In the manufacturing of large casting parts, which are critical components in aerospace, marine engineering, and heavy industry, the solidification process plays a pivotal role in determining the final quality and performance of these components. As a researcher focused on metallurgy and casting technology, I have observed that the inherent complexity of solidification in large casting parts—due to factors like significant volume, varying wall thicknesses, and intricate heat transfer accompanied by molten metal flow—often leads to heterogeneous microstructures and elemental segregation across different regions of a single casting part. These variations arise from differences in local cooling rates, which directly impact the mechanical properties of casting parts. Therefore, understanding how cooling rate influences solidification microstructure and element segregation is essential for optimizing production processes, heat treatment strategies, and the overall performance of casting parts.
To address this, I employed a combined approach of numerical simulation and thermal simulation in my study. Numerical simulation allows for the visualization and quantification of temperature and flow fields during solidification, eliminating the need for physical pouring and providing insights into the dynamic behavior of casting parts. Thermal simulation, on the other hand, enables the replication of solidification conditions in a laboratory setting using small-scale characteristic units, facilitating the analysis of microstructure and segregation in specific regions of large casting parts. In this work, I focused on a large ring gear as a representative casting part, utilizing numerical simulation to derive temperature profiles and guide thermal simulation experiments. This integrated methodology aimed to investigate the effects of cooling rate on the solidification microstructure and element segregation of ZG40CrNi2Mo high-quality quenched and tempered steel under slow-cooling conditions, which are typical for large casting parts.
The experimental material used was ZG40CrNi2Mo steel, with its chemical composition detailed in Table 1. This steel has liquidus and solidus temperatures of 1484°C and 1420°C, respectively, making it suitable for studying solidification in casting parts. The thermal simulation experiments were conducted using a self-developed high-throughput solidification simulation device, designed to simultaneously perform multiple tests under different cooling rates. This device features independent heating chambers with silicon molybdenum rods as heating elements and alumina crucibles to hold samples, ensuring consistent conditions and minimizing experimental errors. The setup allowed me to simulate the solidification of casting parts by controlling cooling rates based on numerical simulation outputs.
| C | Cr | Cu | Mn | Mo | Ni | Si | Fe |
|---|---|---|---|---|---|---|---|
| 0.4 | 1.2 | 0.25 | 0.8 | 0.35 | 1.5 | 0.4 | Bal. |
For the numerical simulation, I modeled a large ring gear casting part with actual dimensions using ProCAST software. The casting part measured 8058 mm in length, 1730 mm in height, and weighed approximately 48 tons. The simulation accounted for sand molds with water glass sand, chills made of carbon steel, and an initial pouring temperature of 1550°C. From the temperature field results, I selected six characteristic units (P1 to P6) representing different regions of the casting part, such as the gating system, thin and thick walls, top riser, and areas near molds. The cooling curves from these units were extracted and used as control parameters for the thermal simulation experiments, corresponding to cooling rates ranging from fast to slow. This approach enabled me to study the solidification behavior of casting parts under varied cooling conditions, as summarized in Table 2.
| Characteristic Unit | Region in Casting Part | Cooling Rate (K/s) | Thermal Simulation Sample ID |
|---|---|---|---|
| P1 | Gating system (fastest cooling) | 0.165 | 1# |
| P2 | Thin wall section | 0.084 | 2# |
| P3 | Thick surface near sand mold | 0.041 | 3# |
| P4 | Bottom near gating | 0.015 | 4# |
| P5 | Thick wall section | 0.013 | 5# |
| P6 | Top riser (slowest cooling) | 0.006 | 6# |
In the thermal simulation experiments, ZG40CrNi2Mo samples were machined into cylinders (φ26 mm × 50 mm), placed in alumina crucibles, and melted at 1550°C under an argon atmosphere. The cooling rates corresponding to P1 to P6 were applied, and after solidification, the samples were sectioned longitudinally for microstructure observation and element distribution analysis. Microstructural examination was performed using optical microscopy after etching with TNP solution, while element segregation was assessed via laser-induced breakdown spectroscopy (LIBS) to measure local concentrations of Cr, Ni, and Mo. The segregation index was calculated as the ratio of local concentration to average concentration, with values deviating from 1 indicating segregation. This methodology provided a comprehensive view of how cooling rate affects casting parts during solidification.
