In my work on wind-power components, I have concentrated on ductile iron casting because it offers a combination of low cost, excellent formability, and the ability to replace large cast steel parts in many structural applications. From a production standpoint, however, ductile iron casting for wind power is not a simple substitution. The same slow cooling and long solidification time that allow large wind-power sections to be cast also promote spheroidization fading and inoculation fading. In my experience, inoculation fading is often the dominant problem. When fading occurs, the nodularity decreases, graphite shape becomes less regular, carbides and shrinkage defects become more likely, and the mechanical property scatter increases. Therefore, I have focused my experimental work on how spheroidizing and inoculation treatment influences the metallographic structure and mechanical properties of wind-power ductile iron casting. My objective is to provide a practical basis for controlling microstructure and performance in industrial production.
I have found that multiple inoculation is necessary for large wind-power ductile iron casting. The melt must be treated not only during spheroidization but also during tapping, transfer, and pouring. The size, mass, feed rate, and number of inoculant additions must be matched to filling and solidification behavior, especially the filling velocity of the liquid iron. When this matching is done correctly, inoculation fading is slowed, graphite nodule count increases, nodularity improves, and the final ductile iron casting becomes more uniform.
A small-dose stream inoculant, which I designate as inoculant C, contains special substances that react with calcium and cerium. These reactions form complex compounds that act as excellent graphite nuclei in the later stage of solidification. They increase the number of graphite nodules, improve the nodularity rate, and produce a high-density graphite distribution. As a result, carbides and shrinkage porosity are reduced, the ferrite matrix is increased, and the chill tendency is lowered. If no instantaneous inoculant is added during pouring, I must lower the pouring temperature appropriately to allow graphitizing expansion to provide self-feeding. However, low-temperature pouring can also cause insufficient spheroidizing and inoculation, and slag holes may appear in the ductile iron casting. For these reasons, a well-designed inoculation schedule is one of the most critical factors in producing wind-power ductile iron casting.
Experimental Materials and Process Design
For my trials, I selected Q8 pig iron and special carbon steel as the charge materials. I used two nodulizers, designated A and B, and three inoculants, designated C, D, and E. The chemical compositions of the raw materials are summarized in Table 1. These values are given as mass fractions in percent, and they guided my charge calculation and process control.
| Material | Grade | C | Si | Mn | S | P | RE | Ca | Ba | Al | Mg |
|---|---|---|---|---|---|---|---|---|---|---|---|
| Pig iron | Q8 | 4.6 | 0.55 | – | 0.020 | 0.02 | – | – | – | – | – |
| Steel scrap | – | 0.05 | 0.005 | 0.23 | 0.008 | 0.015 | – | – | – | – | – |
| Nodulizer | A | – | 45–48 | 0.5–1.5 | 1.5–2.5 | 1–3 | <0.8 | – | – | – | 5.5–6.5 |
| Nodulizer | B | – | 44–48 | 0.4–0.6 | 0.8–1.2 | – | – | – | – | – | 5.5–6.1 |
| Inoculant | C | – | 70–76 | – | 0.7–1.3 | – | 0.7–1.3 | – | – | – | – |
| Inoculant | D | – | 67–72 | – | 1.5–2.0 | 2–4 | – | – | – | – | – |
| Inoculant | E | – | 72–77 | – | 1–2 | 2–3 | 0.8–1.5 | – | – | – | – |
I used the sandwich method for spheroidization. The tapping temperature was approximately 1500 °C. On one side of the pocket dam in the ladle, I first placed 1.2%–1.5% nodulizer and 0.002%–0.01% antimony alloy. Then I covered the nodulizer with 0.6%–1.0% inoculant E, which is a long-acting composite inoculant containing RE, Ba, Ca, and other elements. Finally, I covered the inoculant with a layer of silicon steel sheet. These three layers were used mainly to avoid an excessively fast spheroidization reaction. During tapping, I applied 0.1%–0.2% mixed inoculant on the furnace platform as stream inoculation. After spheroidization, I quickly skimmed the slag, and the pouring temperature was controlled between 1320 °C and 1340 °C. During pouring, I also used 0.1%–0.2% inoculant C or D for instantaneous stream inoculation.
