In the realm of large-scale wind power generation, the bearing housing stands as a critical component. The operational environment for these structures is exceptionally harsh, encompassing factors such as corrosion, wind-blown sand abrasion, high humidity, and sub-zero temperatures. Within this context, the load-bearing scenario becomes profoundly complex. Compounded by the significant challenges associated with installation and maintenance in remote or elevated locations, the quality requirements for wind turbine bearing housing castings are extraordinarily stringent. To ensure safe and reliable operation over a mandated 25-year service life, these components must possess superior fatigue strength and low susceptibility to brittle fracture. The demanding service conditions, which include substantial operational loads from rotation and vibration coupled with environmental degradation, necessitate a material with exceptional and consistent properties. This is where advanced ductile iron casting technology, particularly through precise inoculation control, plays a pivotal role in meeting these extreme demands.
The process of inoculation in ductile iron casting involves the deliberate introduction of specific agents into the molten metal to control the solidification process and the resulting microstructure. The core principle lies in providing potent nucleation sites for graphite precipitation. These nuclei, typically particles with dimensions on the order of a few micrometers, can be endogenous (formed from reactions within the melt) or exogenous (introduced via the inoculant). Elements with a high affinity for oxygen and sulfur are particularly crucial in establishing favorable nucleation conditions. The primary objectives are to regulate the precipitation of graphite—controlling its quantity, size, and morphology—and to promote the formation of a desired metallic matrix, ultimately enhancing the mechanical properties and structural integrity of the casting.
The choice of inoculant significantly influences the final characteristics of the ductile iron casting. Different inoculant compositions lead to variations in graphite nodule count, matrix structure, and consequently, the tensile strength, ductility, and impact toughness. This study investigates and compares the effects of two distinct secondary inoculants—a silicon-barium-zirconium (Si-Ba-Zr) type and a sulfur-oxygen (S-O) type—on the graphite morphology, microstructure, and mechanical properties of castings intended for wind turbine bearing housings. By analyzing these effects, the research aims to provide a scientific basis for optimizing production processes to achieve the high-performance standards required for this application.
1. Experimental Methodology
1.1 Materials and Charge Composition
The melting trial was conducted using a charge consisting of Q10 grade pig iron, high-quality carbon steel, and returns. The key treatment materials included a low rare-earth (RE) magnesium-ferrosilicon nodularizer (Mg6RE), ferrosilicon (75% Si), and the two secondary inoculants under investigation: a silicon-barium-zirconium inoculant and a sulfur-oxygen inoculant. A high-calcium-barium primary inoculant was also used. The chemical compositions of the principal raw materials are summarized in Table 1.
| Material | C | Si | Mn | S | P | Ba | Ca | RE | Al | Zr | Fe |
|---|---|---|---|---|---|---|---|---|---|---|---|
| Q10 Pig Iron | 4.53 | 0.60 | 0.042 | 0.008 | 0.022 | – | – | – | – | – | Bal. |
| Carbon Steel | 0.06 | 0.04 | 0.18 | 0.005 | 0.003 | – | – | – | – | – | Bal. |
| 75SiFe | – | 72.0 | – | – | – | – | – | – | – | – | Bal. |
| Si-Ba-Zr Inoculant | – | 55.45 | – | – | – | 0.80 | 0.73 | 1.83 | 0.85 | 0.012 | Bal. |
| S-O Inoculant | – | 69.74 | – | – | – | 0.04 | 0.98 | 2.35 | 1.14 | – | Bal. |
1.2 Melting, Treatment, and Casting Procedure
The base iron was melted in a 25-ton medium-frequency induction furnace. The tapping temperature was set at 1460°C, and the melt chemistry was verified using optical emission spectrometry. The nodularizing treatment was performed using a sandwich process in the treatment ladle, with the addition of 1.05 wt.% Mg6RE nodularizer along with 0.35 wt.% of the high-Ca-Ba primary inoculant (3-8 mm grain size). For the secondary inoculation, two separate batches were prepared. One batch received a 0.20 wt.% addition of the Si-Ba-Zr inoculant via stream inoculation during pouring, while the other batch received a 0.20 wt.% addition of the S-O inoculant using the same method. The pouring temperature was carefully controlled between 1360°C and 1370°C. A total of 20 bearing housing castings of grade QT400-18AL were produced, each equipped with an attached test block of dimensions 70 mm x 70 mm x 180 mm, following the production schedule for both inoculant types.
