I have developed a practical foundry process for producing complex water-jacket exhaust manifolds as ductile iron castings. The component combines an exhaust gas chamber and a water-cooling jacket in one thin-walled body, and it must survive severe thermal cycling, vibration, and corrosive exhaust conditions. My process uses green sand molding, balanced solidification, high-strength coated sand cores, controlled nodularization, and careful core assembly. Through this work, I have found that successful ductile iron castings of this type depend less on a single breakthrough and more on the systematic control of gating, feeding, core strength, gas venting, mold rigidity, and pouring discipline.
The water-jacket exhaust manifold consists of an air chamber and a water chamber. The air chamber collects high-temperature exhaust gas from each engine cylinder. The water chamber surrounds the air chamber and uses circulating coolant to reduce the temperature of the exhaust wall, the water-jacket wall, and the external radiation. This lowers the exhaust temperature, reduces the engine compartment temperature, and extends the service life of the manifold. Such ductile iron castings are widely used in marine diesel engines and in engines that require explosion-proof performance. The material is normally a ferritic ductile iron, and the rough casting is produced by casting. Because the part has a double-layer structure, the air and water chamber walls are only about 5–6 mm thick, and many isolated hot spots exist, the casting is difficult to feed in a perfectly balanced manner. Shrinkage cavities and porosity are common risks. The water-jacket core is thin, long, and weakly supported, so it can deform, break during pouring, or cause wall thinning. In addition, many cores generate a large volume of gas, and poor venting can lead to misruns, scabs, and back-pressure blow defects. For these reasons, water-jacket exhaust manifolds are among the most difficult ductile iron castings to produce consistently.
Product and Performance Requirements
I began by defining the geometry, mass, material, and performance targets. The casting is a six-cylinder marine diesel water-jacket exhaust manifold. Its external dimensions are 1,135 mm × 186 mm × 178 mm, and its mass is 24.6 kg. The material is QT400-15 ductile iron, which requires a nodularity of at least 80% and a ferrite matrix of at least 90%. The part includes six inlet flanges, one outlet flange, one water inlet, one water outlet, one EGR flange, six process holes, and six bolt bosses. The process holes are critical because they serve three functions: they locate the water-jacket core, they vent gas during pouring, and they provide access for cleaning the water chamber.
| Item | Value or Requirement |
|---|---|
| Outer dimensions | 1,135 mm × 186 mm × 178 mm |
| Casting mass | 24.6 kg |
| Material | QT400-15 ductile iron |
| Minimum tensile strength | 400 MPa |
| Minimum yield strength | 250 MPa |
| Minimum elongation | 15% |
| Air chamber wall thickness | About 6 mm |
| Water chamber wall thickness | About 6 mm |
| Nodularity requirement | ≥80% |
| Ferrite requirement | ≥90% |
| Key features | 6 inlet flanges, 1 outlet flange, 1 water inlet, 1 water outlet, 1 EGR flange, 6 process holes, 6 bolt bosses |
| Typical applications | Marine diesel engines and explosion-proof engines |
The process holes are not merely casting conveniences. In service, they are sealed with bowl-shaped plugs. In production, they allow the water-jacket core to be positioned positively, which reduces core breakage, deformation, thin walls, and parting-line flash in the water chamber. They also act as gas vents: gas generated by the water-jacket core during pouring can escape through the core prints and out through the process holes. Finally, they are essential for cleaning. After knockout, the water-jacket sand can be flushed out through these holes, and shot blasting media can enter the internal water passage to clean the inner surfaces. Without these process holes, the water chamber would be almost impossible to clean reliably in ductile iron castings of this configuration.
