As industrialization progresses, ductile iron has become a primary metallic material in industrial production due to its excellent comprehensive properties. The typical ferritic matrix grade QT450-10 is widely used in metallurgical machinery, petrochemicals, pipeline industry, and transportation. However, the increasingly diverse service environments demand higher performance from ductile iron. Compared with mixed-matrix ferritic ductile iron, silicon solid-solution strengthened ferritic ductile iron exhibits superior strength, toughness, and machinability, driving growing research interest in high-silicon ferritic ductile iron. In our study, we focus on the effect of silicon content on the microstructure and mechanical properties of high-silicon ductile iron produced by the lost foam casting process. The lost foam casting process differs from traditional sand casting in that the gasification of the foam pattern removes some heat, making control of pouring temperature critical. Silicon content also affects the eutectic temperature and consequently the matrix structure. Therefore, obtaining a fully ferritic high-silicon ductile iron via lost foam castings is of great significance. We prepared high-silicon ductile iron with silicon content ranging from 2.9% to 4.6% using the lost foam casting process and investigated its microstructure and mechanical properties, analyzing the mechanism by which silicon influences the matrix and performance, providing a theoretical basis for large-scale high-silicon ductile iron pipe fittings produced by lost foam castings.
1. Experimental Materials and Methods
To study the effect of different silicon contents on the microstructure and properties of ductile iron, we designed one set of QT450-10 ductile iron samples as the control group and three sets of high-silicon ductile iron samples with silicon content in the range of 3.5%–4.5% as the experimental group. According to the carbon equivalent formula:
$$\omega(CE)\% = \omega(C)\% + 0.3[\omega(Si)\% + \omega(P)\%]$$
We maintained the carbon equivalent approximately constant; as silicon content increased, carbon content decreased accordingly. Melting was conducted in a 1 t medium-frequency induction furnace, and pouring was performed using the lost foam casting process. The foam pattern was made of EPS material, the coating mainly consisted of bauxite, and the molding sand was dry silica sand. The pouring temperature was 1,487 °C, negative pressure 0.05 MPa, pouring time 30 s. Nodularization was carried out using the wire feeding method with 1.3% nodulizer, and inoculation used 75 ferrosilicon at 1.5%–2.5%. The wall thickness of the Y-shaped foam pattern was 24 mm, and the dimensions followed the national standard GB/T 1348-2019. The final chemical compositions of the four groups are presented in Table 1.
| Sample | C | Si | Mn | P | S | Mg | RE | CE |
|---|---|---|---|---|---|---|---|---|
| A | 3.52 | 2.92 | 0.183 | 0.029 | 0.0058 | 0.064 | 0.0160 | 4.4 |
| B | 3.09 | 3.68 | 0.273 | 0.027 | 0.0057 | 0.068 | 0.0155 | 4.2 |
| C | 2.88 | 4.22 | 0.311 | 0.026 | 0.0037 | 0.059 | 0.0155 | 4.15 |
| D | 2.60 | 4.59 | 0.265 | 0.024 | 0.0046 | 0.057 | 0.0145 | 3.98 |
In production, the silicon content of as-cast QT450-10 ductile iron is controlled within 2.5%–3.0%; sample A with 2.92% Si meets the requirement. According to the Fe-C phase diagram, under equilibrium conditions, the eutectic composition occurs at a carbon equivalent of 4.34%, but silicon shifts the eutectic point, so we controlled the carbon equivalent in the range of 4.0%–4.4%. The levels of Mn, P, S, Mg, etc., were similar across all four groups. Thus, the chemical compositions of all cast samples satisfy the experimental design.
As-cast samples were machined into 10 mm × 10 mm × 10 mm metallographic specimens, ground, polished, and etched with 4% nitric acid alcohol for 10–15 s. Microstructure observation was performed using an Olympus-DSX500 optical microscope (OM) and a ZEISS Ultra Plus field emission scanning electron microscope (SEM). Image analysis software (ipp-6.0 and OLYCIA.m3) was used to measure ferrite and graphite content and size, and graphite morphology was graded according to GB/T 9441-2021. Macrohardness was measured using a digital Brinell hardness tester (750 N load) and a Vickers hardness tester (300 N load). Tensile tests were conducted on an AG-XPLUS universal testing machine with specimens having a parallel section diameter of 14 mm, gauge length of 70 mm, and total length of 180 mm.
2. Results and Discussion
2.1 Effect of Silicon Content on Microstructure of High-Silicon Ductile Iron in Lost Foam Castings
The graphite morphology of ductile iron samples with different silicon contents is shown in Figure 1 (the image of graphite morphology is provided below). According to GB/T 9441-2021, the nodularity of all samples was above 90%, and the roundness of graphite remained essentially unchanged. Using ipp-6.0, we statistically analyzed the graphite morphology; the results are listed in Table 2. For sample A with 2.92% Si, the average number of graphite nodules per unit area was about 151 nodules/mm², and the average graphite nodule diameter was about 35.6 μm. For sample D with 4.59% Si, the average number of graphite nodules was about 223 nodules/mm², and the average diameter was about 26 μm. With increasing silicon content, the graphite nodules became finer and more numerous. Silicon acts as a nucleation site for graphite, and higher silicon content increases the temperature difference between the stable and metastable eutectic ranges, favoring graphite formation. Additionally, with carbon equivalent held constant, higher silicon means lower carbon, reducing the overall graphite volume. Therefore, the graphite nodules in high-silicon ferritic ductile iron produced by lost foam castings become finer and more uniform.
