In our foundry production of a Y-series four-cylinder light-duty diesel engine cylinder head, we encountered a significant and persistent quality issue. The component is a monolithic (integral) cylinder head cast in grey cast iron, grade HT280, with a single rough casting weight of 45 kg and an annual production volume exceeding 40,000 units. A defining characteristic of this part is its lightweight design with thin walls. Our standard production process employs medium-frequency induction furnace melting with a pouring temperature range of 1410–1430°C. Molding is done using green sand, with a pattern yielding four castings per mold. For coremaking, the upper cover and lower deck cores are produced using the triethylamine cold-box process, while the intake port and water jacket cores are made using a shell (resin-coated sand) hot-box process. The gating system is designed as a bottom-running type. Since the commencement of production, final machining has been followed by a wet-type air pressure leak test. The results were concerning, with leakage defects accounting for over 90% of the total scrap. The leakage was predominantly localized around the injector boss areas.
1. Root Cause Identification and Analysis of Leakage
The primary leakage sites were concentrated at the injector bosses for cylinders 2 and 3, constituting more than 80% of all leakage failures. The injector hole, with a diameter of Ø30 mm, is not cored out during casting, thereby creating a significant structural hot spot within the otherwise thin-walled geometry. After machining, which drills out this section, the underlying defect is exposed. The machined surface clearly reveals a grey, honeycomb-like porous structure. Scanning Electron Microscopy (SEM) examination at 100x magnification confirmed the presence of dendritic crystal structures with localized porous cavities, definitively identifying the defect as shrinkage porosity.
This grey cast iron cylinder head has an average wall thickness of approximately 5 mm. The un-cored Ø30 mm injector boss forms a substantial isolated thermal mass. During solidification, this region cools at a slower rate compared to the surrounding thin sections. This differential cooling creates an unfavorable solidification sequence where the thin walls solidify and contract first. By the time the thermal hotspot finally solidifies, the liquid and eutectic contraction cannot be compensated for due to the lack of accessible molten feed metal, leading to the formation of shrinkage porosity. The driving force can be described by considering the volume change during solidification:
$$ V_{shrinkage} = V_{liquid} \cdot (\beta_{L} + \beta_{E}) – V_{feed} $$
Where:
- $V_{shrinkage}$ is the volume of shrinkage porosity formed.
- $V_{liquid}$ is the volume of liquid metal in the hot spot region.
- $\beta_{L}$ is the volumetric contraction coefficient for the liquid-to-solid transition.
- $\beta_{E}$ is the volumetric contraction coefficient during the eutectic reaction.
- $V_{feed}$ is the volume of liquid metal successfully fed into the region from surrounding areas.
In this case, $V_{feed} \rightarrow 0$ because the feeding paths (the thin walls) freeze early, isolating the hot spot.
2. Design, Application, and Mechanism of Internal Chills
Internal chills are particularly suitable for addressing isolated hot spots in thin-walled grey cast iron castings like cylinder heads. They function by altering the local solidification dynamics, effectively increasing the cooling rate of the hot spot to promote directional solidification or to achieve a more simultaneous solidification pattern with the surrounding walls. This approach was deemed optimal given the constraints of our existing bottom-gating system, which made direct feeding via conventional risers impractical for these internal locations.
2.1 Material Selection for the Internal Chill
While chills are often made from a material similar to the casting to ensure compatibility and fusion, we opted for cost-effectiveness and supply chain simplicity. Tin-plated mild steel sheet, a readily available and low-cost material, was selected. The tin plating is critical as it prevents surface oxidation and rusting, ensuring a clean interface with the molten iron to avoid gas generation. Spectrographic analysis of the chill material composition is summarized below:
| Element | C | Si | Mn | P | S | Cr | Ni | Cu | Sn |
|---|---|---|---|---|---|---|---|---|---|
| ω (%) | 0.11 | 0.10 | 0.523 | 0.016 | 0.008 | 0.032 | 0.016 | 0.017 | 0.006 |
2.2 Geometry and Positioning of the Chill
The chill design prioritized simplicity, reliability, and ease of assembly. A complex shape with welded tabs or fins was avoided due to risks of poor fusion, manufacturing complexity, and high cost. The final design is a slotted, hollow cylinder with a slight taper.
