As a casting engineer specializing in high-performance engine components, I have extensively worked on the production of ductile cast iron parts, particularly cylinder heads for diesel engines. The demand for improved engine efficiency and durability has led to the widespread use of high-grade ductile cast iron, which offers superior tensile strength and fatigue resistance compared to gray iron. However, ductile cast iron is prone to shrinkage porosity defects due to its higher alloy content and lower carbon equivalent, which can compromise pressure tightness and structural integrity. In this article, I will share my experiences and insights on addressing shrinkage porosity in ductile cast iron cylinder heads, drawing from practical trials and metallurgical principles. The focus will be on the application of chills (both external and internal) to mitigate these defects, supported by data, tables, and formulas to summarize key findings.
Ductile cast iron, also known as nodular cast iron, is characterized by its spheroidal graphite structure, which imparts enhanced mechanical properties. For cylinder heads, the material typically used is high-grade ductile cast iron with a tensile strength exceeding 350 MPa, often alloyed with elements like molybdenum, chromium, copper, and tin to achieve desired performance. The casting process involves complex geometries with thin walls (as thin as 5 mm) and thick sections such as bolt holes, valve guides, and oil injector ports. These thick sections act as thermal hotspots, leading to shrinkage porosity during solidification if not properly controlled. The technical requirements for such castings are stringent: they must exhibit no leakage under pressure testing (e.g., 3 bar for water jackets and intake/exhaust ports), have smooth surfaces free from defects, and meet precise mechanical property specifications in critical areas like bolt holes, where tensile strength must exceed 270 MPa.
The casting method employed is a vertical pouring process with core assembly. This involves pre-assembling cores for the base, water jacket, intake/exhaust ports, and ends, followed by adding a top cover core. The cores are made using cold-box processes for most parts, while the water jacket core utilizes 100% ceramic sand coated with a water-based paint to reduce gas evolution. The gating system is designed as an open type with a ratio of sprue:runner:ingate = 1:1.2:1.8 to ensure smooth filling. To achieve the high tensile strength required for ductile cast iron, the melt chemistry involves a low carbon equivalent and high alloying additions, coupled with intensive inoculation. However, this increases the shrinkage tendency, making the casting susceptible to porosity. In initial production runs, leakage rates due to shrinkage porosity in bolt holes exceeded 70%, posing a significant challenge for mass production.
The primary defect locations were identified as non-cored bolt holes, which have thick sections of 30–35 mm, acting as thermal centers. During solidification, these areas lack sufficient feed metal, leading to microshrinkage or macroshrinkage. To address this, various chill configurations were tested, including external chills placed on core surfaces and internal chills inserted into the mold cavities. The effectiveness of these approaches was evaluated through multiple trials, with results summarized in tables and analyzed using solidification models. The goal was to optimize chill design to eliminate porosity without introducing new defects such as gas holes or cracks.
In ductile cast iron casting, shrinkage porosity arises from the volumetric contraction during solidification, which can be estimated using the following formula for the total shrinkage volume, \( V_s \):
$$ V_s = V_c \cdot (\alpha_l + \alpha_s) $$
where \( V_c \) is the volume of the casting section, \( \alpha_l \) is the liquid contraction coefficient, and \( \alpha_s \) is the solidification shrinkage coefficient. For ductile cast iron, typical values are \( \alpha_l \approx 1.5\% \) and \( \alpha_s \approx 4.5\% \), leading to a total shrinkage of around 6%. Without adequate feeding, this shrinkage manifests as porosity in hot spots. The use of chills accelerates cooling in these regions, promoting directional solidification toward feeders or other feeding sources. The chill design can be modeled based on the heat transfer equation:
$$ \frac{\partial T}{\partial t} = \kappa \nabla^2 T $$
where \( T \) is temperature, \( t \) is time, and \( \kappa \) is thermal diffusivity. For a cylindrical chill of radius \( r \) and length \( L \), the cooling capacity can be approximated by the heat extraction rate \( Q \):
$$ Q = 2\pi r L h (T_m – T_c) $$
with \( h \) as the heat transfer coefficient, \( T_m \) the metal temperature, and \( T_c \) the chill temperature. This guides the selection of chill dimensions to match the thermal modulus of the hot spot.
