top-pouring system would cause severe sand erosion, turbulence, and oxidation of the liquid iron. A bottom-pouring system is preferred for tranquil filling, but it does not naturally promote directional solidification. In my design, I combined bottom-side pouring with multiple risers and chills to satisfy the condition of balanced solidification. The bottom-side gating system reduces the risk of splashing and oxidation, which is especially important for ductile iron castings because magnesium in the melt readily forms oxides and nitrides.
Gating System Design
For ductile iron castings, the gating system must deliver the liquid metal quickly, quietly, and without entrapping slag. I adopted a semi-open (open-closed) gating system, which is a compromise between full-open and full-closed systems. In a semi-open system, the total cross-sectional area of the runners is larger than the sprue area, but the ingate area is smaller than the runner area. This arrangement provides good slag trapping and smooth flow. The area ratios I used are:
$$F_{\text{sprue}} : F_{\text{runner}} : F_{\text{ingate}} = 1 : 1.6 : 1.2$$
The weight of the poured metal, including the risers and gating, was 1,480 kg, giving an excellent process yield of 93.2%. From experience with large ductile iron castings, I selected a pouring time of about 50 s. The effective static pressure head was estimated from the mold geometry, and using the well-established “large orifice” discharge theory, I calculated the required sprue cross-sectional area as:
$$A_{\text{sprue}} = \frac{G}{\mu \rho t \sqrt{2 g H}}$$
where \(G\) is the total pouring weight (1,480 kg), \(\mu\) is the discharge coefficient (about 0.8 for a ceramic sprue with well-designed transitions), \(\rho\) is the density of liquid iron (approximately 7.0 × 10³ kg/m³), \(t\) is the pouring time (50 s), \(g\) is the gravitational acceleration (9.81 m/s²), and \(H\) is the effective height of the sprue (around 0.5 m). Substituting these values gives:
$$A_{\text{sprue}} \approx \frac{1480}{0.8 \times 7000 \times 50 \times \sqrt{2 \times 9.81 \times 0.5}} \approx 0.0050 \ \text{m}^2 = 50 \ \text{cm}^2$$
Therefore, I used a ceramic sprue tube with an inner diameter of 80 mm, which provides a cross-sectional area of 50.3 cm². The runner system was machined in the mold as two horizontal runners: the upper runner had a trapezoidal cross-section of 70 mm top / 85 mm bottom × 50 mm height, and the lower runner had 70 mm / 85 mm × 40 mm. The total ingate area was designed to be 60 cm², achieved with six flat ingates, each 80 mm wide and 12.5 mm high. These wide, thin ingates help to trap slag and solidify quickly after filling, thereby isolating the cavity and allowing the internal graphitic expansion of ductile iron castings to feed the solidifying surfaces. The gating parameters are listed in Table 2.
| Parameter | Value |
|---|---|
| Gating system type | Semi-open |
| Area ratio (sprue:runner:ingate) | 1 : 1.6 : 1.2 |
| Total pouring weight | 1,480 kg |
| Process yield | 93.2% |
| Pouring time | 50 s (range 45–55 s) |
| Effective sprue height | ~0.5 m |
| Sprue tube | φ80 mm ceramic |
| Runner cross-section | Upper 70/85 × 50 mm; Lower 70/85 × 40 mm |
| Number of ingates | 6 |
| Ingate dimensions | 80 mm × 12.5 mm each |
To further clean the liquid metal and reduce inclusions, I placed four foam ceramic filters of size 150 mm × 100 mm × 20 mm and two filters of size 100 mm × 100 mm × 20 mm in the runner system. Foam ceramic filters are essential for ductile iron castings because they remove non-metallic inclusions, dross, and re-oxidation products that form due to the presence of magnesium. The filters also stabilize the flow and reduce turbulence. The selected filter capacity was sufficient for the total weight of 1,480 kg, and the filters were positioned so that the metal would fully cover them before entering the cavity, preventing premature clogging.