Casting Process Challenges
I identified the main challenges before fixing the process. The double-wall structure creates many isolated hot spots. The thin water-jacket core is long and has few supports, so it tends to distort. The large number of cores increases gas evolution. The water chamber parting line can produce internal flash or fins, which restrict coolant flow. Sand inclusion, slag inclusion, scabs, blows, misruns, and shrinkage porosity are all possible. My response was to combine balanced solidification, hot-side risers, cold risers, formed chills, high-strength coarse coated sand, integrated core assembly, controlled clearances, and strong mold venting.
| Challenge | Process Response |
|---|---|
| Double-wall structure with isolated hot spots | Apply balanced solidification; use hot-side risers, cold risers, and formed chills; design short, thin, wide, sloped riser necks |
| Thin, long water-jacket core with low strength | Use high-strength, coarse-grain coated sand; connect core ends with ribs; assemble all cores on a dedicated fixture |
| Core deformation and wall thinning | Control core print clearance, process allowance, draft angle, and parting negative; use rigid core assembly |
| High gas evolution and poor venting | Use low-gas coated sand; add core vents; drill mold vent holes at all core prints; control coating penetration depth |
| Scabs, blows, and misruns | Increase sand permeability; keep pouring temperature and time within limits; ensure progressive filling |
| Water-chamber parting-line flash | Place a sealing pad at the parting line of the upper and lower water cores to block liquid iron |
| Cleaning difficulty in the water chamber | Use process holes; shot blast through access holes; use high-pressure water and endoscope inspection; heat stubborn deposits to 600 °C when needed |
| Shrinkage porosity in ductile iron castings | Use graphite expansion self-feeding; reduce riser volume; place formed chills at local hot spots |
Gating and Feeding Design
I used a middle-pouring gating system and calculated the choke area with the Osann relationship. For ductile iron castings, the gating ratio should promote smooth filling, allow inclusions to float into the runner, and close the gate before graphite expansion begins. I selected a semi-closed gating ratio:
$$
S_{\text{sprue}} : S_{\text{runner}} : \sum S_{\text{gate}} = 1.2 : 1.4 : 1
$$
I estimated the choke area from the pouring weight, iron density, pouring time, and gate velocity:
$$
A_{\text{choke}} = \frac{W}{\rho t v}
$$
where W is the pouring weight, ρ is the liquid iron density, t is the pouring time, and v is the average velocity through the choke. I increased the runner cross-section as much as practical so that the iron rises smoothly and inclusions have time to separate. The ingates were designed short, thin, and wide. This geometry allows the gate to freeze soon after filling, which prevents iron from flowing back out of the mold and allows the graphite expansion to feed the casting internally. The gating layout also helps equalize the temperature field and supports the liquid feeding that is still necessary before expansion begins.
I applied balanced solidification technology. In ductile iron castings, graphite expansion during eutectic solidification can partly or fully compensate for liquid shrinkage and solidification contraction. The net volumetric change can be expressed as:
$$
\epsilon_{\text{net}} = \epsilon_{\text{liquid}} + \epsilon_{\text{solidification}} + \epsilon_{\text{graphite}}
$$
When the graphite expansion term is large enough and the mold is rigid, the net shrinkage can approach zero. Under this condition, the casting can self-feed through the gate and through the expanding graphite. I therefore placed hot-side risers at the inlet flanges, bolt bosses, and outlet flange, and I allowed iron to enter the casting through the risers. I also placed two cold risers opposite the hot risers to balance the temperature field and improve feeding at isolated hot spots. Local formed chills were used where feeding was difficult. The riser neck was designed to be thin, wide, short, and sloped. Thinness ensures timely freezing when feeding stops. Width provides sufficient cross-section. Short length keeps the feed path open. The slope improves the final feeding efficiency.
