With the development of the equipment manufacturing industry towards high-power, heavy-tonnage, and high-strength directions, heavy-section nodular cast iron castings have become increasingly important in the mechanical equipment market. The weight of these nodular cast iron castings can reach hundreds of tons, with the maximum wall thickness exceeding one meter. Nodular cast iron, also known as ductile iron, is a high-performance cast iron with excellent mechanical properties, good formability, and relatively low cost. It is widely used both domestically and internationally to manufacture critical components for large mechanical equipment, particularly as a replacement for large steel castings and forgings, offering significant advantages. However, due to the slow cooling rate and prolonged solidification time—ranging from several hours to twenty or thirty hours—heavy-section nodular cast iron castings often exhibit a series of defects, including reduced graphite nodule count, coarse graphite nodule size, nodularization fading, graphite distortion, graphite flotation, element segregation, and intergranular carbides.
To address these challenges, we conducted a series of pouring experiments using test blocks with a material grade of QT450-10, dimensions of 400 mm × 400 mm × 400 mm, and a single mass of approximately 500 kg. The aim was to explore a suitable melting process for producing heavy-section nodular cast iron, ensuring that the microstructure and mechanical properties of the castings meet standard requirements. This article details our investigation, from initial trials and problem identification through to process optimization and production validation, all presented from a first-person perspective as the research team.
The test block was designed with a bottom-gating system, as illustrated in the casting process schematic. A filter screen was placed in the runner to ensure the cleanliness of the molten metal entering the mold cavity. The ingate was connected to the bottom of the test block using ceramic tubes. To allow the test block to solidify under normal sand mold conditions, no chills or risers were used; only vent sheets were placed on the highest surface. We used furan resin self-hardening sand for molding. Given the substantial thickness of the test block itself, special attention was paid during molding to enhance compaction and improve mold strength, allowing for adequate self-feeding during solidification to achieve a dense internal structure and minimize the impact on mechanical properties. The mold surface was coated with zircon flour paint to reduce chemical reactions between the sand mold and the molten iron.
The melting equipment consisted of a 3-ton medium-frequency induction furnace, and the nodularization treatment was performed using the sandwich method (also known as the pouring-over method). The charge comprised 50% Q10 pig iron and 50% low-manganese bundled scrap steel. The titanium content in the pig iron was controlled to ω(Ti) < 0.035%. For nodularization, a mixture of 50% yttrium-based heavy rare earth nodularizer and 50% light rare earth nodularizer was used, with a total addition of 1.2% of the molten iron weight. For inoculation, 75% silicon-barium inoculant was employed, and a sulfur-oxygen containing inoculant was used for stream inoculation during pouring at a ratio of 0.15%. During melting, silicon carbide with a purity above 90% was added as a preconditioner, and low-nitrogen, low-sulfur, high-temperature graphitized petroleum coke carburizer was used to adjust the carbon content. The target chemical composition for the initial trial is summarized in Table 1.
| Element | Target Range |
|---|---|
| C | 3.6 – 3.7 |
| Si | 2.3 – 2.4 |
| Mn | ≤ 0.2 |
| P | ≤ 0.03 |
| S | ≤ 0.012 |
| RE (Rare Earth) | 0.01 – 0.02 |
| Mg | 0.04 – 0.05 |
| Sb (Antimony) | 0.0045 |
Antimony (Sb) is considered an anti-nodularizing element. However, when added as a trace alloying element to heavy-section nodular cast iron, it can produce beneficial effects. Particularly when the rare earth (RE) content is relatively high, Sb can counteract the excessive RE, improving graphite spheroidity and increasing the graphite nodule count. For addition, Sb was crushed into powder and placed in the pouring ladle’s funnel along with the sulfur-oxygen inoculant for stream inoculation during pouring.
For the nodularizing treatment, a ladle with a well-formed dam was used. The nodularizer was placed on one side of the dam, compacted with a rammer, and covered with 0.2% (of the iron weight) of inoculant. Small steel chips were then layered on top and compacted to delay the ignition time and improve magnesium absorption. The molten iron in the furnace was superheated to 1500°C, held for 5 minutes for overheating treatment, and then the power was reduced. A suitable amount of pig iron was added to the furnace to rapidly lower the temperature to 1420°C before tapping for nodularization. Adding pig iron served a dual purpose: introducing additional graphite nuclei to improve iron quality and enabling rapid cooling to enhance production efficiency. Using the sandwich method, the molten iron was poured into the side of the ladle opposite the nodularizer. Inoculation was performed in multiple stages: in the ladle, during tapping, and during pouring. A dedicated inoculant dispensing device was installed above the furnace. Approximately 0.6% silicon-barium inoculant was added to the dispenser’s funnel. When about one-half to two-thirds of the iron had been tapped, the device was activated to uniformly introduce inoculant into the metal stream at the furnace spout, ensuring complete melting and absorption.