Riser and Chill Design
One of the most challenging aspects of producing sound ductile iron castings is controlling solidification to avoid shrinkage porosity. Ductile iron has a unique solidification behavior: the precipitation of graphite during eutectic solidification causes a volumetric expansion that can compensate for the liquid shrinkage, provided that the mold is rigid and the casting has a strong enough wall to resist deformation. This self-feeding behavior allows the use of small or no feeders in many cases. However, to account for unavoidable variations in temperature, chemistry, and molding conditions, I still applied a combination of small safety risers and strategically placed chills.
At the top of the heavy oil-cylinder section, where the wall thickness is 90 mm and the thermal center is large, I placed two safety neck-down risers. Each riser had a diameter of 120 mm and a height of 200 mm. The neck connecting the riser to the casting was 28 mm in diameter and 35 mm in length. The neck-down design ensures that the riser can feed the hot spot until the end of solidification, and then the neck solidifies before the riser, allowing the internal pressure from graphitic expansion to be used for self-feeding. The dimensions of the safety risers are summarized in Table 3.
| Feature | Specification |
|---|---|
| Safety risers (top of oil cylinder) | 2 pieces, φ120 mm × 200 mm |
| Riser neck (shrink neck) | φ28 mm × 35 mm |
| Exhaust riser pads | 4 pieces, top 100 × 25 mm, bottom 80 × 12 mm |
| Overflow riser rings | φ60 mm / φ100 mm × 180 mm |
| Oil-cylinder top and cavity chills | 6 shaped chills, thickness 70 mm |
| Oil-injection pad chills | 2 shaped chills, thickness 80 mm |
| Oil-cylinder bottom chill | 1 cylindrical chill φ180 mm × 80 mm |
On the four support lugs, I placed four flat exhaust riser pads with a top opening of 100 mm × 25 mm and a bottom opening of 80 mm × 12 mm. These pads not only allow gas to escape during pouring but also act as small feeders for those localized hot spots. Additionally, cylindrical overflow riser rings were placed on the side lugs to collect the first, colder metal that may contain slag or dross. This is a common practice for ductile iron castings: the first metal entering the cavity is diverted to overflow risers so that the main cavity is filled with clean, hot metal.
Chills are indispensable for controlling the solidification sequence. I placed six shaped chills of 70 mm thickness on the top surface and inside surface of the oil-cylinder wall. The oil-injection pad received two chills of 80 mm thickness, and the bottom of the oil-cylinder cavity received one cylindrical chill of φ180 mm × 80 mm. The chills increase the cooling rate in these heavy sections, refining the microstructure and shifting the solidification toward the center. This helps prevent shrinkage porosity in the thickest parts of the ductile iron castings. The chills must be clean, dry, and free of rust and oil before being placed in the mold. I also fixed them with nails or hooks to prevent displacement during molding and closing.
To quantify the solidification behavior, I can use the modulus concept. The modulus \(M\) of a casting section is defined as the ratio of its volume \(V\) to its cooling surface area \(A\):
$$M = \frac{V}{A}$$
For the oil-cylinder wall with a thickness of 90 mm and an assumed large plane, the modulus is approximately half the thickness, i.e., \(M \approx 45\) mm. The safety riser with a diameter of 120 mm and a height of 200 mm has a modulus of roughly:
$$M_{\text{riser}} = \frac{V_{\text{riser}}}{A_{\text{riser}}} = \frac{\pi (0.06)^2 (0.2)}{2 \pi (0.06)(0.2) + 2 \pi (0.06)^2} \approx \frac{0.00226}{0.0754 + 0.0226} \approx 0.023 \ \text{m} = 23 \ \text{mm}$$
This is smaller than the casting modulus, which is typical for a neck-down safety riser that is used to feed only the liquid contraction before graphite expansion begins. After the neck freezes, the riser becomes disconnected, and the graphitic expansion inside the casting promotes self-feeding. This approach is widely applied to ductile iron castings to achieve high yield without sacrificing soundness.
Exhaust and Venting System
Gas porosity is a frequent defect in ductile iron castings, especially when large cores are used. The oil-cylinder core is massive and generates a considerable amount of gas when the hot metal comes into contact with the resin-bonded sand. To ensure proper venting, I designed a steel pipe core rod with a diameter of 120 mm for the oil-cylinder core. This pipe serves both as a structural support and as a vent channel. Holes of φ10 mm were drilled along the pipe wall to collect gas from the core body, and the pipe was connected to a φ100 mm vent at the bottom of the mold. When pouring, the bottom of the mold is slightly elevated and the vent is ignited to draw gas out of the core. This ensures that the gas pressure inside the core does not cause blowholes or “choke” defects.
