Among the numerous foundry process methods, green sand casting has found the most extensive application due to its simple process and low production cost, with approximately 80–90% of castings produced via this method. In practical production using clay-bonded green sand, castings are prone to defects such as scabs, burn-on, and sand inclusions. To mitigate these defects and obtain higher-quality castings, foundries commonly add pulverized coal to the green sand. The mechanism involves the added coal dust generating a large volume of reducing gases upon heating, which forms a protective gas film at the mold-metal interface. Furthermore, the thermal decomposition of coal dust leads to the precipitation of a lustrous carbon film at this interface. These combined effects prevent metal penetration into the mold, thereby effectively countering defects like burn-on. However, the use of coal dust also introduces significant problems. On one hand, the toxic and carcinogenic substances produced by coal dust during casting severely impact the ecological environment and workers’ health. On the other hand, at high temperatures, coal dust generates sulfur-containing gaseous compounds that can infiltrate the molten iron. This infiltration poses a particular risk for nodular cast iron (ductile iron) castings, as the sulfur may react with magnesium or rare earth elements present in the surface layer of the iron melt. This reaction can deplete the effective content of these nodularizing elements, potentially leading to poor spheroidization and adversely affecting the final quality of the nodular cast iron casting.
To gain a deeper understanding of the specific impact of traditional coal-dust-bonded sand on the microstructure of castings, our study focused on analyzing the graphite morphology and matrix structure at both the surface and interior of trapezoidal nodular cast iron castings with varying wall thicknesses, all produced using conventional coal dust green sand. This investigation aims to clarify the relationship between the microstructure of nodular cast iron components and their wall thickness, providing practical guidance for industrial production.
1. Experimental Methodology
1.1 Materials
The molding material was conventional coal dust green sand sourced from an operating foundry. The base iron was specified for grade QT450-10 nodular cast iron. Its chemical composition is summarized in Table 1.
| Element | C | Si | Mn | P | S | Mgres | REres |
|---|---|---|---|---|---|---|---|
| Content | 3.70–4.00 | 2.15–2.93 | 0.46–0.66 | 0.010–0.016 | 0.027–0.035 | 0.027–0.050 | 0.026–0.043 |
Other materials included a standard nodularizing alloy (FeSiMg), inoculant, and other standard foundry additives.
1.2 Casting Procedure and Specimen Design
To investigate the effect of wall thickness on the microstructure of nodular cast iron castings, a trapezoidal test casting with an integrated gating system was designed. The trapezoid was parted on its wider base. A sprue and ingate were set up, and the mold was poured using a one-cavity-per-mold pattern. The trapezoidal section had successive step thicknesses of 5 mm, 10 mm, 20 mm, 30 mm, and 40 mm. Each step had a length of 40 mm, and the overall width of the casting was 40 mm. The configuration is illustrated below.
Molds were prepared by hand using the conditioned coal dust green sand. After pouring, the castings were shaken out, shot-blasted, and the gating systems were removed to obtain the final trapezoidal test specimens.
1.3 Metallographic Sample Preparation and Analysis
Sampling for metallographic analysis involved sectioning the trapezoidal casting width-wise through the center of each thickness step using wire electrical discharge machining (EDM). For each step, four distinct locations from the edge to the center of the cross-section were examined, labeled V1 (sub-surface), V2, V3, and V4 (near-center). The sampling scheme is shown conceptually in Figure 2 of the original text (location labels V1-V4).
Metallographic samples were prepared by grinding with abrasive papers, polishing with diamond paste, and cleaning with ethanol. Graphite morphology was observed directly on the polished surface. To reveal the matrix microstructure, samples were etched with a 5% nital solution, followed by ethanol cleaning. Microstructural observation was conducted using a ZEISS inverted optical microscope. Quantification of nodularity, pearlite content, and graphite nodule size grade was performed using dedicated image analysis software.
2. Results and Analysis
2.1 Characteristics of Microstructure in Trapezoidal Castings
1) Graphite Structure from Surface to Interior for Different Wall Thicknesses
The observed graphite structures at locations V1, V2, V3, and V4 for castings of different wall thicknesses are summarized below. Overall, the graphite distribution appeared relatively sparse, and the roundness (nodularity) of the graphite particles was not optimal.
2) Graphite Structure at the Extreme Surface Layer
A critical observation was made at the immediate surface layer (within the first ~100 µm) of all wall thicknesses. As shown in representative micrographs (e.g., Fig. 7 of the original text), spheroidal graphite nodules were virtually absent in this region. This suggests a severe degradation of the nodularizing effect at the metal-mold interface.
3) Matrix Structure from Surface to Interior for Different Wall Thicknesses
After etching, the matrix microstructure was examined. The matrix structures at locations V1 through V4 for different wall thicknesses are summarized. A clear trend was observed: as the wall thickness increased, the ferrite content in the matrix increased significantly, while the pearlite content decreased.