2. Results and Analysis
2.1 Chemical Composition
The chemical composition of the final castings, as determined from the attached test blocks, is presented in Table 2. The results confirm that both processes yielded a ferritic ductile iron casting with closely matched base compositions, ensuring a valid comparison focused on the inoculant effect. Key elements such as carbon (C), silicon (Si), magnesium (Mg), and residual rare earths (RE) are within the typical ranges for high-quality ductile iron.
| Group (Inoculant) | Sample | C | Si | Mn | P | S | RE | Mg |
|---|---|---|---|---|---|---|---|---|
| Group 1 (Si-Ba-Zr) | 1-1 | 3.67 | 1.90 | 0.20 | 0.025 | 0.008 | 0.007 | 0.046 |
| 1-2 | 3.68 | 1.91 | 0.19 | 0.024 | 0.008 | 0.007 | 0.044 | |
| 1-3 | 3.69 | 1.93 | 0.17 | 0.024 | 0.008 | 0.006 | 0.041 | |
| … | … | … | … | … | … | … | … | |
| 1-10 | 3.67 | 1.94 | 0.19 | 0.029 | 0.009 | 0.006 | 0.046 | |
| Avg. (Group 1) | ~3.68 | ~1.93 | ~0.19 | ~0.026 | ~0.009 | ~0.006 | ~0.043 | |
| Group 2 (S-O) | 2-1 | 3.68 | 1.93 | 0.21 | 0.027 | 0.009 | 0.005 | 0.039 |
| 2-2 | 3.68 | 1.95 | 0.20 | 0.022 | 0.008 | 0.006 | 0.042 | |
| … | … | … | … | … | … | … | … | |
| 2-10 | 3.67 | 1.94 | 0.20 | 0.025 | 0.012 | 0.008 | 0.038 | |
| Avg. (Group 2) | ~3.68 | ~1.94 | ~0.20 | ~0.025 | ~0.010 | ~0.007 | ~0.040 |
2.2 Mechanical Performance Analysis
Tensile and impact test specimens were machined from the attached test blocks according to the standard sampling procedure. Tensile tests were conducted at room temperature, and Charpy V-notch impact tests were performed at -40°C to assess low-temperature toughness, a critical property for wind turbine components operating in cold climates. The results are compiled in Table 3.
| Group | Sample | Tensile Strength (MPa) | Yield Strength (MPa) | Elongation (%) | Hardness (HBW) | Impact at -40°C (J) |
|---|---|---|---|---|---|---|
| Group 1 (Si-Ba-Zr) | 1-1 | 361 | 222 | 25.5 | 131 | 12.33 |
| 1-2 | 369 | 229 | 25.5 | 135 | 12.83 | |
| … | … | … | … | … | … | |
| 1-10 | 362 | 222 | 23.5 | 134 | 11.17 | |
| Mean (1) | 363.3 | 224.0 | 23.8 | 134.0 | 12.80 | |
| Std. Dev. (1) | 3.09 | 3.13 | 1.46 | 1.83 | 0.90 | |
| Group 2 (S-O) | 2-1 | 363 | 225 | 23.0 | 130 | 13.17 |
| 2-2 | 363 | 226 | 22.5 | 130 | 12.33 | |
| … | … | … | … | … | … | |
| 2-10 | 366 | 226 | 25.0 | 134 | 10.83 | |
| Mean (2) | 363.1 | 224.0 | 24.1 | 131.1 | 11.93 | |
| Std. Dev. (2) | 1.73 | 3.89 | 1.54 | 1.52 | 0.90 |
Analysis of Mean Values: The data reveals that the average tensile strength, yield strength, and elongation are remarkably similar for both inoculants, with differences well within typical experimental scatter. Both successfully meet the requirements for QT400-18AL grade iron. The most notable difference is observed in the low-temperature impact energy. The ductile iron casting treated with the Si-Ba-Zr inoculant exhibited a mean impact value of approximately 12.8 J, which is about 0.9 J higher than the mean value of 11.9 J for the S-O inoculant-treated iron. This represents a roughly 7% improvement, which could be significant for fracture-critical applications.