For the four side-lug cores and four guide-pillar bore cores, I provided φ15 mm vent channels that were aligned with φ20 mm vents in the upper mold. These vents release gas from the cores to the atmosphere. The self-penetrating cores (automatic cores) also required careful venting. Before closing the mold, I always dried the cores and the mold cavities, especially around the chills, to minimize moisture and gas evolution. Even though furan resin sand generates less gas than green sand, the total gas volume from such a large mold is considerable. A well-designed venting system is essential for producing sound ductile iron castings without gas-related defects.
Melting and Inoculation Practice
The production of high-quality ductile iron castings begins with the liquid metal. I used an induction furnace and selected high-quality low-sulfur pig iron and clean scrap steel. The scrap steel proportion was increased to 40%–50% of the charge to dilute the trace elements and provide a fully pearlitic matrix after final treatment. For the base iron, the target composition was: carbon 3.75%–3.85%, silicon 1.4%–1.5%, manganese 0.4%–0.5%, phosphorus ≤0.06%, and sulfur ≤0.025%. The carbon equivalent, calculated as:
$$CE = w(\mathrm{C}) + \frac{1}{3} w(\mathrm{Si}) + \frac{1}{3} w(\mathrm{P})$$
was maintained between 4.4% and 4.6%. This level is high enough to promote graphite nucleation and adequate graphitic expansion, but not so high as to cause graphite flotation. If the CE were too low, the contraction would outweigh the expansion, leading to shrinkage porosity in the final ductile iron castings.
For nodulization, I used the sandwich (or tundish) treatment in the ladle. A mixture of a low-rare-earth magnesium nodulizer (70%) and a yttrium-based heavy-rare-earth nodulizer (30%) was placed in a pocket at the bottom of the ladle and covered with steel chips and silicon inoculant. The total nodulizer addition was 1.3%–1.4% by weight of the treated iron. The residual magnesium was kept between 0.03% and 0.05%, and the residual rare earth was kept between 0.01% and 0.02%. This careful balance avoids both under-treatment (poor nodularity) and over-treatment (excessive carbides and shrinkage). The heavy rare earth elements help to counteract the harmful effects of trace elements and improve the fatigue resistance of the ductile iron castings.
Inoculation is critical for achieving high nodule count and eliminating chill. I used a barium-containing silicon inoculant for the ladle addition (0.6%–0.7% of the melt weight) and also performed a secondary inoculation during tapping. Finally, a fine-grained barium-silicon inoculant (0.3–0.8 mm) was added as a stream inoculant during pouring, at a rate of 0.1%. This triple inoculation strategy ensures that graphite nucleates uniformly and that the solidified microstructure is free from brittle carbide phases. The result is a fully ferritic-pearlitic matrix with excellent strength and elongation.
I also applied a graphitization pre-treatment before nodulization. This pre-treatment adds a small amount of a graphitizing agent to the base iron, which provides stable nucleation sites and reduces the tendency for undercooling. This practice has been shown to increase nodule count and improve the machinability and pressure tightness of ductile iron castings. The pouring temperature was strictly controlled at 1,320–1,350°C. Higher temperatures increase liquid contraction and risk dross formation, while lower temperatures cause misruns and cold-shuts. The pouring time was kept between 45 and 55 s, using the earlier calculated 50 s as the target. After pouring, the castings were allowed to cool in the mold for at least 12 hours before shakeout, preventing residual stress and cracking.
Defect Prevention in Ductile Iron Castings
During the development of this process, I encountered and resolved several typical defects. One of the most troublesome was shrinkage at the oil-injection port. This area requires a loose piece in the pattern, and any movement of the loose piece or the chill can cause the hole to become eccentric, leading to machining problems and localized shrinkage. To prevent this, I ensured that the oil-injection-port loose piece was firmly fixed to the pattern and that the shaped chill was tightly attached to the port and nailed to the core. Additionally, the core print for the oil-injection port was rammed solidly so that the core could not shift during mold closing.