To compare the effect of inoculation, I conducted a reference condition without stream inoculation. In the stream-inoculated condition, the spheroidization temperature was 1460–1480 °C and the pouring temperature was 1330–1350 °C. In the condition without stream inoculation, the spheroidization temperature was reduced to 1430–1440 °C and the pouring temperature was 1320–1340 °C. I also considered the influence of alloying elements such as antimony, high-temperature holding time in the furnace, holding time in the ladle, chills, and casting wall thickness on the graphite morphology of wind-power ductile iron casting. In the molding process, I designed the use of chills reasonably. In the melting process, I reduced high-temperature holding time in the furnace and holding time in the ladle. According to the casting wall thickness and structure, I controlled the carbon content after spheroidization to 3.1%–3.7%, the carbon equivalent to 4.1%–4.5%, the silicon content to 1.5%–1.9%, Mn ≤ 0.3%, P ≤ 0.03%, S ≤ 0.015%, residual RE to 0.01%–0.03%, and residual Mg to 0.03%–0.05%.
The carbon equivalent was evaluated using the following expression:
$$CE = C + \frac{Si + P}{3}$$
Here, CE is the carbon equivalent, C is the carbon content, Si is the silicon content, and P is the phosphorus content, all expressed as mass fractions in percent. This formula helped me keep the ductile iron casting within a composition window that favors graphite spheroidization and reduces chill.
My main melting equipment was a 15 t/h, 8000 kW medium-frequency coreless induction furnace. I used ICP analysis to determine the composition of the casting body and attached test blocks. I used an optical metallographic microscope to observe graphite nodule size and graphite morphology in the body and attached test blocks, and I compared the results with standard metallographic charts to evaluate the nodularity grade and pearlite content. I used a scanning electron microscope to study the microstructural features and the surface morphology of degenerate graphite. I used an EDX spectrometer for local chemical analysis. I used a computer-controlled electronic universal testing machine to measure tensile strength and elongation, and a low-temperature impact tester to measure low-temperature impact toughness.
Metallographic and Mechanical Property Results
I examined 20 heats, labeled 001# through 021#. The chemical compositions of the trial ductile iron casting test blocks are given in Table 2. The metallographic structures and mechanical properties are given in Table 3. These data form the basis of my comparison between stream inoculation and no stream inoculation.
| Heat | Test block | C | Si | Mn | S | P | Mg residual | Sb |
|---|---|---|---|---|---|---|---|---|
| 1 | 001# | 3.69 | 1.75 | 0.12 | 0.015 | 0.023 | 0.034 | 0.002 |
| 2 | 003# | 3.77 | 1.76 | 0.15 | 0.016 | 0.024 | 0.068 | 0.002 |
| 3 | 004# | 3.76 | 1.78 | 0.14 | 0.017 | 0.025 | 0.095 | 0.002 |
| 4 | 005# | 3.80 | 1.77 | 0.14 | 0.019 | 0.022 | 0.066 | 0.002 |
| 5 | 006# | 3.75 | 1.81 | 0.14 | 0.019 | 0.022 | 0.066 | 0.002 |
| 6 | 007# | 3.80 | 1.72 | 0.13 | 0.017 | 0.023 | 0.050 | 0.002 |
