In the production of large-scale marine diesel engine components, such as crankcases and cylinder blocks, the material of choice is often high-grade nodular cast iron (ductile iron). These castings are characterized by complex geometries, significant variations in wall thickness (ranging from 10 mm to over 180 mm), and stringent quality requirements, including high impact toughness and rigorous non-destructive testing standards. Producing sound, defect-free heavy-section castings in nodular cast iron presents a significant challenge. Three predominant defects persistently threaten the integrity and performance of these components: shrinkage porosity, graphite flotation, and the formation of degenerate graphite structures, notably chunky graphite. This article details a systematic, first-person investigation into the influencing factors behind these defects, culminating in optimized production parameters for high-quality marine castings.
The core challenge lies in balancing the chemical composition, particularly the Carbon Equivalent (CE), to mitigate these interconnected defects. A high CE promotes good graphite nucleation and expansion, which can counteract the inherent shrinkage tendency of nodular cast iron. However, an excessively high CE in thick sections leads to graphite flotation. Conversely, a lower CE increases the risk of shrinkage porosity. Furthermore, heavy sections are prone to the formation of chunky graphite, which severely degrades mechanical properties. Therefore, defining an optimal “window” for CE and understanding the role of minor elements and cooling control is critical.
1. Materials and Experimental Methodology
To simulate the thermal conditions of actual large castings, a step-block geometry was designed. This design featured five distinct sections with wall thicknesses of 40 mm, 80 mm, 120 mm, 180 mm, and 220 mm, each with a length of 250-300 mm. The step-block pattern and gating system are illustrated in the schematic.
The base iron was prepared in a 1.5-ton medium-frequency induction furnace using a charge of 65% high-purity pig iron and 35% premium steel scrap, with recarburizer added to achieve the target carbon content. The molten iron was treated via the standard sandwich method using a rare-earth magnesium ferroalloy as the nodulizer and a Si-Ba alloy as the inoculant. The treatment was conducted at 1470-1490°C, with one ton of iron processed per ladle.
Four distinct melts with varying Carbon Equivalent values were prepared. The chemical composition of the treated iron for each trial is summarized in Table 1. The Carbon Equivalent was calculated using the standard formula for cast iron:
$$
CE = \%C + \frac{1}{3}\%Si
$$
| Trial Scheme | CE (%) | Mn (%) | P (%) | S (%) |
|---|---|---|---|---|
| Scheme 1 | 4.01 | 0.22 | 0.015 | 0.010 |
| Scheme 2 | 4.21 | 0.22 | 0.015 | 0.010 |
| Scheme 3 | 4.42 | 0.23 | 0.015 | 0.012 |
| Scheme 4 | 4.61 | 0.22 | 0.015 | 0.012 |
Each ladle was used to pour four step-block molds at different pouring temperatures: 1370°C, 1350°C, 1330°C, and 1310°C. Thermal analysis was performed on the treated iron to record its cooling curve. After shakeout and shot blasting, the step-blocks were sectioned. Each thickness section was visually inspected for macroscopic shrinkage and graphite flotation. Metallographic samples were extracted from both the upper region and the center of each section to evaluate graphite morphology (nodularity, size, presence of degenerate forms) and matrix structure using optical microscopy.
2. Results and Discussion
2.1 The Dual Role of Carbon Equivalent: Shrinkage vs. Graphite Flotation
Thermal analysis curves for the different CE irons, all poured at 1370°C, provided initial insights. Key parameters derived from these curves are the undercooling during the austenite formation stage (G1) and the recalescence during the eutectic graphite precipitation stage (G3). A higher G1 value indicates a larger austenite formation range, typical of hypoeutectic compositions with higher shrinkage propensity. A higher G3 value signifies a strong graphitization potential, which promotes internal feeding and reduces micro-shrinkage. The analysis confirmed that as CE increased from 4.01% to 4.61%, G1 decreased significantly while G3 increased, predicting a reduction in shrinkage tendency with higher CE.