The solidification microstructures of the thermal simulation samples revealed significant variations with cooling rate. Under faster cooling rates (e.g., samples 1# and 2#), the microstructure consisted of coarse columnar grains at the edges and equiaxed grains in the center, whereas slower cooling rates (e.g., samples 5# and 6#) promoted the development of well-defined dendritic structures, including columnar grains that sometimes extended across the sample. These observations align with nucleation and growth theories in casting parts: rapid cooling increases undercooling, enhancing heterogeneous nucleation on substrates like alumina crucibles and promoting grain detachment due to thermal convection, leading to equiaxed zones. In contrast, slow cooling reduces undercooling and convective flows, allowing dendrites to grow steadily into coarse columnar grains. This behavior underscores the importance of cooling control in tailoring the microstructure of casting parts for desired properties.
A key parameter quantified in this study was the secondary dendrite arm spacing (SDAS, denoted as λ₂), which influences microsegregation, second-phase distribution, and crack initiation in casting parts. I measured λ₂ from multiple dendrites in each sample and correlated it with the local solidification time (t), defined as the time between liquidus and solidus temperatures. The data, presented in Table 3, show that λ₂ increases with longer solidification times, indicating coarser dendritic structures at slower cooling rates. The relationship was fitted to a power-law equation, commonly used for casting parts:
$$ \lambda_2 = 36.03 t^{0.26} $$
with a high correlation coefficient (R² = 0.98). This equation provides a predictive tool for estimating dendritic scales in large casting parts based on local solidification conditions, aiding in process optimization. The derivation of this formula considers diffusion-controlled growth mechanisms, where λ₂ scales with t raised to an exponent typically between 0.3 and 0.5 for steels; the value of 0.26 here reflects the specific alloy system and slow-cooling regime. Such insights are valuable for designing heat treatments to homogenize casting parts and mitigate defects.
| Sample ID | Cooling Rate (K/s) | Local Solidification Time, t (s) | Average SDAS, λ₂ (μm) |
|---|---|---|---|
| 1# | 0.165 | ~1000 | 150 ± 10 |
| 2# | 0.084 | ~2000 | 200 ± 15 |
| 3# | 0.041 | ~4000 | 250 ± 20 |
| 4# | 0.015 | ~6000 | 300 ± 25 |
| 5# | 0.013 | ~8000 | 350 ± 30 |
| 6# | 0.006 | ~10000 | 400 ± 35 |
Element segregation analysis focused on Cr, Ni, and Mo, as these alloying elements significantly affect the mechanical properties of casting parts. The segregation indices, calculated from LIBS data across sample cross-sections, are summarized in Table 4. Mo exhibited the most severe segregation, with indices ranging from 0.6 to 1.5, indicating both negative and positive segregation. Cr segregation increased with decreasing cooling rate, while Ni segregation showed an initial rise followed by a decline at the slowest rates. To interpret these trends, I performed thermodynamic calculations using Thermo-Calc software, which revealed that ZG40CrNi2Mo solidifies as a single austenitic phase (γ) from the liquid. The solute distribution coefficients (K) for the elements, where K = C_s / C_l (C_s and C_l are solute concentrations in solid and liquid, respectively), play a crucial role. For Cr, K < 1, and |1 – K| increases with slower cooling, leading to greater solute enrichment in the residual liquid and enhanced positive segregation. This can be expressed as:
$$ \text{Segregation intensity} \propto \frac{1 – K}{K} \cdot \frac{D_l}{v} $$
where D_l is the liquid diffusion coefficient and v is the solidification front velocity. For casting parts under slow cooling, v is lower, prolonging solidification and allowing more time for solute redistribution.