Analysis of Property Stability (Scatter): For high-reliability components like wind turbine bearings, consistency (low scatter) in properties is as crucial as the mean performance. The standard deviation (Std. Dev.) for each property, calculated from the 10 samples per group, provides a measure of this stability.
- Tensile Strength: The S-O inoculant produced a lower standard deviation (1.73 MPa vs. 3.09 MPa), indicating superior consistency in tensile strength for this ductile iron casting.
- Yield Strength & Elongation: The Si-Ba-Zr inoculant showed a lower standard deviation in yield strength (3.13 MPa vs. 3.89 MPa) and a marginally lower one in elongation (1.46% vs. 1.54%), suggesting better stability in these properties.
- Impact Toughness: Both inoculants resulted in identical standard deviations (0.90 J) for the -40°C impact values, indicating equivalent stability in low-temperature toughness under these test conditions.
The variability can be further quantified using the Coefficient of Variation (CV), which normalizes the standard deviation by the mean: $$CV = \frac{\sigma}{\mu} \times 100\%$$ where $\sigma$ is the standard deviation and $\mu$ is the mean. Applying this confirms the observations above, showing a lower CV for tensile strength with the S-O inoculant and lower CVs for yield strength and elongation with the Si-Ba-Zr inoculant.
2.3 Microstructural and Graphite Morphology Analysis
The microstructure, particularly the graphite morphology, is the fundamental determinant of mechanical properties in ductile iron casting. Samples for metallographic examination were taken from the residual parts of the tensile test bars. Analysis was performed using optical microscopy, and quantitative assessment of graphite nodule count and nodularity was conducted using image analysis software according to GB/T 9441-2009. The key results are summarized in Table 4.
| Group | Sample | Nodularity (%) | Graphite Size (Primary) | Nodule Count (mm⁻²) | Matrix (Est.) |
|---|---|---|---|---|---|
| Group 1 (Si-Ba-Zr) | 1-1 | 91 | Type 5,6,7 | 133 | >95% Ferrite |
| 1-2 | 92 | Type 5,6,7 | 151 | >95% Ferrite | |
| … | … | … | … | … | |
| 1-10 | 92 | Type 5,6,7 | 145 | >95% Ferrite | |
| Mean (1) | 91.9 | – | 143.8 | >95% Ferrite | |
| Std. Dev. (1) | 1.20 | – | 27.12 | – | |
| Group 2 (S-O) | 2-1 | 91 | Type 5,6,7 | 146 | >95% Ferrite |
| 2-2 | 91 | Type 5,6,7 | 137 | >95% Ferrite | |
| … | … | … | … | … | |
| 2-10 | 91 | Type 5,6,7 | 134 | >95% Ferrite | |
| Mean (2) | 91.3 | – | 140.2 | >95% Ferrite | |
| Std. Dev. (2) | 1.16 | – | 43.50 | – |
Analysis of Mean Values:
- Nodularity: Both inoculants achieved excellent and virtually identical average nodularity levels (~91.5%), ensuring the high-quality spheroidal graphite structure essential for a superior ductile iron casting.
- Nodule Count: The Si-Ba-Zr inoculant produced a slightly higher average graphite nodule count (143.8 mm⁻²) compared to the S-O inoculant (140.2 mm⁻²). A higher nodule count generally refines the matrix structure and can enhance mechanical properties, particularly ductility and impact toughness, which correlates with the observed impact energy advantage.
- Matrix: All samples exhibited a predominantly ferritic matrix (>95%), as expected for the QT400-18AL grade, confirming effective annealing or slow cooling in the test block.
Analysis of Microstructural Stability:
- Nodularity Stability: The standard deviation for nodularity is nearly the same for both groups (~1.18%), indicating both inoculants provide equally consistent and high-quality graphite spheroidization in this ductile iron casting process.
- Nodule Count Stability: A significant difference is observed in the consistency of the nodule count. The Si-Ba-Zr inoculant yielded a standard deviation of 27.1 mm⁻², whereas the S-O inoculant resulted in a much larger scatter of 43.5 mm⁻². This suggests that the Si-Ba-Zr inoculant provides more uniform and reliable nucleation potency, leading to a more reproducible graphite dispersion.
3. Discussion: Mechanisms of Inoculant Action
The observed differences in the performance of the ductile iron casting can be traced back to the distinct mechanisms of the two inoculants. The efficacy of an inoculant is governed by its ability to provide stable, heterogeneous nucleation sites that survive in the molten iron and promote graphite precipitation at a high rate.