I also estimated the modulus of hot spots and risers. The modulus is:
$$
M = \frac{V}{A}
$$
where V is the volume and A is the cooling surface area. For conventional feeding, the riser modulus would be larger than the casting hot-spot modulus:
$$
M_r \ge 1.1 M_c
$$
However, for ductile iron castings produced with balanced solidification, I used a lower effective riser modulus:
$$
M_r = (0.6 \sim 1.0) M_c
$$
The exact value depends on carbon equivalent, mold rigidity, pouring temperature, and graphite expansion potential. I also controlled the riser neck modulus:
$$
M_n = (0.5 \sim 0.7) M_h
$$
where M_h is the hot-spot modulus. This approach reduced the number and size of risers and lowered the risk of shrinkage porosity without creating excessive iron consumption.
| Gating and Feeding Parameter | Design Value or Practice |
|---|---|
| Pouring position | Middle pouring |
| Molding method | Green sand, two-box molding |
| Flask size | 1,500 mm × 800 mm × 300 mm |
| Casting per mold | One |
| Gating ratio | \(S_{\text{sprue}}:S_{\text{runner}}:\sum S_{\text{gate}}=1.2:1.4:1\) |
| Ingate geometry | Short, thin, wide |
| Riser type | Hot-side risers at flanges and bosses; cold risers opposite |
| Riser neck geometry | Thin, wide, short, sloped |
| Local feeding | Formed chills at difficult hot spots |
| Solidification strategy | Balanced solidification with graphite expansion self-feeding |
Core Design and Core Making
The core package is the heart of the process. I used a lower shell core, an air-chamber core, a lower water-chamber core, and an upper water-chamber core. The mold cavity is formed by the upper and lower sand molds, the lower shell core, the air-chamber core, and the upper and lower water-chamber cores. The assembly sequence is critical: lower shell core, lower water core, air-chamber core, and upper water core. During design, I controlled core location, clearance, process allowance, draft angle, and parting negative. The clearance between each core and the sand mold or lower shell core was set at 0.3–0.5 mm. The draft angle was 3°–5°. The process negative was 0.3–0.5 mm, and the parting negative was 0.5 mm. To avoid interference and sand dropping during core setting, I increased the clearance between the lower shell core and the lower mold to 1 mm. I also increased the clearance at the water-core parting line to 1.5 mm. This prevents the upper and lower water cores from being compressed and broken during mold closing and leaves room for adhesive.
I used the hot-box core-making method. The cores were made from high-strength, low-gas coated sand. The base sand was silica sand with a SiO₂ content of at least 85%. The grain size was 50/70 mesh. This coarse grain improves permeability and venting, while the high strength reduces core breakage and deformation. The coated sand properties are summarized below.
| Property | Value |
|---|---|
| Grain size | 50–70 mesh |
| SiO₂ content | ≥85% |
| Hot tensile strength | 2.2–2.3 MPa |
| Cold tensile strength | 4.8–5.2 MPa |
| Hot bending strength | ≥2.5 MPa |
| Cold bending strength | 7.0–8.0 MPa |
| Gas evolution | ≤13 mL·g⁻¹ |
| Loss on ignition | ≤3.0% |
The air-chamber core is hollow and has vent pins in the core box, so gas can escape freely. Its core prints are fixed directly to the lower shell core, which gives good connection strength and resists deformation. The water-chamber cores are more difficult. They are long, thin, and weakly supported. I added vent grooves at the parting surface between the upper and lower water cores. The cores are located by core prints and ribs in the lower shell core. This makes the assembly rigid and reduces the chance of deformation or fracture. The water-chamber core has one end that forms the inlet flange holes and another end that forms the process holes. The two ends are connected by transverse ribs, which increase overall strength and improve location reliability.
