In recent years, research in the field of ductile iron casting has gradually shifted towards large-section, super-large-section, and lightweight applications. Lightweight ductile iron castings have emerged as a hot research topic, successfully opening up new markets for ductile iron. The rationality of casting wall thickness design largely determines the difficulty of the casting process, with performance non-uniformity being one of the primary defects. Therefore, precise control of the solidification process is highly demanded, an issue that has garnered widespread attention in academia. With the rapid development of the wind power industry, the structural design of wind turbine components has become increasingly complex, leading to widespread occurrences of uneven wall thickness, which poses significant challenges in production and application. To address this, this study focuses on investigating the variations in microstructure and mechanical properties of ductile iron casting stepped specimens poured under the same process. Through experimentation and analysis, the influence of uneven wall thickness on the microstructure and mechanical properties of castings is derived. These findings not only contribute to improving the performance and quality of wind power castings but also promote the sustainable development of the wind power industry.
The key objective of this experiment is to analyze the impact of casting wall thickness variations on microstructure and properties. The chemical composition of the iron melt for the test castings is presented in Table 1.
| Element | Content (wt%) |
|---|---|
| C | 3.80 |
| Si | 2.10 |
| Mn | <0.25 |
| P | <0.035 |
| S | ≤0.020 |
| Mg | 0.035–0.045 |
| Re | <0.025 |
| Sb | 0.006 |
An induction furnace was used for melting the iron melt. When the iron melt temperature reached the predetermined range of 1420–1480°C, tapping was performed. To ensure nodulization, the sandwich method was employed for spheroidization treatment. Specifically, 0.9–1.05% nodulizer was added to the bottom of a preheated ladle pocket, covered with a layer of steel scrap, and then topped with 0.15–0.50% primary inoculant. After spheroidization, slag removal was conducted to ensure melt purity. At a suitable temperature range of 1350–1370°C, 0.20% Si-Al inoculant was added during pouring, following the gating system design. Each step dimension was 200 mm × 200 mm × t mm, where t represents the thickness of each step.
Prior to cutting samples, non-destructive testing was performed on the poured stepped blocks to ensure comprehensive sampling and data integrity. Due to the large size of the specimens, sampling considered comprehensiveness and representativeness. Referring to MAGMA simulation solidification modulus, sampling positions were set symmetrically on both sides of the axis. Each layer included 6 impact specimens, 7 metallographic specimens, and 1 tensile specimen, as detailed in the sampling distribution diagram.
Metallographic samples were collected and identified according to the specified steps, with images arranged from surface to core and from lower to upper steps. Observations revealed that as the step height increased, the diameter of graphite spheres in the upper layers tended to enlarge. This phenomenon is primarily due to increased section thickness slowing solidification, allowing more time for graphite flotation and growth. At step junctions, graphite size was grade 5, while in other areas, it was grade 6. Graphite at junctions exhibited larger diameters and fewer graphite nodules. Analysis combined with simulation indicated that areas远离热节点 cooled faster, promoting nucleation of smaller-diameter core particles and resulting in more graphite nodules. In contrast, graphite near hot spots remained液态 for longer, leading to growth and flotation, reducing nodule count and increasing diameter. Thus, the design of wall thickness transitions significantly influences graphite morphology distribution.
On each step of the trapezoidal specimen, three layers of samples were cut according to the layout. These were observed under a metallographic microscope at 100x magnification across the entire inspected surface. From the worst areas, three consecutive fields were examined, and nodularity and nodule count were measured using an image analyzer, with results shown in Figure 5. Average nodularity grades mainly ranged between grades 4 and 5. For step thicknesses less than 170 mm, fluctuations in nodularity from surface to core were relatively consistent. However, when step thickness exceeded 170 mm, the fluctuation amplitude increased. Graphite nodule counts showed larger fluctuations from surface to core for thicknesses below 205 mm, but decreased significantly beyond 205 mm. Comprehensive analysis indicated that both nodularity and graphite nodule count trends initially decreased and then increased. Notably, graphite nodule count variations were relatively large. Therefore, when specific requirements for graphite nodule count exist, excellent molding processes must be paired to shorten solidification time and achieve targets.
The relationship between graphite nodule count (N) and wall thickness (t) can be approximated by the following empirical formula, which captures the initial decrease and subsequent increase:
$$ N(t) = a \cdot t^2 + b \cdot t + c $$
where a, b, and c are constants derived from experimental data. For instance, based on observations, a negative value for a and positive for b might fit the trend where N decreases initially and then rises after a certain thickness.