| 7 | 015# | 3.71 | 1.82 | 0.12 | 0.016 | 0.023 | 0.049 | 0.002 |
| 8 | 021# | 3.69 | 1.80 | 0.15 | 0.015 | 0.020 | 0.062 | 0.002 |
| 9 | 002# | 3.70 | 1.71 | 0.14 | 0.018 | 0.022 | 0.035 | 0.002 |
| 10 | 008# | 3.70 | 1.73 | 0.10 | 0.017 | 0.021 | 0.047 | 0.002 |
| 11 | 009# | 3.70 | 1.83 | 0.14 | 0.017 | 0.021 | 0.047 | 0.002 |
| 12 | 010# | 3.69 | 1.69 | 0.16 | 0.017 | 0.022 | 0.052 | 0.002 |
| 13 | 011# | 3.88 | 1.78 | 0.12 | 0.016 | 0.027 | 0.047 | 0.002 |
| 14 | 012# | 3.82 | 1.75 | 0.14 | 0.016 | 0.026 | 0.046 | 0.002 |
| 15 | 013# | 3.67 | 1.84 | 0.15 | 0.015 | 0.021 | 0.067 | 0.002 |
| 16 | 014# | 3.78 | 1.72 | 0.13 | 0.010 | 0.027 | 0.045 | 0.002 |
| 17 | 016# | 3.75 | 1.67 | 0.11 | 0.015 | 0.024 | 0.054 | 0.002 |
| 18 | 018# | 3.73 | 1.73 | 0.14 | 0.014 | 0.021 | 0.059 | 0.002 |
| 19 | 019# | 3.78 | 1.71 | 0.13 | 0.015 | 0.023 | 0.060 | 0.002 |
| 20 | 020# | 3.72 | 1.73 | 0.12 | 0.012 | 0.020 | 0.061 | 0.002 |
| Heat | Test block | Instantaneous inoculant | Test block type | Nodularity grade | Graphite size grade | Pearlite (%) | Hardness (HBW) | Tensile strength (MPa) | Elongation (%) |
|---|---|---|---|---|---|---|---|---|---|
| 1 | 001# | – | Body | 3 | 7 | 5 | 114 | 365 | 17.0 |
| 2 | 003# | – | Body | 3 | 7 | 5 | 121 | 365 | 16.5 |
| 3 | 004# | – | Body | 3 | 7 | 5 | 121 | 375 | 17.5 |
| 4 | 005# | – | Body | 3 | 7 | 5 | 114 | 370 | 16.0 |
| 5 | 006# | – | Body | 3 | 7 | 5 | 121 | 375 | 17.0 |
| 6 | 007# | – | Body | 3 | 7 | 5 | 121 | 375 | 16.5 |
| 7 | 015# | – | Body | 3 | 8 | 5 | 121 | 370 | 13.5 |
| 8 | 021# | – | Body | 3 | 7 | 5 | 121 | 365 | 16.0 |
| 9 | 002# | Inoculant D | Body | 2 | 7 | 5 | 114 | 380 | 16.5 |
| 10 | 008# | Inoculant D | Body | 2 | 7 | 5 | 121 | 370 | 21.0 |
| 11 | 009# | Inoculant D | Body | 2 | 7 | 5 | 114 | 380 | 20.0 |
| 12 | 010# | Inoculant D | Body | 2 | 7 | 5 | 121 | 370 | 18.0 |
| 13 | 011# | Inoculant D | Body | 2 | 7 | 5 | 121 | 375 | 17.5 |
| 14 | 012# | Inoculant D | Body | 2 | 7 | 5 | 114 | 375 | 18.5 |
| 15 | 013# | Inoculant C | Body | 2 | 7 | 5 | 121 | 385 | 18.5 |
| 16 | 014# | Inoculant C | Body | 2 | 7 | 5 | 121 | 370 | 19.0 |
| 17 | 016# | Inoculant C | Body | 2 | 7 | 5 | 121 | 370 | 18.5 |
| 18 | 018# | Inoculant C | Body | 2 | 7 | 5 | 121 | 370 | 17.0 |
| 19 | 019# | Inoculant C | Body | 2 | 7 | 5 | 121 | 370 | 19.0 |
| 20 | 020# | Inoculant C | Body | 2 | 8 | 5 | 121 | 400 | 17.0 |
From the data comparison, I observed a clear difference. When stream inoculation was applied, the nodularity of the body attached test blocks was grade 2, and the graphite nodule size remained stable at grade 7. Without stream inoculation, the nodularity of the body attached test blocks was grade 3, while the graphite nodule size also remained stable at grade 7. This indicates that stream inoculation slows the onset of inoculation fading. In addition, Table 3 shows that with stream inoculation, both the elongation and tensile strength of the wind-power ductile iron casting body specimens increased. The improvement in ductility was especially noticeable in several heats treated with inoculant D and inoculant C.