After core making, I assembled the cores on a dedicated fixture. I applied adhesive at the core joints and cured the assembly. Because the coated sand is coarse, the core surface is relatively rough. I therefore coated the cores. I placed the cores vertically during coating so that the coating penetration depth could be controlled to less than 1 mm. If the coating penetrates too deeply, it can cause scabs, blows, or back-pressure defects during pouring. I also placed a sealing pad at the parting line of the upper and lower water cores. This pad blocks liquid iron from entering the parting line and prevents internal flash or fins that would restrict water flow.
| Core Parameter | Design Value |
|---|---|
| Core-making method | Hot-box process |
| Core sand type | High-strength, low-gas coated sand |
| Base sand | Silica sand with SiO₂ ≥85% |
| Grain size | 50/70 mesh |
| Core-to-mold clearance | 0.3–0.5 mm |
| Lower shell core clearance | 1 mm |
| Water-core parting clearance | 1.5 mm |
| Draft angle | 3°–5° |
| Process negative | 0.3–0.5 mm |
| Parting negative | 0.5 mm |
| Coating penetration | ≤1 mm |
| Core assembly | Dedicated fixture, adhesive bonded, cured |
| Water-core parting seal | Sealing pad to prevent internal flash |
Molding
I used a 148 jolt-squeeze molding machine with green sand. The sand is a bituminous coal and bentonite bonded system, mixed on a DISA line. I selected high-quality coal dust and bentonite to obtain good flowability, strength, and permeability. The molding sand properties are shown below. To improve venting, I controlled the upper mold hardness to about 80 on the C-scale. I also drilled vent holes at every core print in the upper mold. These vent holes allow gas generated by the cores and the mold to escape during pouring. After molding, I placed the assembled core package into the lower mold. Because the core package is heavy and large, it can easily knock loose sand from the mold. I therefore cleaned all loose sand from the mold cavity before closing. After closing, I clamped the upper and lower flasks with mold clamps and prepared for pouring.
| Molding Sand Property | Value |
|---|---|
| Permeability | ≥90 |
| Moisture | 4.5%–5.0% |
| Compactability | 40%–50% |
| Clay content | 16%–18% |
| Wet compressive strength | ≥0.1 MPa |
| Upper mold hardness | About 80 on C-scale |
| Vent holes | At all core prints in upper mold |
| Mold closing | Clamped with mold clamps |
Melting and Nodularization
The material is QT400-15 ductile iron. The target is a nodularity of at least 80% and a ferrite matrix of at least 90%. I melted the iron in a medium-frequency induction furnace. I used a heavy rare-earth magnesium ferrosilicon alloy as the nodulizer. The addition rate was 1.0%–1.2% FeSiMg7Re3. For in-furnace inoculation, I used 0.3%–0.5% FeSi75-C. The nodularization treatment was performed by the sandwich method. I placed the nodulizer in the pocket, covered it with part of the inoculant, compacted it with a flat tool, covered it with dry ductile iron chips, and added perlite on top. The tapping temperature was at least 1,540 °C. I avoided directing the iron stream onto the nodulizer. When the iron had filled about two-thirds of the ladle, I stopped tapping, added the second inoculant during the middle and late stages of the reaction, and then added more iron to the required mass. The reaction time was controlled to 45–100 s. After the reaction, I added a slag remover, skimmed the slag, and covered the iron surface with more slag remover before pouring.
The carbon equivalent is an important control parameter for ductile iron castings. I used the following relationship:
$$
CE = C + \frac{Si}{3} + \frac{P}{3} + \frac{Mn}{6} + \frac{Ni}{15} + \frac{Cu}{13} + \frac{Cr}{4}
$$
For the final composition after nodularization, the carbon equivalent was about 4.60. This value favors graphite expansion and helps balanced solidification. The chemical composition before and after nodularization is given below.
| Element | Base Iron (wt.%) | After Nodularization (wt.%) |
|---|---|---|
| C | 3.79 | 3.61 |
| Si | 1.92 | 2.84 |
| Mn | 0.192 | 0.187 |
| P | 0.029 | 0.029 |
| S | 0.011 | 0.007 |
| Re | — | 0.013 |
| Mg | — | 0.038 |
The magnesium and rare-earth contents are kept within a narrow range. Too little magnesium leads to incomplete nodularization. Too much magnesium increases dross, shrinkage tendency, and gas defects. The sulfur content is reduced from 0.011% to 0.007%, which improves magnesium recovery and nodule formation. The silicon content increases after inoculation, which promotes ferrite formation and graphite expansion. I also control the residual elements to avoid carbides and to maintain the required ferrite matrix in the final ductile iron castings.