| Sample | Si (wt%) | Graphite nodule count (nodules/mm²) | Graphite nodule diameter (μm) |
|---|---|---|---|
| A | 2.92 | 151 | 35.6 |
| B | 3.68 | 165 | 30.8 |
| C | 4.22 | 208 | 27.7 |
| D | 4.59 | 223 | 26.0 |
The matrix microstructure of the samples is illustrated in the figure. All samples, including the QT450-10 with 2.92% Si and the high-silicon ones, consist entirely of ferrite. From a thermodynamic perspective, higher silicon raises the eutectoid transformation temperature, widening the temperature range between the stable and metastable eutectoid reactions, which favors ferrite formation. From a kinetic viewpoint, during solidification of lost foam castings, the austenite formed at high temperature transforms upon cooling to the eutectoid temperature. Because high-silicon ductile iron has densely distributed graphite, the distance between austenite and graphite is shortened, allowing carbon in austenite to easily diffuse to the eutectic graphite. Once carbon diffuses out, ferrite nuclei readily form at the austenite interface, promoting ferrite formation. We measured the ferrite content and grain size (Table 3). For sample A, the average ferrite grain size was about 43.77 μm; for sample D, it was about 35.38 μm. As silicon increased, the ferrite grain refined because silicon raises the eutectic temperature, increasing the undercooling at the solidification front and enhancing the primary nucleation rate of ferrite.
| Sample | Si (wt%) | Ferrite content (%) | Average ferrite grain size (μm) |
|---|---|---|---|
| A | 2.92 | 88.8 | 43.766 |
| B | 3.68 | 89.9 | 42.064 |
| C | 4.22 | 90.0 | 38.630 |
| D | 4.59 | 91.7 | 35.381 |
2.2 Effect of Silicon Content on Mechanical Properties of High-Silicon Ferritic Ductile Iron in Lost Foam Castings
Tensile strength and Brinell hardness are key indicators for evaluating ductile iron performance. The mechanical properties of the as-cast samples are summarized in Table 4. Sample A (2.92% Si) exhibited a tensile strength of 441 MPa, yield strength of 325 MPa, elongation of 17%, and hardness of 128 HBW. Sample D (4.59% Si) showed a tensile strength of 683 MPa, yield strength of 488 MPa, elongation of 17.5%, and hardness of 186.1 HBW. With increasing silicon content, the strength and hardness of the high-silicon ductile iron increased, and the yield ratio also increased. The yield ratio (yield strength/tensile strength) indicates the material’s resistance to plastic deformation; a higher ratio is beneficial for saving material and reducing cost. The finer graphite nodules reduce the stress concentration on the metal matrix, and the increased silicon content raises the eutectic temperature, increasing the undercooling at the solidification front, thus refining the ferrite grains. More grain boundaries impede dislocation motion, enhancing strength, hardness, and elongation. However, as silicon atoms form Cottrell atmospheres around dislocations, the plastic deformation capacity decreases, so elongation remains essentially unchanged under the combined effect. The scanning electron microscopy fractographs of the tensile specimens are described (without referencing specific figure numbers): at 2.92% Si, the fracture surface consisted mainly of dimples, indicating ductile fracture; at 3.68% Si, fewer but deeper dimples appeared, still ductile but with some transgranular fracture; at 4.22% Si, large cleavage facets dominated, indicating brittle fracture; at 4.59% Si, dimples were few and cleavage facets and intergranular fracture appeared. The strong solid-solution strengthening effect of silicon causes lattice distortion in ferrite, reducing its plastic deformation ability. Thus, the fracture mode transitions from ductile to brittle with increasing silicon, but since the matrix is fully ferritic and graphite is fine and uniform, the high-silicon ductile iron still exhibits good macroscopic plasticity.
| Sample | Si (wt%) | Tensile strength (MPa) | Yield strength (MPa) | Yield ratio | Elongation (%) | Hardness (HBW) |
|---|---|---|---|---|---|---|
| A | 2.92 | 441 | 325 | 0.737 | 17 | 128 |
| B | 3.68 | 527 | 425 | 0.806 | 18 | 149 |
| C | 4.22 | 601 | 512 | 0.852 | 17 | 174.6 |
| D | 4.59 | 683 | 588 | 0.861 | 17.5 | 186.1 |
3. Mechanism of Silicon Influence on Microstructure and Properties of High-Silicon Ductile Iron in Lost Foam Castings
To understand the mechanism, we measured the microhardness of the ferrite matrix, analyzed the element distribution by SEM-EDS, and determined the lattice constant of ferrite by X-ray diffraction. The Fe-Si phase diagram under equilibrium shows that in the silicon range of 2.9%–4.6%, only the body-centered cubic (BCC) ferrite phase exists, and silicon is completely soluble in ferrite as substitutional atoms. The atomic radius of silicon is 0.118 nm, while the largest interstitial site in BCC ferrite is 0.0766 nm (0.633R), so silicon cannot form an interstitial solid solution. Since the atomic radius of iron is 0.121 nm, slightly larger than silicon, silicon substitutes for iron in the lattice. Table 5 lists the ferrite microhardness. Sample A had an average ferrite microhardness of 186.16 HV, while sample D reached 253.17 HV. The increase in microhardness with silicon content indirectly confirms the enhanced solid-solution strengthening effect. Variations in hardness at different locations within the same sample indicate segregation of silicon in the ferrite matrix.