When no stream inoculation was used, the tensile fracture surface contained holes left by detached graphite nodules and a white heterophase. I examined this white heterophase with EDX, and the result is given in Table 4. The white heterophase had low Fe content and high O content, together with small amounts of Mg, Si, Ca, Y, and Ce. I interpret this phase as an oxide inclusion containing Y and Ce. When stream inoculation was used, the tensile fracture surface exhibited many uneven dimples, and no white heterophase was observed. In general, cleavage steps form because impurities exist in the middle of cleavage regions. The tips of these impurities create microcrack sources, and stress concentration at the microcrack tip causes rapid crack propagation, leading to local transgranular fracture. Without stream inoculation, transgranular fracture may be caused by inclusions of Y- and Ce-rich rare-earth oxides. With stream inoculation, the content of these Y- and Ce-rich rare-earth oxide inclusions is reduced, so the tendency for transgranular fracture is lower, and the mechanical properties of the body test blocks are improved.
| C | O | Mg | Si | Ca | Fe | Y | La | Ce |
|---|---|---|---|---|---|---|---|---|
| 80.28 | 19.22 | 0.08 | 0.02 | 0.07 | 0.25 | 0.06 | 0.00 | 0.02 |
Graphite Nucleation Mechanism
When inoculant C is used as an instantaneous inoculant, its composition contains balanced calcium and cerium. This balance minimizes chill tendency and neutralizes harmful trace elements that prevent graphite nodule formation. Inoculant C also contains small amounts of sulfur and oxygen. When it is added to the liquid iron, sulfur and oxygen react with calcium and cerium to form special Ce-Ca-sulfide-oxide composite compounds. These special compounds form excellent graphite crystallization nuclei in ductile iron casting. They provide a large number of nucleus sites and initial Mg-Si oxides, which together strongly promote graphite nucleation. They also stabilize new nucleus sites. As a result, a high-density graphite nodule distribution is obtained, chill tendency is effectively eliminated, and the formation of carbides and shrinkage porosity is reduced. The use of inoculant C effectively increases and refines the ferrite matrix and improves the ductility of ferritic ductile iron casting.
When inoculant D is used as an instantaneous inoculant, its composition contains a relatively high average content of the rare-earth element Y. When added to liquid iron, Y oxidizes to Y2O3. Y2O3 has a hexagonal crystal structure and can act as a heterogeneous nucleus for graphite precipitation. Because the melting point of Y2O3 is as high as 2690 °C, it remains in the liquid iron for a long time and can become an effective nucleus. It therefore has a certain resistance to graphite fading, and the nodularity is high, the graphite morphology is good, and a favorable graphite nodule distribution can be obtained. However, in the later stage of liquid iron solidification, its nucleation ability is inferior to that of inoculant C. It cannot produce a high-density graphite nodule distribution. It also has no obvious effect on the ferrite matrix, and it cannot refine grains or increase the ferrite matrix.
| Feature | Inoculant C | Inoculant D |
|---|---|---|
| Main active elements | Balanced Ca and Ce; small amounts of S and O | Relatively high Y |
| Nucleus type | Ce-Ca-sulfide-oxide composite; initial Mg-Si oxides | Y2O3 with hexagonal structure |
| Nucleus stability | Strong; stable new nucleus sites | High melting point; long residual time |
| Graphite distribution | High-density graphite nodules | Good nodularity, but lower density in later stage |
| Chill tendency | Effectively reduced | Reduced, but less effective late |
| Carbide and shrinkage | Reduced | Limited reduction |
| Ferrite matrix | Increased and refined | No obvious effect |
| Anti-fading ability in late stage | Strong | Weaker than C |
I also express the average nodule count and average nodularity in quantitative terms. If the number of graphite nodules counted in a field is \(n\) and the field area is \(A\), then the nodule count per unit area is
$$N = \frac{n}{A}$$
If the area occupied by nodular graphite is \(A_{\text{nodule}}\) and the total graphite area is \(A_{\text{total}}\), then the nodularity can be estimated as
$$N_{g} = \frac{A_{\text{nodule}}}{A_{\text{total}}} \times 100\%$$
These simple relationships helped me compare the effectiveness of different inoculation treatments in my ductile iron casting trials. The practical result was that inoculant C produced a more stable and long-acting nucleation effect, especially in the later stage of solidification, while inoculant D provided good initial nodularity but weaker late-stage nucleation.