Pouring
The pouring temperature was at least 1,380 °C. The entire process from the start of iron treatment to the last mold poured was kept within 15 min. The pouring time itself was kept within 10 min. During pouring, I kept the sprue full at all times. This prevents air and slag from being drawn into the mold and helps maintain a smooth, progressive fill. I used in-stream inoculation with 0.1%–0.15% long-acting inoculant FYJ-1a. The stream was not interrupted. The pouring operator maintained a steady pouring rate so that the metal front advanced uniformly through the thin walls and around the cores. If pouring is interrupted or too slow, cold shut, misrun, and scab defects can occur. If pouring is too fast, mold erosion and gas entrapment can occur. Therefore, I specified a controlled pouring rate and verified it during trial production.
| Pouring Parameter | Control Value |
|---|---|
| Pouring temperature | ≥1,380 °C |
| Time from treatment to last pour | ≤15 min |
| Pouring time | ≤10 min |
| Sprue condition | Always full |
| In-stream inoculation | 0.1%–0.15% FYJ-1a |
| Stream condition | Continuous, no interruption |
Cleaning
After knockout, the casting still contains the water-chamber core, air-chamber core, and shell core. The air chamber is relatively easy to clean compared with the water chamber. The water chamber is small, and internal flash, sand, and shot can become trapped. I use the process holes as cleaning access points. First, I shot blast the internal water passage through the process holes. Then I blow out the steel shot with an air gun. For local dead corners where flash or sand cannot be removed by shot blasting, I use a high-pressure water jet. I inspect the water chamber with an endoscope to confirm that the passage is clear. If there are stubborn deposits or flash that cannot be removed by shot blasting or water jet, I heat the casting to 600 °C and hold it for 1 h. After air cooling, I vibrate the casting on a shakeout machine for 10 min. This usually removes the remaining material. This cleaning sequence is essential because residual internal flash or sand in the water chamber can restrict coolant flow and cause hot spots in service.
| Cleaning Step | Purpose |
|---|---|
| Knockout and shakeout | Remove bulk sand and core material |
| Shot blast through process holes | Clean internal water passage |
| Air blow | Remove loose shot and sand |
| High-pressure water jet | Remove local flash and adhering sand |
| Endoscope inspection | Verify water chamber cleanliness |
| Heat to 600 °C for 1 h, air cool, vibrate 10 min | Remove stubborn deposits and flash |
Quality Results
I produced trial castings according to the process described above. I sectioned the manifold wall for metallographic examination and tested the mechanical properties with an attached test block. The nodularization grade was 2, the nodularity was 92%, and the graphite size was 7. The ferrite content was at least 94%. The tensile strength was 475 MPa, the yield strength was 357 MPa, the elongation was 21%, and the hardness was 168 HB. These values meet the technical requirements for QT400-15 ductile iron castings. The finished casting passed the host engine manufacturer’s inspection and validation tests, including thermal cycling and pressure checks. The process has since been used in batch production. The finished product is shown below.
| Property | Measured Value | Requirement |
|---|---|---|
| Nodularization grade | 2 | Acceptable |
| Nodularity | 92% | ≥80% |
| Graphite size | 7 | Acceptable |
| Ferrite content | ≥94% | ≥90% |
| Tensile strength | 475 MPa | ≥400 MPa |
| Yield strength | 357 MPa | ≥250 MPa |
| Elongation | 21% | ≥15% |
| Hardness | 168 HB | Acceptable |
Defect Control
During trial production, I observed several defects: wall thinning, sand holes, inclusions, blows, scabs, shrinkage porosity, and internal flash. Most of these defects came from process control rather than from the fundamental design. By tightening the process, I reduced them significantly. The finished casting yield has been stabilized above 90%. The table below summarizes the main defects, their causes, and the countermeasures I use.