The pouring temperature was set between 1290°C and 1310°C. Two test blocks were cast in one mold according to the described process. Samples for mechanical testing and metallographic analysis were taken from the test blocks by drilling at specified locations, as depicted in the sampling diagram. To ensure data reliability, additional sampling positions were included.
Upon sectioning the test block through its center, a distinct circular dark spot was observed in the central region. Metallographic examination of samples taken from this dark spot area revealed severe graphite distortion, with a substantial amount of chunky graphite (also referred to as exploded graphite or碎块状石墨) present in the core of the test block. Literature indicates that depending on the volume fraction of chunky graphite in the microstructure, the tensile strength of the casting can be reduced by 20% to 40%, elongation by 50% to 80%, and impact toughness by 50%. The mechanical properties of the test blocks from the initial trial are presented in Table 2.
| Sample ID | Tensile Strength (MPa) | Yield Strength (MPa) | Elongation (%) |
|---|---|---|---|
| Target (QT450-10) | ≥ 390 | ≥ 260 | ≥ 8 |
| Upper-1 | 331.1 | 260.64 | 4.5 |
| Lower-1 | 339.1 | 264.35 | 4.0 |
| Upper-2 | 324.3 | 258.91 | 4.0 |
| Lower-2 | 317.2 | 261.58 | 3.0 |
| Upper-3 | 363.0 | 273.10 | 5.5 |
| Lower-3 | 333.6 | 265.95 | 5.0 |
As evident from Table 2 and the metallographic observations, the mechanical properties of samples taken from the large test block were unsatisfactory and failed to meet the standard requirements, primarily due to the formation of extensive chunky graphite.
The formation mechanism of chunky graphite in heavy-section nodular cast iron is complex. Two primary theories exist. One suggests that excessively high carbon and silicon contents lead to the continuous growth and subsequent fragmentation of graphite nodules. The other theory posits that rare earth elements destabilize the austenite shell surrounding the graphite nodules. The process of chunky graphite formation can be described as follows: In slowly cooled regions, the eutectic graphite clusters continue to grow. Under the erosive action of liquid metal convection, graphite at the boundaries of these clusters may become detached, forming free fragments. Larger fragments break into smaller ones due to convection. Furthermore, graphite nodules formed during slow solidification are much larger than typical primary graphite nodules. When they exceed a critical size, internal stresses within the graphite nodule increase continuously. Upon exceeding a certain threshold, the graphite nodule ruptures into fragments. During solidification, liquid metal convection further breaks these fragments into even smaller pieces. Upon final solidification, these manifest as chain-like, dendritic, or other abnormal graphite morphologies. Production experience strongly indicates that the chunky graphite defect in heavy-section nodular cast iron is directly related to the rare earth content in the molten iron. A high concentration of rare earth elements segregates at the austenite grain boundaries, preventing the austenite shell that envelops the graphite from closing for an extended period. Carbon atoms diffuse towards the graphite nodule and grow out along the unclosed paths, leading to graphite distortion and the formation of chunky graphite. Analyzing our initial production process in light of this theoretical understanding, we identified the need to adjust both the nodularizer and the chemical composition.
With advancements in Chinese foundry pig iron production technology, the purity of common Q10 pig iron has significantly improved compared to the past. Following previous experience and using nodularizers with high ω(RE) can lead to significant adverse effects. In our initial process, we used a DF-4 nodularizer with ω(RE) of 3.5%, which proved unsuitable for producing heavy-section nodular cast iron castings. We switched to a DY-8 heavy rare earth nodularizer with ω(RE) of 1.0% and adjusted the mixing ratio with the light rare earth nodularizer (ω(RE) ≈ 0.6%) from 50%/50% to 30% (heavy RE) and 70% (light RE). The objective was to reduce the residual rare earth content in the molten iron to ω(RE) < 0.015%, thereby preventing the formation of chunky graphite. It is noteworthy that some manufacturers producing heavy-section nodular cast iron castings have adopted high-purity pig iron in combination with rare-earth-free nodularizers.