2.2 Quantitative Analysis and Discussion
The nodularity percentages measured at locations V1 to V4 for different wall thicknesses are consolidated in Table 2. A general trend is evident: nodularity decreases progressively from the near-center (V4) towards the sub-surface region (V1). This gradient is consistent with the detrimental effect of sulfur infiltration from the decomposing coal dust, which consumes nodularizing elements (Mg, RE) primarily at and near the surface. The near-complete absence of nodules at the extreme surface strongly confirms that the decomposition products of coal dust at high temperature indeed cause significant nodularization deterioration in nodular cast iron.
| Wall Thickness (mm) | 5 | 10 | 20 | 30 | 40 |
|---|---|---|---|---|---|
| Extreme Surface | Nearly no spheroidal graphite observed | ||||
| V1 (Sub-surface) | 64.33 | 66.06 | 65.29 | 67.52 | 65.55 |
| V2 | 65.88 | 69.25 | 67.53 | 66.12 | 68.26 |
| V3 | 66.13 | 68.72 | 68.26 | 69.25 | 70.89 |
| V4 (Near-center) | 68.59 | 71.52 | 69.80 | 69.63 | 71.54 |
The matrix structure, characterized by the pearlite grade (which correlates with volume fraction), is compared in Table 3. The results show that as the wall thickness decreases, the pearlite grade increases (i.e., more pearlite, less ferrite). This is primarily governed by the cooling rate, which is inversely related to wall thickness. Faster cooling (in thinner sections) suppresses the diffusion-controlled transformation of austenite to ferrite during the eutectoid reaction, favoring the formation of pearlite. The relationship between cooling rate ( $${dT}/{dt}$$ ), transformation kinetics, and the resulting phase fraction can be conceptually described by the Time-Temperature-Transformation (TTT) diagram. The driving force for pearlite formation increases with undercooling below the eutectoid temperature (A1):
$$ \Delta G \propto (T_{eutectoid} – T_{actual}) $$
where a higher cooling rate results in a lower $T_{actual}$ at the transformation point, increasing $ \Delta G $ and promoting the pearlite reaction over ferrite formation.
| Wall Thickness (mm) | 5 | 10 | 20 | 30 | 40 |
|---|---|---|---|---|---|
| V1 | Pearlite Grade 35 | Pearlite Grade 35 | Pearlite Grade 25 | Pearlite Grade 25 | Pearlite Grade 15 |
| V2 | Pearlite Grade 35 | Pearlite Grade 35 | Pearlite Grade 25 | Pearlite Grade 25 | Pearlite Grade 15 |
| V3 | Pearlite Grade 35 | Pearlite Grade 35 | Pearlite Grade 25 | Pearlite Grade 25 | Pearlite Grade 15 |
| V4 | Pearlite Grade 35 | Pearlite Grade 35 | Pearlite Grade 25 | Pearlite Grade 25 | Pearlite Grade 15 |
The core issue with coal dust lies in its high-temperature decomposition. The complex organic compounds in coal break down, releasing volatile matter containing sulfur in forms such as H2S, SO2, and organic sulfides. At the casting temperature, these gases can permeate the boundary layer and react with the nodularizing elements in the iron melt at the interface. The primary detrimental reaction involves the desulfurization of the melt surface by magnesium:
$$ [Mg]_{(in\ Fe)} + H_2S_{(g)} \rightarrow MgS_{(s)} + H_{2(g)} $$
$$ [Mg]_{(in\ Fe)} + [S]_{(in\ Fe)} \rightarrow MgS_{(s)} $$
where $[Mg]$ and $[S]$ represent dissolved magnesium and sulfur in the iron melt. The formation of stable MgS slag removes active, nodule-forming magnesium from the melt in the surface region. The effective magnesium loss, $ \Delta [Mg]_{eff} $, can be modeled as proportional to the sulfur activity at the interface and the exposure time:
$$ \Delta [Mg]_{eff} \propto \int_{0}^{t} k \cdot a_S(t) \cdot dt $$
where $k$ is a rate constant and $a_S(t)$ is the time-dependent sulfur activity at the metal surface, which is directly influenced by the sulfur-bearing gases from the coal dust. This loss mechanism creates a concentration gradient of nodularizing elements from the interior to the surface, leading to the observed reduction in nodularity towards the casting surface and its complete failure at the extreme surface layer. This phenomenon underscores a critical trade-off in using coal dust for nodular cast iron casting: while it improves surface finish by preventing burn-on, it simultaneously impairs the very graphite structure that defines the material’s properties.
3. Conclusions
Through a systematic study of the graphite morphology and matrix structure at both the surface and interior of trapezoidal nodular cast iron castings with varying wall thicknesses produced using traditional coal dust green sand, key relationships between microstructure and processing parameters have been elucidated. The findings offer practical insights for production guidance:
- nodular cast iron castings produced with coal dust green sand generally exhibit a relatively sparse distribution of graphite with sub-optimal nodularity. A distinct gradient exists, with nodularity decreasing progressively from the center towards the sub-surface regions of the casting for all wall thicknesses.
- The most significant finding is the virtual absence of spheroidal graphite in the immediate surface layer (within approximately 100 µm) of the castings. This provides direct metallographic evidence that the high-temperature decomposition products of coal dust severely impair the nodularization process in nodular cast iron, likely through sulfur-induced depletion of nodularizing elements.
- The matrix structure is strongly influenced by cooling rate, which is a function of wall thickness. As the wall thickness decreases, the pearlite grade increases (higher pearlite volume fraction), correspondingly decreasing the ferrite content. This is a direct consequence of faster cooling rates in thinner sections, which thermodynamically and kinetically favor the formation of pearlite over ferrite during the eutectoid transformation. The phase fraction can be related to cooling rate via models derived from TTT diagram analysis.
- This study confirms that coal dust green sand casting presents a dual challenge: environmental and workplace pollution alongside a tangible negative impact on the critical graphite structure of nodular cast iron castings. Therefore, the development and implementation of environmentally friendly molding sand binders and additives that do not compromise casting quality—particularly for high-value materials like nodular cast iron—must be a priority for the foundry industry and researchers. Future work should focus on quantifying the depth of the affected surface layer and correlating it with mechanical property degradation.