| Defect | Main Cause | Countermeasure |
|---|---|---|
| Wall thinning | Core deformation, core shift, insufficient core support | Use high-strength coated sand; connect water-core ends with ribs; assemble cores on a fixture; control clearances and process negatives |
| Sand holes | Loose sand in mold cavity, weak core surface, sand wash | Clean mold cavity thoroughly; control mold hardness; improve core coating; reduce pouring turbulence |
| Inclusions | Slag or reaction products entering mold | Skim slag thoroughly; use semi-closed gating; keep sprue full; use in-stream inoculation |
| Blows and scabs | High gas evolution, poor venting, deep coating penetration | Use low-gas coated sand; add core vents and mold vent holes; control coating penetration to ≤1 mm |
| Shrinkage porosity | Isolated hot spots, weak feeding, insufficient graphite expansion | Apply balanced solidification; use hot-side risers and cold risers; place formed chills; control carbon equivalent |
| Internal flash in water chamber | Liquid iron entering water-core parting line | Place sealing pad at water-core parting line; control core parting clearance; verify with endoscope |
| Core breakage | Weak core, rough handling, excessive mold closing force | Use high-strength coated sand; assemble on fixture; increase clearance at critical joints; handle cores carefully |
| Misrun | Low pouring temperature, slow pouring, poor venting | Maintain pouring temperature ≥1,380 °C; keep pouring time ≤10 min; improve mold and core venting |
Process Control Summary
For consistent production of these ductile iron castings, I rely on a fixed set of process parameters. The most important parameters are summarized below. I use these values for process monitoring and for troubleshooting. When a defect appears, I check the parameter that is most likely to have drifted. For example, if blows or scabs appear, I check core gas evolution, coating penetration, mold venting, and pouring temperature. If shrinkage porosity appears, I check carbon equivalent, pouring temperature, riser condition, and chill placement. If wall thinning appears, I check core assembly, core clearance, and core deformation.
| Process Stage | Key Parameter | Control Value |
|---|---|---|
| Melting | Tapping temperature | ≥1,540 °C |
| Nodularization | Nodulizer addition | 1.0%–1.2% FeSiMg7Re3 |
| Nodularization | Reaction time | 45–100 s |
| Inoculation | Furnace inoculation | 0.3%–0.5% FeSi75-C |
| Inoculation | In-stream inoculation | 0.1%–0.15% FYJ-1a |
| Pouring | Pouring temperature | ≥1,380 °C |
| Pouring | Total treatment-to-pour time | ≤15 min |
| Pouring | Pouring time | ≤10 min |
| Molding | Sand permeability | ≥90 |
| Molding | Moisture | 4.5%–5.0% |
| Molding | Compactability | 40%–50% |
| Molding | Wet compressive strength | ≥0.1 MPa |
| Molding | Upper mold hardness | About 80 on C-scale |
| Core making | Coated sand grain size | 50/70 mesh |
| Core making | Gas evolution | ≤13 mL·g⁻¹ |
| Core making | Coating penetration | ≤1 mm |
| Core assembly | Core-to-mold clearance | 0.3–0.5 mm |
| Core assembly | Water-core parting clearance | 1.5 mm |
| Core assembly | Draft angle | 3°–5° |
| Core assembly | Process negative | 0.3–0.5 mm |
| Core assembly | Parting negative | 0.5 mm |
| Cleaning | Heat treatment for stubborn sand | 600 °C for 1 h, air cool, vibrate 10 min |
Why This Process Works for Ductile Iron Castings
The success of this water-jacket exhaust manifold comes from matching the process to the solidification behavior of ductile iron castings. Ductile iron does not behave like a simple shrinkage-prone alloy. If the carbon equivalent is high enough, the graphite expansion during eutectic solidification can offset much of the liquid and solidification shrinkage. However, this self-feeding effect only works when the mold is rigid, the temperature field is balanced, and the gates freeze at the right time. If the gate is too large or freezes too late, iron can flow back out of the mold and the expansion pressure is lost. If the gate is too small or freezes too early, the casting may not fill or may not receive enough liquid feed before expansion begins. The gating ratio and ingate geometry are therefore not arbitrary. They are part of the feeding strategy.