The initially set carbon content of ω(C) = 3.6%–3.7% is generally suitable for standard heavy-section nodular cast iron but is considered high for castings with wall thicknesses exceeding 400 mm. For ultra-thick nodular cast iron with prolonged solidification times, excessively high carbon leads to the precipitation of excessive graphite. Carbon atoms continuously diffuse and promote graphite growth, which can eventually lead to fragmentation, resulting in chunky and exploded graphite. Therefore, we adjusted the carbon content down to 3.3%–3.5%. To prevent the carbon equivalent (CE) from dropping too low, aiming for a CE between approximately 4.25% and 4.35%, we appropriately increased the silicon content to ω(Si) = 2.4%–2.6%. The carbon equivalent can be estimated using the formula:
$$CE = C + \frac{Si}{3}$$
For our adjusted composition, with C=3.4% and Si=2.5%, the approximate CE is 3.4 + 2.5/3 ≈ 4.23%, which falls within the desired range. Additionally, the antimony addition was increased from ω(Sb) = 0.0045% to 0.0075%. The revised target chemical composition is detailed in Table 3.
| Element | Target Range |
|---|---|
| C | 3.3 – 3.5 |
| Si | 2.4 – 2.6 |
| Mn | ≤ 0.2 |
| P | ≤ 0.03 |
| S | ≤ 0.012 |
| RE (Rare Earth) | 0.008 – 0.015 |
| Mg | 0.04 – 0.05 |
| Sb (Antimony) | 0.0075 |
Following the improved formulation, we cast four additional test blocks. After shakeout, samples were taken from designated locations for metallographic and mechanical testing. Sectioning of the improved test blocks revealed no circular dark spots; the machined surfaces exhibited a normal color. Metallographic analysis of samples from the core region showed a nodularization grade of 2, graphite size of 5 to 6, and a pearlite content of less than 5%. Chunky graphite was completely eliminated. The corresponding mechanical properties are listed in Table 4.
| Sample ID | Tensile Strength (MPa) | Yield Strength (MPa) | Elongation (%) |
|---|---|---|---|
| Upper-1 | 404 | 283 | 13.5 |
| Lower-1 | 412 | 283 | 19.0 |
| Upper-2 | 403 | 282 | 15.0 |
| Lower-2 | 409 | 281 | 18.5 |
| Upper-3 | 412 | 285 | 20.5 |
| Lower-3 | 408 | 283 | 19.0 |
All properties from the improved test blocks comfortably exceeded the QT450-10 specifications. The successful suppression of chunky graphite directly correlated with the enhanced ductility (elongation) and strength. This validates the critical importance of controlling rare earth content and optimizing the base composition for heavy-section nodular cast iron.
To further validate the industrial applicability of our optimized melting process, it was implemented in the production of a large casting: a two-plate component for a 3200-ton clamping force machine. The single casting weight was approximately 30 tons, with a maximum wall thickness of 730 mm. Test bars were taken from the thick sections of this production casting. Mechanical testing and microstructural examination confirmed the absence of graphite abnormalities, and the casting met all specified requirements. This successful production run demonstrates the robustness and transferability of our developed melting process for heavy-section nodular cast iron castings.
In conclusion, our systematic investigation into the melting process for heavy-section nodular cast iron has yielded significant insights. The formation of detrimental chunky graphite in thick-walled nodular cast iron castings is predominantly influenced by excessive residual rare earth elements and suboptimal carbon and silicon levels. Through controlled adjustments—specifically, reducing the rare earth content in the nodularizer to achieve a final ω(RE) < 0.015% in the iron, lowering the carbon content to 3.3%–3.5%, increasing silicon to 2.4%–2.6%, and strategically employing a higher antimony addition of 0.0075%—we successfully eliminated chunky graphite. This resulted in a high-quality microstructure characterized by a nodularization grade of 2, graphite size of 5-6, and minimal pearlite, which translated into excellent and consistent mechanical properties exceeding the QT450-10 standard. The process was successfully scaled to produce an industrial-scale casting with a wall thickness over 730 mm. This research underscores that meticulous control over nodularization, inoculation, and base composition is paramount for achieving reliable performance in demanding heavy-section nodular cast iron applications. Future work could explore the precise kinetics of graphite growth under slow cooling conditions, potentially modeled by equations describing diffusion-controlled growth, such as:
$$r(t) = k \sqrt{D t}$$
where \( r(t) \) is the graphite nodule radius at time \( t \), \( k \) is a rate constant, and \( D \) is the diffusion coefficient of carbon in austenite. Understanding these fundamentals will further advance the production of defect-free, heavy-section nodular cast iron components.