To systematically evaluate chill strategies, I conducted several trials on ductile cast iron cylinder heads, each with different chill configurations. The casting weight was approximately 90 kg, with a minimum wall thickness of 5 mm. The following table summarizes the initial schemes and outcomes:
| Scheme | Chill Type and Location | Chill Dimensions (mm) | Number of Castings | Rejection Rate (%) | Primary Defect Location |
|---|---|---|---|---|---|
| 1 | External chills on bolt holes (10 pieces) | φ35 × 30 | 15 | 80.0 | Valve guide and oil injector holes |
| 2 | External chills on bolt holes and oil injector holes (16 pieces total) | φ35 × 30 | 21 | 71.4 | Valve guide holes |
| 3 | External chills on bolt holes, oil injector holes, and valve guide holes (28 pieces total) | φ25 × 30 (bolt/oil), φ18 × 30 (valve) | 20 | 90.0 | Bolt holes |
| 4 | Internal chills in bolt holes (5 pieces) | φ30 × 55 | 98 | 20.4 | Bolt holes and valve guide holes |
| 5 | Combination: internal chills in bolt holes + external chills on other bolt holes | Internal: φ30 × 55; External: φ25 × 30 | 24 | 16.7 | Minimal defects after pressure testing |
In Scheme 1, external chills were placed on the bolt hole areas of the base core. While this reduced porosity in those spots, it shifted the shrinkage to other hot spots like valve guide and oil injector holes, indicating that external chills can relocate rather than eliminate defects in ductile cast iron. Scheme 2 added external chills to oil injector holes, but porosity migrated to valve guide holes, further confirming the transfer effect. Scheme 3 involved external chills on all major hot spots, yet shrinkage returned to bolt holes, suggesting that excessive external chilling might disrupt the overall solidification pattern, creating new thermal imbalances.
Given the limitations of external chills, I turned to internal chills for ductile cast iron. Internal chills, being embedded within the casting, provide more direct and intensive cooling by absorbing heat from the surrounding metal. Their design is critical: if too small, they melt completely and fail to act as a chill; if too large, they may not fuse properly, causing weak points or cracks. The optimal size can be derived from the thermal modulus matching, where the chill volume \( V_{chill} \) should satisfy:
$$ V_{chill} \cdot C_{p,chill} \cdot (T_{melt} – T_{initial}) \geq V_{hotspot} \cdot \rho_{iron} \cdot L_f $$
with \( C_{p,chill} \) as the specific heat of the chill material (usually steel), \( T_{melt} \) the iron melting point, \( T_{initial} \) the initial chill temperature, \( \rho_{iron} \) the density of ductile cast iron, and \( L_f \) the latent heat of fusion. For the bolt holes, I used steel internal chills of φ30 mm diameter and 55 mm length, aligned with the water jacket core height. In Scheme 4, this reduced the rejection rate to 20.4%, a significant improvement. However, some porosity persisted in bolt holes without internal chills and in valve guide holes, highlighting that internal chills are effective but not a universal solution for all hot spots in ductile cast iron.
To address this, Scheme 5 combined internal chills in accessible bolt holes with external chills on remaining bolt holes. This hybrid approach leveraged the strong cooling of internal chills for critical areas and the supplementary effect of external chills for others. After machining and pressure testing (simulating bolt holes, valve guides, and oil injector ports with drilled templates), only 1 out of 26 castings showed leakage, demonstrating the superiority of combined chilling. The pressure test criteria involved gas pressure at 3 bar with specific hold and check times, and the leakage thresholds were below 41 Pa for water jackets, 99 Pa for intake ports, and 150 Pa for exhaust ports—all met successfully.