Another common defect was gas porosity caused by the chills. Chills that are not properly prepared can generate steam or gas when the molten metal contacts them. I established a strict chill management procedure: all chills are shot-blasted, checked for rust, oil, and moisture, and are limited to ten uses before being reconditioned or scrapped. Before mold closing, the mold and cores are thoroughly dried, particularly around the chill locations. This dramatically reduces the risk of blowholes and “chill-induced” gas defects.
Slag and dross inclusions are a persistent challenge for ductile iron castings because magnesium reacts with oxygen and sulfur to form a viscous slag. To minimize this, I selected high-quality raw materials with low sulfur and control the sulfur content in the base iron to ≤0.025%. The nodulizer addition was carefully weighed, and the reaction time was maintained within 50–120 s to avoid excessive magnesium loss. The melt was skimmed thoroughly before pouring, and the foam ceramic filters in the gating system provided an additional barrier. During mold closing, I carefully cleaned the mold cavity of any loose sand or dust and applied a mold seal to prevent sand erosion. These measures eliminated slag inclusions in the final ductile iron castings.
Another potential defect in thick-walled ductile iron castings is the degeneration of graphite at the surface due to sulfur or oxygen in the mold atmosphere. The use of furan resin sand with good compaction helps to keep the mold rigid and reduces the likelihood of surface pinholes. I also applied a graphite-based mold coating to the cores and mold surfaces that come into contact with the heavy sections. This coating creates a reducing atmosphere at the casting surface, which promotes nodular graphite and prevents the formation of a ferritic rim with flake graphite. The result is a clean, dense surface with excellent machining characteristics.
Finally, I addressed the issue of nodularity fading. Ductile iron castings with long pouring times can experience a loss of magnesium and rare earth elements, leading to flake or vermicular graphite in the last portions of the casting. The compacted pouring time of 45–55 s minimizes fading. The yttrium-based heavy rare earth nodulizer also provides a longer-lasting nodulizing effect than pure magnesium. The stream inoculant at the pouring station ensures that the last metal entering the mold has enough nucleation sites to form fine, well-shaped graphite nodules. Periodic thermal analysis and sampling of the melt confirmed that the nodularity remained high throughout the pour.
Solidification Simulation and Thermal Analysis
In addition to the empirical calculations, I performed a thermal solidification simulation to verify the riser and chill placement. The simulation model used a finite-difference method to solve the heat conduction equation:
$$\rho c \frac{\partial T}{\partial t} = \nabla \cdot (k \nabla T) + \dot{q}$$
where \(\rho\) is the density, \(c\) is the specific heat, \(k\) is the thermal conductivity, \(T\) is the temperature, \(t\) is time, and \(\dot{q}\) represents the latent heat source due to phase transformation. For ductile iron castings, the latent heat of eutectic solidification is a combination of the heat released by the austenite and graphite formation. I applied the “lever rule” to distribute the latent heat over the solidification temperature range. The simulation predicted that the hot spots were located at the oil-cylinder top and at the oil-injection port, which matched my practical observations. The chills effectively moved the hot spots into the safety risers, allowing complete solidification without internal porosity. The simulation also confirmed that the overflow risers would fill with the first 5–10 kg of metal, preventing cold metal from entering the main cavity.
I also used a simple thermal modulus analysis to verify that the solidification sequence was progressive from thin sections to thick sections. The modulus of each region was calculated as \(M = V/A\). For the 30 mm wall, \(M \approx 15\) mm; for the 90 mm wall, \(M \approx 45\) mm. The chills on the 90 mm wall lower its local modulus so that it solidifies at a similar time to the 30 mm wall. This balanced solidification is the key to producing sound ductile iron castings without relying on large feeders. The combination of a rigid furan resin mold, strategic chills, and small safety risers allowed the internal graphite expansion to do the feeding work.