The visual inspection of the sectioned blocks validated this prediction. The relationship between CE, wall thickness, and the occurrence of shrinkage is summarized in Table 2. For a given CE, shrinkage propensity increased with wall thickness. For a given wall thickness, increasing the CE effectively reduced or eliminated shrinkage cavities. This is due to the increased graphite expansion force which compensates for the liquid and solidification contraction, a fundamental principle of self-feeding in nodular cast iron.
| CE (%) | 40 mm | 80 mm | 120 mm | 180 mm | 220 mm |
|---|---|---|---|---|---|
| 4.01 | None | None | None | Minor | Significant |
| 4.21 | None | None | None | Minor | Significant |
| 4.42 | None | None | None | None | None |
| 4.61 | None | None | None | None | None |
However, the benefit of a high CE is counteracted by the risk of graphite flotation. As CE increases, the quantity of graphite nodules precipitating during solidification increases. In thick sections with long solidification times, these nodules have sufficient opportunity to float upwards due to their lower density compared to the liquid iron, accumulating at the upper surfaces of the casting. The correlation is clearly shown in Table 3. Graphite flotation became pronounced in sections of 120 mm and above when the CE reached 4.61%.
| CE (%) | 40 mm | 80 mm | 120 mm | 180 mm | 220 mm |
|---|---|---|---|---|---|
| 4.01 | None | None | None | Minor | Minor |
| 4.21 | None | None | None | Minor | Minor |
| 4.42 | None | None | None | Significant | Significant |
| 4.61 | None | Minor | Significant | Heavy | Heavy |
From this experimental data, a critical processing window was identified. For the tested pouring temperatures (~1370°C) and for wall thicknesses relevant to typical castings (40-120 mm), a CE range of approximately 4.0% to 4.4% provided an effective balance, avoiding both significant shrinkage and graphite flotation. This finding is crucial for defining the melting specifications for heavy-section nodular cast iron components.
2.2 Controlling Degenerate Graphite: The Role of Antimony and Chill Cooling
Metallographic examination of the central regions of thick sections revealed another critical issue: the formation of chunky graphite. This degenerate form appears as irregular, clustered, blocky graphite aggregates that drastically reduce ductility and impact toughness. The incidence of chunky graphite was found to be strongly dependent on both CE and section size, as summarized in Table 4.
| CE (%) | 40 mm | 80 mm | 120 mm | 180 mm | 220 mm |
|---|---|---|---|---|---|
| 4.01 | None | None | None | Minor | Minor |
| 4.21 | None | None | None | Minor | Minor |
| 4.42 | None | None | None | Significant | Significant |
| 4.61 | None | None | Minor | Significant | Heavy |
To combat this, the influence of the trace element Antimony (Sb) was investigated. Sb is known to segregate at the graphite/liquid interface, altering interfacial energy and diffusion kinetics. It is proposed that Sb reduces the diffusion rate of carbon atoms in the melt, thereby slowing the growth rate of graphite nodules. This suppression of excessive growth helps maintain a spherical morphology and prevents the transition to chunky forms. Additional trial schemes were conducted with controlled Sb additions, with and without the use of external chills placed at the bottom of specific step-block sections (80mm and 120mm). The chemical compositions for these trials are listed in Table 5.
| Trial Scheme | CE (%) | Sb (%) | Chill Used? |
|---|---|---|---|
| Scheme 5 | 4.42 | 0.005 | No |
| Scheme 6 | 4.42 | 0.005 | Yes |
| Scheme 7 | 4.32 | 0.007 | No |
| Scheme 8 | 4.32 | 0.007 | Yes |
The results were conclusive. In all schemes where Sb was added (0.005-0.007%), chunky graphite was effectively eliminated from the centers of even the thickest sections (180mm and 220mm). The combined effect of Sb and chills was particularly beneficial. The use of chills dramatically increased the local cooling rate, which significantly improved graphite nodule count, nodularity (roundness), and size distribution. Graphite size rating improved from level 4-5 (without chill) to level 6 (with chill) according to standard ASTM A247. This can be described by a simplified relationship for nodule count (N_v):
$$
N_v \propto \frac{1}{t_s^n}
$$
where $t_s$ is the local solidification time and $n$ is a positive exponent. The chill reduces $t_s$, thereby increasing $N_v$.