Effect of Silicon on Low-Temperature Impact Toughness
To study the influence of Si on the low-temperature impact toughness of wind-power ductile iron casting, I conducted comparative trials with different Si contents. The main raw materials, nodulizer, and inoculant composition were kept consistent. The process was as follows: the tapping temperature was approximately 1500 °C. On one side of the pocket dam in the ladle, I first placed the nodulizer. Then I covered the nodulizer with inoculant. Finally, I covered the inoculant with a layer of silicon steel sheet. These three layers were used mainly to avoid an excessively fast spheroidization reaction. During tapping, stream inoculation was performed on the furnace platform. After spheroidization, the slag was quickly removed, and the pouring temperature was controlled between 1320 °C and 1340 °C. During pouring, instantaneous stream inoculation was also applied.
I tested 10 heats, labeled 028#–031#, 033#, 034#, 036#, and 050#–052#. The chemical compositions are given in Table 5, and the metallographic structures, tensile properties, and low-temperature impact toughness are given in Table 6. The impact toughness values are reported as individual measurements and as an average. The average impact toughness was calculated as
$$\bar{a}_{k} = \frac{1}{n}\sum_{i=1}^{n} a_{k,i}$$
where \(\bar{a}_{k}\) is the average impact toughness, \(a_{k,i}\) is the individual impact toughness value, and \(n\) is the number of measurements.
| Heat | Test block | C | Si | Mn | S | P | Mg residual |
|---|---|---|---|---|---|---|---|
| 1 | 028# | 3.78 | 1.75 | 0.13 | 0.015 | 0.023 | 0.069 |
| 2 | 029# | 3.73 | 1.70 | 0.15 | 0.015 | 0.015 | 0.044 |
| 3 | 030# | 3.76 | 2.00 | 0.14 | 0.014 | 0.021 | 0.091 |
| 4 | 031# | 3.74 | 1.71 | 0.14 | 0.015 | 0.023 | 0.052 |
| 5 | 033# | 3.75 | 1.74 | 0.13 | 0.013 | 0.021 | 0.056 |
| 6 | 034# | 3.72 | 1.67 | 0.13 | 0.013 | 0.028 | 0.058 |
| 7 | 036# | 3.72 | 1.71 | 0.14 | 0.017 | 0.023 | 0.062 |
| 8 | 050# | 3.69 | 1.82 | 0.13 | 0.018 | 0.029 | 0.047 |
| 9 | 051# | 3.74 | 1.87 | 0.14 | 0.018 | 0.024 | 0.038 |
| 10 | 052# | 3.73 | 1.61 | 0.14 | 0.018 | 0.025 | 0.032 |
| Heat | Test block | Type | Nodularity/Graphite | Pearlite (%) | Hardness (HBW) | Tensile strength (MPa) | Elongation (%) | Impact toughness (J/cm²) |
|---|---|---|---|---|---|---|---|---|
| 1 | 028# | Body | 2/7 | 5 | 121 | 370 | 19.0 | 15.0/13.0/11.0, avg 13.0 |
| 2 | 029# | Body | 2/7 | 5 | 121 | 365 | 18.5 | 17.0/15.0/14.0, avg 15.3 |
| 3 | 030# | Body | 2/7 | 5 | 121 | 385 | 19.5 | 13.0/11.0/11.0, avg 11.7 |
| 4 | 031# | Body | 2/7 | 5 | 121 | 370 | 19.0 | 17.0/14.0/14.0, avg 15.0 |
| 5 | 033# | Body | 2/7 | 5 | 121 | 365 | 19.0 | 13.5/13.0/20.0, avg 15.5 |
| 6 | 034# | Body | 2/7 | 5 | 121 | 370 | 20.0 | 16.5/14.0/15.0, avg 15.2 |
| 7 | 036# | Body | 2/7 | 5 | 121 | 375 | 18.0 | 15.5/14.0/13.5, avg 14.3 |
| 8 | 050# | Body | 2/7 | 5 | 121 | 380 | 17.5 | 14.0/12.0/13.0, avg 13.0 |
| 9 | 051# | Body | 2/7 | 5 | 121 | 385 | 17.5 | 12.0/13.0/11.0, avg 12.0 |
| 10 | 052# | Body | 2/7 | 5 | 121 | 365 | 19.0 | 15.0/16.0/16.0, avg 15.7 |
From Table 5 and Table 6, I found that when the Si content of the ductile iron casting was between 1.6% and 2.0%, the average low-temperature impact toughness of the body attached test blocks generally decreased as the Si content increased. This means that within the range of 1.6%–2.0% Si, the low-temperature impact performance of the wind-power ductile iron casting decreases with increasing Si content. Therefore, an appropriate reduction of Si content helps ensure the low-temperature impact toughness of the ductile iron casting. In my trials, the highest Si content, 2.00%, produced the lowest average impact toughness of 11.7 J/cm², while lower Si contents around 1.61%–1.71% generally produced higher average values, around 15.0–15.7 J/cm², although scatter was present.