Table 4 summarizes the key solidification-control features and their purposes.
| Feature | Purpose | Location |
|---|---|---|
| Safety neck-down risers | Feed liquid contraction; relieve pressure before graphite expansion | Oil-cylinder top |
| Exhaust riser pads | Vent gases and provide minor feeding | Support lugs |
| Overflow riser rings | Collect first cold/slaggy metal | Side lugs |
| Shaped chills (70 mm) | Accelerate cooling in heavy oil-cylinder wall | Oil-cylinder top & inside |
| Shaped chills (80 mm) | Prevent shrinkage at oil-injection pad | Oil-injection port |
| Cylindrical chill (φ180×80) | Promote directional solidification from bottom | Oil-cylinder bottom |
Quality Verification and Mechanical Testing
To verify that the process produces acceptable ductile iron castings, I attached a test coupon to the casting as specified. The coupon dimensions were 180 mm × 80 mm × 50 mm, which is a common size for evaluating the mechanical properties of thick-section ductile iron. After shakeout, the coupon was machined and tested. The results are shown in Table 5.
| Property | Required minimum | Measured value | Status |
|---|---|---|---|
| Ultimate tensile strength | 500 MPa | 528 MPa | Pass |
| Yield strength | 320 MPa | 338 MPa | Pass |
| Elongation | 7% | 8.5% | Pass |
| Brinell hardness | 170–230 HB | 187 HB (average) | Pass |
| Nodularity | Grade 1–3 | Grade 2 | Pass |
| Graphite size | Class 5–7 | Class 6 | Pass |
The tensile strength and elongation both exceeded the minimum requirements, indicating a well-inoculated, fully nodular structure with a mixed ferritic-pearlitic matrix. The hardness was uniform across the test coupon, which is important for machining the oil-cylinder bore. The nodularity grade of 2 and graphite size class of 6 confirmed that the nodulization and inoculation treatments were effective. These properties are typical of high-quality ductile iron castings with a pearlite/ferrite balance that gives both strength and ductility.
After machining, the casting was subjected to the standard pressure test. The oil-cylinder bore was sealed and pressurized with hydraulic oil to 25 MPa for 30 minutes. No leaks or surface exudations were observed. This pressure tightness is the ultimate proof that the casting has no interconnected shrinkage porosity or gas channels. The mating surfaces were also checked for dimensional accuracy, and the oil-injection port was inspected for concentricity and soundness. All parts met the drawing requirements.
Conclusion
Through this project, I demonstrated that complex hydraulic press injection bases can be reliably manufactured as ductile iron castings using a well-designed process that combines: (1) a semi-open gating system with foam ceramic filters; (2) a side-bottom pouring position with the oil-cylinder opening upward; (3) small safety neck-down risers and overflow risers; (4) strategically placed shaped chills; (5) thorough venting of the large oil-cylinder core; and (6) a robust melting and inoculation practice with low-rare-earth magnesium nodulizer plus yttrium-based heavy-rare-earth nodulizer. The final product has no shrinkage, gas holes, slag inclusions, or leaks. The mechanical properties comfortably exceed the specification, and the casting process yield is high (93.2%), offering significant cost savings.
This experience reinforces the principle that for ductile iron castings, solidification control is not about simply feeding with large risers, but about using the graphitic expansion to achieve self-feeding. The combination of a rigid mold, proper gating, controlled cooling with chills, and small safety risers provides a robust and economical process. I have successfully transferred this process to batch production, and the parts have received excellent feedback from the customer. The lessons learned here are applicable to many other heavy-section ductile iron castings that require pressure tightness and high mechanical integrity. By paying attention to every detail—from raw materials to pouring temperature—one can achieve world-class quality in ductile iron castings.
The following photograph shows a finished and machined hydraulic press injection base produced with this process. The smooth surface and precise geometry illustrate the capability of the chosen casting method.
In summary, the successful production of these ductile iron castings demonstrates that a scientifically based approach to gating, risering, chilling, and melt treatment can eliminate the age-old problems of shrinkage and gas porosity even in complex, thick-walled components. I hope that the details shared here will provide valuable guidance for other foundry engineers working with ductile iron castings, especially those facing similar challenges with oil-hydraulic components. The combination of theoretical analysis, practical experience, and careful quality control is the key to producing flawless ductile iron castings in a modern foundry.