| Parameter | Value | Design Rationale |
|---|---|---|
| Wall Thickness | 0.8 mm | Optimized based on hot spot modulus; too thin provides insufficient cooling, too thick risks incomplete fusion or cracking. |
| Diameter | Slightly > core print diameter | Ensures a snug, self-retaining fit within the core print of the sand core via the spring-back of the slotted cylinder. |
| Length | Determined by machining allowance | Ensures the chill remains entirely within the casting wall and is not exposed after final machining. |
| Taper & Slot Angle | 1-2° taper, ~30° slot | Taper aids insertion; slot provides clearance for core geometry and acts as a vent path for core gases. |
The chill is manually inserted into the corresponding print on the water jacket core during the core assembly process. Its elastic nature holds it firmly in place without the need for additional adhesives or supports.
2.3 Metallurgical Mechanism and Effects
Upon casting, the chill acts as a powerful heat sink. The solid steel rapidly extracts heat from the surrounding molten grey cast iron, accelerating the solidification of the hot spot. The goal is to shift the location of any potential shrinkage from the periphery (where it would be exposed by machining) to the very center of the hot spot, or ideally, to eliminate it entirely.
Metallographic examination of a sectioned casting reveals three distinct zones:
- Chill Zone (Steel): The original mild steel microstructure (ferrite-pearlite). Only a thin outer layer melts.
- Transition Zone (Fusion Line): A narrow band (~0.3 mm) where mutual diffusion and melting occur, resulting in a pearlitic structure with minimal graphite.
- Base Metal Zone (Grey Cast Iron): The standard pearlitic matrix with flake graphite of the HT280 grade.
The interface is continuous and free from micro-cracks or gas pores, indicating excellent metallurgical bonding. Microhardness (HV) measurements across the interface further illustrate the gradient:
| Zone | Microstructure | Hardness (HV) |
|---|---|---|
| Chill (Steel Core) | Pearlite + Ferrite | 270 – 279 |
| Transition Zone | Pearlite (fine) | 240 – 260 |
| Base Grey Cast Iron | Pearlite + Graphite | 206 – 230 |
The higher hardness of the chill and transition zone also contributes to leak-tightness by providing a denser, less permeable barrier compared to the graphitic base iron. The effectiveness of the chill can be modeled by considering the enhanced heat extraction. The local solidification time ($t_f$) for the hot spot with a chill can be approximated by:
$$ t_{f,chill} = \frac{\rho L V}{A (h_{eff} \Delta T)} $$
Where $\rho$ is density, $L$ is latent heat, $V$ is volume of the hot spot, $A$ is the effective interfacial area with the chill, $h_{eff}$ is the effective heat transfer coefficient (greatly increased by the metal-to-metal contact), and $\Delta T$ is the temperature difference. Compared to solidification against a sand core ($h_{sand} \approx 500-1000 W/m^2K$), the presence of a steel chill ($h_{steel} \approx 2000-5000 W/m^2K$ during initial contact) drastically reduces $t_{f,chill}$, promoting earlier solidification of the hot spot.
2.4 Potential Risks and Mitigation
While effective, internal chills introduce specific risks that require process control:
- Poor Fusion: Changes in pouring temperature, metal chemistry, or coating on the chill could lead to a cold lap or non-fused interface, creating a leak path itself. Regular destructive audits of castings are necessary.
- Machining Hard Spots: Misplacement of the chill could lead to its hard zone appearing in the machined bore, causing tool wear or damage.
- Stress Concentration: The differential contraction between the chill and the iron could, under certain conditions, promote micro-cracking. Proper chill sizing and avoiding sharp corners are essential.
- Moisture/Gas: Chills must be stored dry and used promptly after core assembly to prevent condensation or gas evolution at the interface.