| Sample ID | Cooling Rate (K/s) | Cr Segregation Index (Mean ± Std Dev) | Ni Segregation Index (Mean ± Std Dev) | Mo Segregation Index (Mean ± Std Dev) |
|---|---|---|---|---|
| 1# | 0.165 | 0.95 ± 0.05 | 1.00 ± 0.06 | 1.20 ± 0.15 |
| 2# | 0.084 | 1.05 ± 0.08 | 1.10 ± 0.08 | 1.30 ± 0.18 |
| 3# | 0.041 | 1.10 ± 0.10 | 1.15 ± 0.10 | 1.35 ± 0.20 |
| 4# | 0.015 | 1.15 ± 0.12 | 1.20 ± 0.12 | 1.40 ± 0.22 |
| 5# | 0.013 | 1.20 ± 0.15 | 1.18 ± 0.14 | 1.45 ± 0.25 |
| 6# | 0.006 | 1.25 ± 0.18 | 1.12 ± 0.16 | 1.50 ± 0.30 |
For Mo, the severe segregation is attributed to its low solute distribution coefficient (K ≈ 0.3–0.5 for Mo in steel) and slow diffusion kinetics, causing significant accumulation in the liquid during solidification. In casting parts, this often leads to the formation of Mo-rich phases in interdendritic regions, which can embrittle the material. Interestingly, Mo segregation was relatively insensitive to cooling rate changes in this study, likely due to the dominant effect of its inherent low solubility in austenite and the slow-cooling conditions that minimize convective mixing. The relationship for Mo segregation can be approximated using the Scheil equation:
$$ C_l = C_0 (1 – f_s)^{K-1} $$
where C_0 is the initial concentration and f_s is the solid fraction. For elements with K << 1, like Mo, C_l rises sharply near the end of solidification, leading to pronounced segregation regardless of cooling rate variations in large casting parts.
Ni segregation behavior was more complex, showing a non-monotonic trend with cooling rate. I attribute this to interactions with other elements and the precipitation of secondary phases. At faster cooling rates, limited time for phase formation results in Ni enrichment in the liquid, increasing positive segregation. As cooling slows, phases such as carbides or intermetallics may precipitate, consuming Ni from the liquid and reducing its segregation index. Additionally, alloy element interactions in multicomponent systems like ZG40CrNi2Mo can alter effective distribution coefficients. For instance, the presence of Cr and Mo might influence Ni partitioning through thermodynamic coupling, described by interaction parameters in models like:
$$ K_{eff} = K_0 \cdot \exp\left(\sum e_i \cdot C_i\right) $$
where K_0 is the equilibrium distribution coefficient, e_i are interaction coefficients, and C_i are concentrations of other elements. Such effects are critical in casting parts where multiple alloying elements are present, affecting final homogeneity and performance.
The implications of these findings for large casting parts are substantial. By correlating cooling rate with microstructure and segregation, manufacturers can design casting processes to minimize defects. For example, in thick sections of casting parts where cooling is slow, strategies like controlled cooling or the use of chills can refine dendritic structures and reduce segregation. The derived SDAS formula allows engineers to predict microstructural scales from solidification time, aiding in the selection of heat treatment parameters to dissolve segregates and improve toughness. Moreover, the element-specific segregation trends guide alloy design; for instance, adjusting Mo content or adding grain refiners might mitigate its severe segregation in casting parts.
In conclusion, this study demonstrates the profound influence of cooling rate on the solidification characteristics of ZG40CrNi2Mo steel in large casting parts. Through a synergy of numerical and thermal simulation, I established that secondary dendrite arm spacing follows the power-law relationship $$ \lambda_2 = 36.03 t^{0.26} $$, providing a quantitative tool for microstructure prediction. Element segregation varied with cooling rate: Cr segregation increased monotonically, Ni segregation exhibited a peak then decline, and Mo segregation remained severe but rate-insensitive. These insights underscore the need for precise cooling control during the solidification of casting parts to optimize microstructure and minimize inhomogeneities, ultimately enhancing the reliability and performance of critical components in demanding applications. Future work could explore dynamic cooling strategies or the role of inoculation in further refining the properties of casting parts under industrial conditions.