Metallographic specimens were corroded with 4% nitric alcohol for 15 seconds, rinsed with anhydrous ethanol, dried, and observed under a 100x metallographic microscope. The microstructure consisted mainly of spheroidal graphite, ferrite, and pearlite. White areas represented ferrite, while dark areas indicated pearlite. From steps a to h, pearlite content did not show a clear trend with increasing step thickness. The matrix structure can be described by the phase fraction equation:
$$ V_f = V_{\text{ferrite}} + V_{\text{pearlite}} + V_{\text{graphite}} $$
where $V_f$ is the total volume fraction, and the proportions vary with solidification conditions.
Tensile test bars were cut from three layers of each step, and data is presented in Figure 7. Tensile strength fluctuated between 365 MPa and 415 MPa with increasing step thickness, but overall variation was small. This is attributed to the use of ferritic matrix ductile iron casting. As step thickness increased, changes in graphite nodule count affected carbide formation tendencies, but primary cementite was not eliminated, resulting in minor strength variations. Elongation fluctuated between 15% and 26%, with larger amplitudes, possibly due to effects on grain size and homogeneity. Thus, step thickness changes have a relatively small impact on tensile strength but a larger effect on elongation. To improve elongation uniformity, structural design and molding processes must work synergistically to achieve microstructural homogeneity.
The tensile strength ($\sigma_t$) can be modeled as a function of wall thickness (t) and graphite parameters:
$$ \sigma_t = \sigma_0 – k_1 \cdot \frac{1}{N} + k_2 \cdot d_g $$
where $\sigma_0$ is the base strength, $k_1$ and $k_2$ are material constants, N is the graphite nodule count, and $d_g$ is the average graphite diameter.
| Wall Thickness (mm) | Tensile Strength (MPa) | Elongation (%) | Impact Energy (J) | Hardness (HBW) |
|---|---|---|---|---|
| 100 | 395 | 22 | 14.5 | 165 |
| 135 | 385 | 18 | 12.0 | 160 |
| 170 | 375 | 16 | 10.5 | 155 |
| 205 | 370 | 15 | 9.5 | 150 |
| 240 | 380 | 20 | 11.0 | 145 |
| 275 | 390 | 24 | 13.5 | 140 |
| 310 | 400 | 26 | 15.0 | 135 |
| 345 | 410 | 25 | 16.5 | 130 |
Low-temperature impact tests were conducted at -20°C on Charpy impact specimens, with three layers cut per step. Results from the impact testing machine are shown in Figure 8. As step thickness increased, average impact energy first decreased and then increased, with a turning point at 305 mm, consistent with simulation results. Solidification speed affects the microstructure of ductile iron casting, thereby influencing impact performance. Impact values for the entire stepped specimen fluctuated between 9 J and 16.5 J, with relatively large amplitudes. Thus, improving impact toughness requires favorable solidification speeds to achieve excellent metallographic structures.
The impact energy (E) can be related to solidification rate (S) and wall thickness (t) by:
$$ E = E_{\text{max}} \cdot \left(1 – e^{-k \cdot S}\right) $$
where $E_{\text{max}}$ is the maximum achievable energy and k is a constant. Solidification rate S is inversely proportional to wall thickness, i.e., $S \propto 1/t$.
Brinell hardness tests were performed on the metallographic specimens, with three points measured per sample and averages taken. Results are shown in Figure 9. Hardness values trended relatively steadily with increasing step thickness, with minor influence. Average hardness values remained above 130 HBW. The hardness trend and influencing factors are crucial for understanding internal structure and mechanical properties. Thus, in material selection and application, knowledge of hardness characteristics is essential.
Hardness (H) can be expressed as a function of pearlite content ($V_p$) and graphite parameters:
$$ H = H_f \cdot (1 – V_p) + H_p \cdot V_p – k_h \cdot d_g $$
where $H_f$ and $H_p$ are hardness contributions from ferrite and pearlite, respectively, and $k_h$ is a constant.
In summary, across all step thicknesses, the matrix structure exhibited F+P characteristics, with graphite morphology primarily types VI and V, nodularity grades 4–5, and graphite sizes 5–6. Increasing step thickness led to relatively small changes in tensile strength due to incomplete elimination of primary cementite. Solidification speed positively influenced impact performance, with faster speeds yielding better results. Brinell hardness gradually decreased with increasing step thickness. Elongation fluctuated significantly with wall thickness increase, likely due to effects on grain size and homogeneity. Therefore, to achieve superior ductile iron casting properties, melting and molding processes must be coordinated, rather than relying on either alone.
The comprehensive analysis of ductile iron casting demonstrates that wall thickness plays a critical role in determining microstructural and mechanical behavior. Future work should focus on optimizing process parameters to minimize variations and enhance performance in applications such as wind power components. The integration of simulation tools with experimental validation can further advance the understanding and control of ductile iron casting properties.