| Sample | Si (wt%) | Ferrite microhardness (HV) – individual measurements | Average (HV) |
|---|---|---|---|
| A | 2.92 | 180.88, 187.47, 188.17, 187.48, 186.79 | 186.16 |
| B | 3.68 | 230.59, 223.68, 211.63, 222.72, 227.58 | 223.24 |
| C | 4.22 | 237.59, 244.34, 230.51, 229.08, 249.87 | 238.28 |
| D | 4.59 | 244.62, 245.65, 266.16, 259.36, 249.78 | 253.17 |
We performed EDS point analysis on ferrite grain interiors and grain boundaries. For sample A (2.92% Si), the segregation index \( K_{Si} \) at grain boundaries was 0.72 and 0.68, while inside grains it was 1.01 and 1.04. For sample D (4.59% Si), the segregation index at grain boundaries was 1.12 and 1.09, and inside grains it was 0.83 and 0.91. The segregation index is defined as:
$$ K_{Si} = \frac{w_{Si}}{w_{Si0}} $$
where \( w_{Si} \) is the local silicon content at the measurement point, and \( w_{Si0} \) is the average silicon content in the ductile iron. At low silicon content, silicon tends to segregate inside grains; at high silicon content, it segregates more strongly at grain boundaries. This segregation increases lattice distortion, raising strength and hardness while reducing plasticity and toughness. However, because the graphite nodules are fine and the matrix is fully ferritic, the elongation decreases only slightly.
X-ray diffraction patterns of the samples were analyzed. Only ferrite peaks corresponding to (110), (200), and (211) planes were observed, confirming that silicon is dissolved in the ferrite lattice. The lattice constant \( a \) of the BCC ferrite was calculated using the formula for cubic crystals when \( h+k+l \) is even:
$$ d = \frac{a}{\sqrt{h^2 + k^2 + l^2}} $$
where \( d \) is the interplanar spacing. Table 7 summarizes the calculated lattice constants. For sample A (2.92% Si), the average lattice constant was 0.28647 nm; for sample D (4.59% Si), it was 0.28602 nm. Compared with the lattice constant of pure iron (0.28664 nm), the distortion degree increased from 0.06% to 0.22%. The decreasing lattice constant with increasing silicon content indicates enhanced lattice distortion, which increases dislocation resistance and thus improves strength and hardness.
| Si (wt%) | a(110) (nm) | a(200) (nm) | a(211) (nm) | a(AVG) (nm) | Distortion (%) |
|---|---|---|---|---|---|
| 2.92 | 0.28628 | 0.28655 | 0.28659 | 0.28647 | 0.06 |
| 3.68 | 0.28642 | 0.28612 | 0.28620 | 0.28625 | 0.14 |
| 4.22 | 0.28607 | 0.28638 | 0.28608 | 0.28618 | 0.16 |
| 4.59 | 0.28606 | 0.28600 | 0.28600 | 0.28602 | 0.22 |
4. Conclusions
We successfully produced high-silicon ductile iron with silicon content ranging from 2.9% to 4.6% using the lost foam casting process. The matrix structure was fully ferritic. When the silicon content was 2.92%, the average graphite nodule diameter was about 35.6 μm, the number of graphite nodules per unit area was about 151 nodules/mm², and the average ferrite grain size was about 43.77 μm. When the silicon content was 4.59%, the average graphite nodule diameter decreased to about 26 μm, the nodule count increased to about 223 nodules/mm², and the ferrite grain size refined to about 35.38 μm. The mechanical properties improved significantly: at 2.92% Si, tensile strength was 441 MPa, Brinell hardness 128 HBW, and elongation 17%; at 4.59% Si, tensile strength reached 683 MPa, hardness 186.1 HBW, and elongation 17.5%, indicating excellent comprehensive mechanical properties. For silicon content in the range of 3.5%–4.6%, the increase in silicon content refined the graphite nodules, reducing the stress concentration on the metal matrix. The ferrite grain size decreased, leading to fine-grain strengthening. Additionally, silicon atoms substituted for iron atoms in the ferrite lattice, forming a substitutional solid solution. With increasing silicon content, the segregation of silicon at ferrite grain boundaries became more pronounced, causing greater lattice distortion and reducing the lattice constant. The distortion degree increased from 0.06% to 0.22%. These combined mechanisms—refined graphite, fine grain size, and solid-solution strengthening—contributed to the enhancement of strength and hardness while maintaining good elongation in high-silicon ductile iron produced by lost foam castings.