The water-jacket core is another critical factor. In many failed ductile iron castings of this type, the water-jacket core deforms during handling or pouring. The deformation changes the local wall thickness. A thin wall cools faster, which changes the solidification sequence and can create a new hot spot elsewhere. If the core breaks, the water chamber can be blocked by core fragments, and the casting must be scrapped. I solved this by using high-strength coated sand, by adding transverse ribs, by assembling the cores on a fixture, and by controlling the clearances and process negatives. The sealing pad at the water-core parting line is also important because it prevents internal flash, which is difficult to remove and can cause coolant flow restriction.
Venting is equally important. The air-chamber core, water-chamber cores, shell core, and green sand mold all generate gas when they contact liquid iron. If the gas cannot escape, it creates back pressure, which can cause blows, scabs, misruns, and incomplete filling. I used coarse 50/70 mesh sand to increase core permeability, added vent pins and vent grooves, drilled vent holes in the upper mold at every core print, and controlled coating penetration. I also used a low-gas coated sand. These measures reduced gas-related defects substantially. This is a key reason why the process is stable in production.
Finally, the process depends on strict melting and pouring control. The magnesium and rare-earth contents must be within range to produce good nodules. The sulfur content must be low enough to avoid magnesium consumption. The pouring temperature must be high enough to fill the thin walls but not so high that it increases gas evolution and shrinkage. The time from treatment to pouring must be short enough to avoid magnesium fade and inoculation fade. The in-stream inoculation must be continuous. All of these factors affect the final microstructure and properties of the ductile iron castings.
Conclusion
I have successfully developed a production process for ductile iron water-jacket exhaust manifolds with a double-wall structure, 6 mm walls, and multiple isolated hot spots. The process uses green sand molding, balanced solidification, hot-side risers, cold risers, formed chills, high-strength coarse coated sand cores, integrated core assembly, controlled core clearances, strong mold venting, and strict melting and pouring control. The main conclusions are as follows.
(1) Balanced solidification technology allows the feeding and gating systems to be designed with smaller risers and fewer risers. Hot-side risers feed the flanges and bosses, while cold risers and formed chills balance the temperature field and reduce shrinkage porosity in the ductile iron castings.
(2) The core package must be designed as an integrated system. The lower shell core, air-chamber core, and upper and lower water-chamber cores must have the correct clearances, draft angles, process negatives, and parting negatives. Assembling the cores on a dedicated fixture and bonding them before setting them into the mold improves strength, location reliability, and dimensional accuracy.
(3) High-strength, coarse-grain, low-gas coated sand is essential. The 50/70 mesh sand improves permeability and venting, while the high strength reduces core deformation and breakage. Vent pins, vent grooves, mold vent holes, and controlled coating penetration prevent blows, scabs, and misruns.
(4) The process holes serve three purposes: core location, gas venting, and water-chamber cleaning. They are not optional features. They are necessary for reliable production of these ductile iron castings.
(5) Strict control of melting, nodularization, inoculation, pouring temperature, pouring time, mold hardness, and sand properties is required. The final casting achieved 92% nodularity, 94% ferrite, 475 MPa tensile strength, 357 MPa yield strength, and 21% elongation. The casting yield has been stabilized above 90%, and the product meets the customer’s technical requirements and service validation.
This process has been used in batch production and has proven to be reliable for this difficult class of ductile iron castings. When defects do occur, I use the process control table to trace the cause and correct the specific parameter that has drifted. This systematic approach has allowed me to produce a complex water-jacket exhaust manifold with consistent quality and performance.