Sulfur-Oxygen (S-O) Inoculant: This type typically contains a significant amount of Rare Earth (RE) elements, primarily Cerium (Ce), in the range of 1.5-2.0%. Cerium has a very high affinity for both sulfur and oxygen. In the melt, it reacts to form stable compounds:
$$ \text{Ce} + \text{O} \rightarrow \text{CeO}_x $$
$$ \text{Ce} + \text{S} \rightarrow \text{CeS} $$
$$ 2\text{Ce} + 2\text{O} + \text{S} \rightarrow \text{Ce}_2\text{O}_2\text{S} $$
These cerium oxy-sulfides and sulfides are potent, high-melting-point nucleation substrates for graphite. They are effective in increasing the nodule count and reducing chilling tendency. The presence of active S and O in the inoculant composition ensures the formation of these nuclei. The high density of these Ce-based compounds minimizes their flotation, aiding in their distribution.
Silicon-Barium-Zirconium (Si-Ba-Zr) Inoculant: This inoculant leverages a different set of elements. Barium (Ba) forms stable oxides and sulfides (BaO, BaS) which are also effective nucleation sites. Barium-containing nuclei are known for their relatively slow flotation rates due to favorable density, contributing to a uniform distribution of graphite nodules and potentially better property stability, as seen in the yield strength and elongation results. Zirconium (Zr) has a strong affinity for nitrogen (N₂):
$$ \text{Zr} + \text{N}_2 \rightarrow \text{ZrN} $$
This can be beneficial in neutralizing any detrimental nitrogen picked up from resin-bonded molds or cores, although this effect is often minor. The combination of Ba and Zr appears to offer a robust and consistent nucleation package, resulting in the observed higher stability in nodule count and certain mechanical properties for this ductile iron casting.
The correlation between microstructure and the slightly higher impact energy for the Si-Ba-Zr group, despite a similar nodularity, may be linked to the combined effect of a marginally higher and more uniform nodule count and the potential matrix-refining effect of the uniformly distributed fine graphite nodules. The impact toughness of ferritic ductile iron casting can be empirically related to nodule count and nodularity through relationships that consider the mean free path in the ferrite matrix. A simplified view is that toughness tends to increase with higher nodule count (finer inter-nodular spacing) and higher nodularity.
4. Conclusion
This investigation into the effect of secondary inoculants on the properties of QT400-18AL ductile iron casting for wind turbine bearing housings yields the following key conclusions:
- Mechanical Performance: Both the silicon-barium-zirconium (Si-Ba-Zr) and sulfur-oxygen (S-O) inoculants produce castings that comfortably meet the standard grade requirements. The Si-Ba-Zr inoculant provided a consistent advantage in low-temperature (-40°C) impact toughness, with an average value approximately 7% higher than that achieved with the S-O inoculant.
- Performance Stability: The inoculants differ in their consistency across different properties. The S-O inoculant demonstrated superior stability (lower scatter) in tensile strength. Conversely, the Si-Ba-Zr inoculant provided better stability in yield strength and elongation, and significantly greater consistency in graphite nodule count throughout the casting.
- Microstructural Influence: Both inoculants achieved excellent and equivalent average nodularity (>91%). The Si-Ba-Zr inoculant promoted a slightly higher and, more importantly, a much more uniform distribution of graphite nodules per unit area, which is a critical factor for homogeneous mechanical behavior in large or complex ductile iron casting.
- Inoculant Selection: The choice between inoculants for a critical application like wind turbine bearings depends on the specific performance priority. If maximizing and ensuring consistent low-temperature impact toughness is paramount, the Si-Ba-Zr inoculant is favorable. If the utmost consistency in tensile strength is the primary concern, the S-O inoculant may be preferred. The superior microstructural uniformity provided by the Si-Ba-Zr inoculant, as evidenced by the stable nodule count, makes it a strong candidate for producing reliable, high-integrity ductile iron casting components subjected to complex and demanding service conditions.
This study underscores that beyond achieving nominal property averages, the strategic selection of inoculant technology is essential for controlling the statistical distribution and microstructural homogeneity of properties, which is fundamental to the guaranteed long-term reliability of wind turbine components manufactured via ductile iron casting.