3. Melt Chemistry Control and Inoculation Practice
The composition and treatment of the grey cast iron melt are fundamental factors influencing shrinkage tendency and microstructure. We implemented targeted adjustments to complement the chilling action.
3.1 Control of Carbon Equivalent and Alloying
For grey cast iron, a lower Carbon Equivalent (CE) generally increases the tendency for shrinkage porosity due to increased volume of the eutectic transformation and reduced graphitization expansion. CE is calculated as:
$$ CE = \%C + \frac{\%Si + \%P}{3} $$
While maintaining the required mechanical properties for HT280 (Tensile Strength ≥ 280 MPa), we aimed for a controlled, mid-range CE. Furthermore, we adjusted the alloying strategy. The original Cu-Cr-Ni combination was modified by removing nickel and introducing a small, controlled addition of tin (Sn). Tin is a potent pearlite stabilizer and promotes undercooled graphite formation, which can refine the eutectic cell structure and improve density. The compositional shift before and after optimization is conceptually summarized below:
| Element | Previous Focus | Optimized Focus | Intended Effect |
|---|---|---|---|
| Carbon (C) | Standard | Slight decrease | Control CE, increase strength |
| Silicon (Si) | Standard | Slight increase | Counteract lower C, promote graphitization |
| Alloys | Cu, Cr, Ni | Cu, Cr, Sn | Replace Ni with Sn for efficient pearlite stabilization and finer graphite. |
3.2 Optimization of Inoculation Practice
Inoculation is critical for achieving the desired graphite morphology and distribution in grey cast iron. However, there is a complex relationship between inoculation and shrinkage. Effective inoculation promotes Type A graphite and a larger number of smaller eutectic cells, which improves the liquid feeding characteristics in the final stages of solidification (enhancing “self-feeding” via graphitization expansion). However, excessive or overly potent inoculation can lead to excessive eutectic undercooling and increased shrinkage tendency.
We use a Zr-bearing ferrosilicon inoculant. To combat fading, a two-stage inoculation process is employed:
- Ladle Inoculation: Primary treatment during tapping.
- Late-Stream Inoculation: Secondary treatment during mold pouring.
The key was precisely controlling the late-stream inoculation amount. Through iterative batch trials, the optimal addition range was established at 0.06–0.08% of the stream weight. This dosage provided sufficient graphite nucleation for good microstructure without exacerbating the shrinkage propensity in the sensitive hot spots. The effect can be related to the number of eutectic cells ($N_{ec}$):
$$ N_{ec} \propto I_{eff} \cdot [Inoculant]_{opt} $$
Where $I_{eff}$ is the inoculant efficiency. An optimal $N_{ec}$ maximizes the self-feeding compensation from graphitization expansion, which can be represented as a compensating pressure ($P_{comp}$):
$$ P_{comp} \approx \alpha \cdot G_{vol} \cdot E_{mod} $$
Where $\alpha$ is a factor related to graphite morphology, $G_{vol}$ is the volumetric expansion due to graphite precipitation, and $E_{mod}$ is the effective modulus of the mushy zone. Proper inoculation maximizes $P_{comp}$ to counteract the hydraulic pressure drop in the feeding liquid.
4. Investigation of Supplementary Process Measures
Parallel to the primary solution, we explored additional methods to mitigate shrinkage in grey cast iron castings for comprehensive knowledge.
4.1 Application of Chill-Intensifying (Te-Based) Coatings
Tellurium (Te) is a powerful carbide stabilizer (anti-graphitizing agent) in grey cast iron. When applied as a coating on a sand core surface, it causes severe undercooling of the iron in contact, resulting in a hard, white iron (chilled) layer and significantly accelerating local solidification. The theory is attractive for replacing metallic chills. We conducted trials by brush-coating a Te-based slurry around the relevant core prints.
Result: The effect was negligible in our specific case. Metallography showed only undercooled graphite (Types D/E) at the interface, not the intended carbidic white layer. The root cause was attributed to the shell core substrate. The fine-grained, low-permeability resin-coated sand prevented adequate penetration and build-up of the viscous Te coating, resulting in an insufficiently thick and concentrated Te layer to induce a true chill. This method proved effective on more permeable cold-box core substrates in other castings.