I can express the general trend as an empirical decreasing relationship:
$$\bar{a}_{k}(Si) = \bar{a}_{k,0} – k_{Si}(Si – Si_{0})$$
where \(\bar{a}_{k}(Si)\) is the average impact toughness at a given Si content, \(\bar{a}_{k,0}\) is a reference average impact toughness, \(Si_{0}\) is a reference Si content, and \(k_{Si}\) is a positive coefficient describing the loss of toughness per unit increase in Si. In my data, \(k_{Si}\) was positive when Si was varied from 1.6% to 2.0%, confirming the downward trend.
Discussion
My results show that stream inoculation is essential for stable production of wind-power ductile iron casting. The slow cooling and long solidification time of large wind-power castings create favorable conditions for spheroidization and inoculation fading. When stream inoculation is absent, the nodularity grade drops to 3, even though the graphite size may remain at grade 7. When stream inoculation is present, the nodularity grade improves to 2 and the graphite size remains stable at grade 7. This improvement is accompanied by higher tensile strength and elongation. The fracture analysis supports this finding: without stream inoculation, Y- and Ce-rich oxide inclusions act as microcrack sources and promote transgranular fracture; with stream inoculation, these inclusions are reduced, and the fracture surface shows more dimples, indicating better ductility.
The nucleation mechanism explains why inoculant C is superior for late-stage fading resistance. Inoculant C contains balanced Ca and Ce, and it also contains small amounts of S and O. These elements react to form Ce-Ca-sulfide-oxide composite nuclei. These composite nuclei are stable and long-acting, and they provide many sites for graphite nucleation. They also help stabilize new nucleus sites. As a result, the graphite nodule count increases, the nodule distribution becomes dense, carbides and shrinkage porosity are reduced, and the ferrite matrix is increased and refined. In contrast, inoculant D relies on Y2O3 nuclei. Y2O3 has a high melting point and remains in the liquid iron for a long time, so it provides good initial nodularity. However, in the later stage of solidification, its nucleation ability is weaker than that of the Ce-Ca-sulfide-oxide system. It does not produce the same high-density graphite distribution and does not refine the ferrite matrix as effectively.
I also confirmed that Si content must be controlled carefully in wind-power ductile iron casting. Although Si is a strong graphitizing element and helps prevent chill, excessive Si reduces low-temperature impact toughness. In my trials, Si in the range of 1.6%–2.0% showed a clear negative effect on average impact toughness as Si increased. Therefore, I recommend keeping Si in the lower part of the specification range when low-temperature toughness is a critical requirement. This must be balanced against the need for sufficient graphitization and nodularity. The carbon equivalent formula
$$CE = C + \frac{Si + P}{3}$$
is useful for this balance because it allows the caster to adjust C and Si together while maintaining a suitable CE for spheroidization and feeding.