However, a critical trade-off was observed with Sb addition. Antimony is a potent pearlite stabilizer. While an addition of 0.005% Sb successfully suppressed chunky graphite without severely affecting the ferritic matrix, increasing the Sb content to 0.007% caused a substantial increase in the pearlite fraction in the matrix (from less than 5% to over 35-45%), as shown in the mechanical and microstructural data from 80mm sections (Table 6). This is unacceptable for ferritic grade nodular cast iron like QT400-18.
| Scheme (Sb, Chill) | Tensile Strength (MPa) | Yield Strength (MPa) | Elongation (%) | Nodularity (%) | Graphite Size | Matrix (F+P) |
|---|---|---|---|---|---|---|
| 5 (0.005%, No) | 426 | 279 | 18 | 80 | 5 | F + P~15% |
| 6 (0.005%, Yes) | 415 | 281 | 20 | 90 | 6 | F + P<5% |
| 7 (0.007%, No) | 457 | 298 | 12 | 80 | 5 | F + P~35% |
| 8 (0.007%, Yes) | 464 | 291 | 11 | 80 | 5 | F + P~45% |
Therefore, for ferritic grades, the use of Sb must be carefully controlled and minimized, relying more on enhanced cooling (chills) and optimal CE control to manage chunky graphite.
3. Industrial Application and Process Optimization
Based on the empirical data from this study, the melting and processing parameters for large marine engine block castings in nodular cast iron (QT400-18 grade) were successfully optimized. The revised production protocol is as follows:
Charge Makeup: 65% High-purity pig iron + 35% Selected steel scrap.
Base Iron Target: CE: 4.2% – 4.5%; Mn ≤ 0.25%; P ≤ 0.05%; S < 0.02%; Trace Sb (controlled, minimal addition).
Treatment: Sandwich method with low-RE MgFeSi alloy (1.2% addition) and Si-Ba inoculant (0.6-0.7% addition) at 1470-1490°C.
Molding: Strategic use of chills in isolated heavy sections and thick junctions to enhance local cooling rates.
This optimized procedure was implemented for the production of five large engine block castings. The properties from separately cast test coupons and from cast-on coupons (attached to the actual castings) were evaluated. All results, summarized in Table 7, met or exceeded the specified requirements for QT400-18AR. Tensile strength, elongation, and impact values were consistent and satisfactory. Microstructural evaluation confirmed a predominantly ferritic matrix (>95% ferrite) with good nodularity (Grade 2) and fine graphite size (Grade 5-6). Non-destructive testing (RT, UT, PT) and hydrostatic pressure tests were also successfully passed.
| Sample Source | Tensile Strength (MPa) | Yield Strength (MPa) | Elongation (%) | Impact Energy (J) | Nodularity Grade | Graphite Size Grade | Ferrite Fraction |
|---|---|---|---|---|---|---|---|
| Cast Coupon (P01) | 407 | 274 | 25.5 | 20 (avg) | 2 | 6 | >95% |
| Attached Block (E01) | 401 | 279 | 27.5 | 21 (avg) | 2 | 6 | >95% |
| Casting Body (E01) | 402 | 263 | 27.0 | 18 (avg) | 2-3 | 5 | >95% |
4. Conclusions
This investigation provides a clear, data-driven framework for controlling defects in heavy-section nodular cast iron castings. The key findings are:
1. The Carbon Equivalent (CE) is the primary controlling factor for the shrinkage-porosity vs. graphite-flotation balance. For the studied conditions, a CE window of 4.2% to 4.5% is recommended to produce sound castings in sections up to 120mm thick. The relationship can be conceptually modeled by defining a soundness index $S$ that must be maximized:
$$
S(CE, d) = f_{shrink}(CE, d) \cdot f_{float}(CE, d)
$$
where $d$ is section thickness, $f_{shrink}$ decreases with increasing CE, and $f_{float}$ decreases with decreasing CE.
2. The formation of chunky graphite in heavy sections is a major concern that increases with both CE and section size. It can be effectively suppressed by the careful addition of the trace element Antimony (Sb). The mechanism involves Sb altering the interfacial kinetics and carbon diffusion, expressed as a reduction in the graphite growth rate constant $k_g$:
$$
\frac{dr}{dt} = k_g \cdot \Delta C \quad \text{where } k_g \downarrow \text{ with Sb addition}
$$
where $r$ is nodule radius and $\Delta C$ is the driving force.
3. For ferritic grades of nodular cast iron, the Sb addition must be strictly minimized (e.g., ≤0.005%) due to its strong pearlite-promoting effect. The primary defense against chunky graphite should be enhanced cooling via strategically placed chills, which refine the graphite structure, increase nodule count, and improve mechanical properties.
4. A holistic approach combining optimized CE, minimal controlled Sb addition (if necessary), and aggressive cooling design with chills is essential for the reliable production of high-integrity, heavy-section nodular cast iron components for demanding marine applications. This systematic methodology transforms the production of large nodular cast iron castings from an empirical art into a more controlled engineering practice.