4.2 Modification of Gating and Feeding System
Redesigning the feeding system to directly supply liquid metal to the hot spot until it solidifies is a fundamental approach. Our current process uses horizontal molding and vertical pouring. Placing a top riser directly over the internal injector boss is geometrically impossible. We investigated side-feeding via a riser attached to the injector boss pad.
Analysis: Preliminary trials and simulations indicated poor efficiency. The main challenges were:
- Restricted Feed Path: Designing an effective “feed neck” geometry that remains open long enough to feed the hot spot is difficult in thin-sectioned grey cast iron castings.
- Inadequate Riser Modulus: The riser size required for effective feeding might be impractically large or interfere with other casting features.
- Gating Complications: Re-routing the gating to feed through the boss area would complicate mold design, increase cleaning costs, and potentially disrupt the overall thermal profile of the casting, affecting mechanical properties elsewhere.
The feeding efficiency $FE$ of such a riser can be expressed as:
$$ FE = \frac{V_{riser, feed}}{V_{hotspot, shrink}} = f(M_r, M_h, t_{open}) $$
Where $M_r$ is the riser modulus, $M_h$ is the hot spot modulus, and $t_{open}$ is the duration the feed path remains open. Achieving a high $FE$ was not practical with the existing layout constraints.
5. Implementation Results and Quantitative Impact
The synergistic implementation of internal chills for targeted hot spot control, coupled with optimized melt carbon equivalent and inoculation practices, yielded a dramatic improvement in product quality.
| Process Parameter | Initial State | Optimized State | Key Impact |
|---|---|---|---|
| Hot Spot Control | None | Internal Steel Chill Insert | Major. Accelerates hot spot solidification, relocates/eliminates shrinkage. |
| Carbon Equivalent (CE) | Standard | Controlled, slightly lowered with Si adjustment | Supportive. Balances strength and shrinkage tendency. |
| Alloying | Cu-Cr-Ni | Cu-Cr-Sn | Supportive. Enhances pearlite stability with efficient Sn addition. |
| Late Inoculation | Variable / High | Precisely controlled at 0.06-0.08% | Critical. Optimizes graphite structure for self-feeding without over-inoculation. |
| Leakage Scrap Rate | Base Level (90%+ of total scrap) | Reduction of ~80% from base | Primary Outcome: Significant quality and cost improvement. |
6. Conclusions and Foundry Engineering Principles
This case study in solving leakage defects in a high-volume grey cast iron component reinforces several key principles in foundry engineering:
- Internal Chills as a Precision Tool: For thin-walled grey cast iron castings with isolated, internally-located thermal hot spots where conventional risers are ineffective, well-designed internal chills are a highly effective solution. Their design must holistically consider material compatibility (surface condition is paramount), cooling capacity (geometry/thickness), ease of placement, and potential risks like fusion integrity and stress.
- Systematic Process Optimization: Defect resolution, particularly for shrinkage-related issues in grey cast iron, rarely relies on a single silver bullet. It requires a systems approach. The successful outcome here was the result of integrating a primary mechanical solution (chills) with supporting metallurgical controls (targeted chemistry and inoculation). The chemical composition defines the inherent shrinkage propensity, while inoculation controls the solidification mechanism that can partially compensate for it.
- Context-Dependent Solution Set: The arsenal for combating shrinkage porosity in grey cast iron includes adjusting melt chemistry, applying chills (internal or external), optimizing feeding systems, using specialty coatings, and controlling pouring parameters. The optimal combination is entirely dependent on the specific product geometry, the location and nature of the defect, the existing production process (molding and coring methods), and economic constraints. What works for a shell core may not work for a cold-box core; what works for a top-feeding system may not be viable for a bottom-gating system.
In conclusion, through the disciplined application of thermal analysis (identifying the hot spot), mechanical intervention (internal chill design), and metallurgical process control, we achieved a substantial reduction in a critical quality defect, enhancing the reliability and manufacturing yield of the grey cast iron cylinder head.