Based on my trials, I recommend the following process window for wind-power ductile iron casting. The carbon content after spheroidization should be 3.1%–3.7%. The carbon equivalent should be 4.1%–4.5%. The silicon content should be 1.5%–1.9%, with lower values preferred for low-temperature impact toughness. Mn should be ≤0.3%, P ≤0.03%, S ≤0.015%, residual RE 0.01%–0.03%, and residual Mg 0.03%–0.05%. The tapping temperature should be about 1500 °C. The spheroidization temperature should be 1460–1480 °C when stream inoculation is used. The pouring temperature should be 1330–1350 °C with stream inoculation, or 1320–1340 °C without stream inoculation. Instantaneous stream inoculation with inoculant C or D should be 0.1%–0.2%. The sandwich method should include 1.2%–1.5% nodulizer, 0.002%–0.01% Sb alloy, 0.6%–1.0% inoculant E, and a silicon steel sheet cover. This process window gave me the best combination of nodularity, graphite size, tensile strength, elongation, and low-temperature impact toughness.
| Parameter | Recommended range | Purpose |
|---|---|---|
| C after spheroidization | 3.1%–3.7% | Promote graphite formation and feeding |
| CE | 4.1%–4.5% | Balance graphitization and chill |
| Si | 1.5%–1.9% | Control graphitization; lower for toughness |
| Mn | ≤0.3% | Limit pearlite and segregation |
| P | ≤0.03% | Reduce embrittlement |
| S | ≤0.015% | Control nodulizer consumption |
| Residual RE | 0.01%–0.03% | Support spheroidization |
| Residual Mg | 0.03%–0.05% | Ensure nodularity |
| Tapping temperature | ~1500 °C | Ensure dissolution and reaction |
| Spheroidization temperature with stream | 1460–1480 °C | Control reaction rate |
| Pouring temperature with stream | 1330–1350 °C | Improve filling and inoculation |
| Pouring temperature without stream | 1320–1340 °C | Use graphitizing expansion for feeding |
| Instantaneous stream inoculant | 0.1%–0.2% C or D | Delay fading |
| Sandwich treatment | 1.2%–1.5% nodulizer, 0.002%–0.01% Sb, 0.6%–1.0% E, Si steel sheet | Control spheroidization |
Conclusions
1. In my trials, ductile iron casting with stream inoculation had better overall quality. Under the same experimental conditions, stream inoculation effectively slowed inoculation fading, raised the nodularity rate to grade 2, kept graphite nodules finer and more stable at grade 7, and significantly improved tensile strength, elongation, and low-temperature impact toughness.
2. Inoculant C was the better choice for wind-power ductile iron casting. It contains balanced Ca and Ce, which form Ce-Ca-sulfide-oxide compounds and initial Mg-Si oxides. These act as stable, long-acting crystallization nuclei with strong resistance to spheroidization and inoculation fading. They produce high nodularity, stable graphite nodule size, and a high-density graphite nodule distribution. They also eliminate chill tendency and reduce carbides, shrinkage cavities, shrinkage porosity, and coarse grains. Inoculant D contains more Y, and although it provides good initial nodularity, its late-stage anti-fading and nucleation ability is weaker than that of inoculant C. It does not produce the same high-density graphite distribution and does not refine the ferrite matrix or increase the ferrite content as effectively.
3. Si content should be controlled to improve toughness. When Si is in the range of 1.6%–2.0%, the low-temperature impact toughness of wind-power ductile iron casting decreases as Si content increases. Therefore, an appropriate reduction of Si content helps ensure low-temperature impact toughness. This finding is important for wind-power applications in cold climates, where low-temperature toughness is a key service requirement.
4. The combination of stream inoculation, appropriate nodulizer and inoculant selection, and tight control of C, Si, CE, residual Mg, and residual RE provides a practical route to stable production of high-quality wind-power ductile iron casting. The Ce-Ca-sulfide-oxide nuclei formed by inoculant C are especially effective for late-stage nucleation, while the Y2O3 nuclei from inoculant D are less effective in the later stage of solidification. By applying these findings, I was able to improve the uniformity of nodularity, graphite size, tensile properties, and low-temperature impact toughness in my production trials.
5. In future work, I would further quantify the relationship between nodule count, nodularity, and impact toughness using image analysis, and I would refine the empirical coefficients for the Si effect on low-temperature impact toughness. I would also examine the interaction between chills, wall thickness, and inoculation fading in large wind-power ductile iron casting, because these factors determine the local cooling rate and therefore the local nucleation and growth behavior of graphite. The results of such work would help make ductile iron casting for wind power even more reliable and cost-effective.