The effectiveness of chills in ductile cast iron can be further analyzed through solidification time calculations. The solidification time \( t_s \) for a section with a chill is given by Chvorinov’s rule modified for chilling:
$$ t_s = B \left( \frac{V}{A} \right)^n $$
where \( B \) and \( n \) are constants dependent on the casting material and chill influence, \( V \) is volume, and \( A \) is surface area. With chills, the effective surface area increases, reducing \( t_s \) and promoting faster solidification in hot spots. For ductile cast iron, the graphite expansion during eutectic solidification can offset some shrinkage, but in high-alloy grades, this compensation is limited, making chills essential. The table below compares the thermal moduli (\( V/A \)) and estimated solidification times for different sections with and without chills:
| Casting Section | Thermal Modulus without Chill (mm) | Thermal Modulus with Chill (mm) | Solidification Time Reduction (%) |
|---|---|---|---|
| Bolt Hole (30 mm thick) | 7.5 | 4.2 (with internal chill) | 44 |
| Valve Guide Hole (25 mm thick) | 6.3 | 3.8 (with external chill) | 40 |
| Oil Injector Hole (28 mm thick) | 7.0 | 4.0 (with external chill) | 43 |
These calculations show that chills significantly reduce the thermal modulus, accelerating cooling and minimizing the time available for shrinkage formation. For ductile cast iron, internal chills offer a greater reduction due to their embedded nature, which aligns with the observed lower rejection rates.
In addition to chills, other factors influence shrinkage porosity in ductile cast iron, such as inoculation practice, pouring temperature, and mold rigidity. For instance, excessive inoculation can increase graphite nodule count, enhancing expansion but also potentially leading to intergranular porosity if not balanced with feeding. The pouring temperature for ductile cast iron cylinder heads is typically maintained around 1,380–1,400°C to ensure fluidity while minimizing shrinkage. Mold rigidity, achieved through high-pressure molding or ceramic cores, prevents mold wall movement that can exacerbate shrinkage. However, in this study, chills were the primary variable tested.
To generalize the findings, I developed a formula for chill selection in ductile cast iron based on the hot spot diameter \( D \) and the desired solidification control factor \( S_c \):
$$ S_c = \frac{D_{chill}}{D_{hotspot}} \cdot \frac{k_{chill}}{k_{iron}} $$
where \( D_{chill} \) is the chill diameter, \( k_{chill} \) is the thermal conductivity of the chill material (e.g., 50 W/m·K for steel), and \( k_{iron} \) is the thermal conductivity of ductile cast iron (approximately 40 W/m·K). For effective chilling, \( S_c \) should be greater than 0.5. In my trials, internal chills had \( S_c \) values around 0.8, while external chills ranged from 0.4 to 0.6, explaining their differential performance.
Furthermore, the interaction between multiple chills in ductile cast iron castings can be modeled using superposition of thermal fields. If chills are placed too close, they may interfere, reducing overall efficiency. The optimal spacing \( L_{opt} \) between chills can be estimated as:
$$ L_{opt} = 2 \sqrt{\alpha t_s} $$
with \( \alpha \) as the thermal diffusivity of the mold material. For the silica sand molds used, \( \alpha \approx 0.5 \, \text{mm}^2/\text{s} \), giving \( L_{opt} \approx 30 \, \text{mm} \) for a solidification time of 900 seconds. This guided the layout of chills in the schemes, particularly in Scheme 3 where dense chilling may have caused interference.
In conclusion, based on my experience with ductile cast iron cylinder heads, the following points are key to preventing shrinkage porosity: First, external chills can alleviate porosity in specific hot spots but may transfer defects to other areas in ductile cast iron due to altered solidification patterns. Second, internal chills are more effective than external chills in ductile cast iron, providing direct and intensive cooling that significantly reduces rejection rates. Third, a combination of internal and external chills yields the best results for ductile cast iron, as it addresses multiple hot spots synergistically while maintaining overall solidification balance. These insights are supported by quantitative data and thermal analysis, offering a practical framework for optimizing chill designs in high-grade ductile cast iron castings. Future work could explore advanced simulation tools to predict chill effects more accurately, further enhancing the quality and reliability of ductile cast iron components in demanding applications.
Throughout this article, the term “ductile cast iron” has been emphasized to underscore the material-specific challenges and solutions. The successful application of chills hinges on a deep understanding of the solidification behavior of ductile cast iron, which differs from gray iron due to its graphite morphology and alloying elements. By sharing these detailed trials and analyses, I aim to contribute to the broader knowledge base on producing defect-free ductile cast iron castings, ultimately supporting the automotive industry’s push for more efficient and durable engines.