The performance of castings is primarily determined by their microstructural characteristics, making the simulation and control of microstructure a critical focus in modern research. In high manganese steel casting, which is renowned for its high strength and wear resistance, challenges such as low thermal conductivity and slow heat diffusion often lead to coarse grains, developed dendrites, and segregation of impurities at grain boundaries. These issues can cause defects like spalling and cracking, significantly compromising the service life of high manganese steel casting components. To address this, numerical simulations integrating cellular automaton (CA) and finite element methods (FEM) have been employed to predict and optimize the solidification microstructure in high manganese steel casting, particularly for thick-section applications like railway frogs.
In this study, we utilize the Procast software’s microstructure module, which couples CA and FEM, to simulate the solidification process of high manganese steel casting. The CA method, grounded in nucleation physics and dendrite growth kinetics, enables the prediction of grain size, distribution, and the transition from columnar to equiaxed crystals. Meanwhile, FEM handles macroscopic transport phenomena, including heat conduction and momentum transfer, to compute temperature and velocity fields. The integration of these approaches allows for accurate reproduction of nucleation and growth under varying solidification conditions, providing insights into the evolution of microstructure in high manganese steel casting.
The mathematical foundation for microstructure simulation involves nucleation and growth models. For nucleation, a continuous model based on Gaussian distribution is applied, expressed as:
$$ n = n_{\text{max}} \frac{1}{\sqrt{2\pi} \Delta T_{\sigma}} \exp\left(-\frac{1}{2}\left(\frac{\Delta T – \Delta T_{\text{max}}}{\Delta T_{\sigma}}\right)^2\right) $$
where \( n \) represents the nucleation density, \( n_{\text{max}} \) is the maximum nucleation site density, \( \Delta T \) is the undercooling, \( \Delta T_{\text{max}} \) is the mean undercooling for maximum nucleation rate, and \( \Delta T_{\sigma} \) is the standard deviation. For dendrite tip growth kinetics, a simplified KGT model is used:
$$ V_{\text{tip}}(\Delta T_c) = 5.85 \times 10^{-6} (\Delta T_c)^2 $$
where \( V_{\text{tip}} \) is the tip growth velocity and \( \Delta T_c \) is the constitutional undercooling. These equations are essential for capturing the dynamic behavior of high manganese steel casting during solidification.
Our experimental setup involved a thick-section high manganese steel casting for a railway frog component, with a total length of 2000 mm and a maximum thickness of 132 mm. The casting process employed conventional sand casting with insulating risers and chills to manage solidification. For microstructure refinement, we introduced a process of adding metal particles during pouring to enhance nucleation sites. The simulation focused on a critical region of the high manganese steel casting to analyze microstructure evolution under different conditions, such as varying pouring temperatures and nucleation site densities.
The simulation results for conventional high manganese steel casting, with a pouring temperature of 1450°C and an average cooling rate of 0.1°C/s, indicated a nucleation site density of approximately \( 2 \times 10^7 \, \text{m}^{-3} \). This led to a microstructure dominated by coarse columnar and equiaxed grains, as validated by experimental observations. The table below summarizes the key parameters and outcomes for different scenarios in high manganese steel casting:
| Parameter | Conventional Casting | With Particle Addition (Density \( 5 \times 10^7 \, \text{m}^{-3} \)) | With Particle Addition (Density \( 1 \times 10^8 \, \text{m}^{-3} \)) |
|---|---|---|---|
| Pouring Temperature (°C) | 1450 | 1450 | 1450 |
| Nucleation Site Density (m⁻³) | \( 2 \times 10^7 \) | \( 5 \times 10^7 \) | \( 1 \times 10^8 \) |
| Grain Size | Coarse | Fine and Uniform | Fine with Porosity |
| Microstructure Homogeneity | Low | High | Moderate |
Increasing the pouring temperature to 1500°C and 1550°C resulted in coarser columnar grains and extended columnar growth in high manganese steel casting, due to enhanced thermal gradients. The temperature profiles at different points in the casting region showed that higher superheat promoted dendritic coarsening, while reducing the temperature gradient favored equiaxed crystal formation. The evolution of microstructure under these conditions highlighted the competition between columnar and equiaxed growth, which is critical for controlling the properties of high manganese steel casting.
To achieve grain refinement in high manganese steel casting, we simulated the effect of adding metal particles during pouring, which increases nucleation site density. At a density of \( 5 \times 10^7 \, \text{m}^{-3} \), the microstructure became finer and more uniform, with reduced columnar grain size and enhanced equiaxed crystal formation. This aligns with experimental results where adding 2% by mass of 1 mm diameter metal particles yielded similar refinement. However, when the nucleation site density reached \( 1 \times 10^8 \, \text{m}^{-3} \), simulations predicted the formation of shrinkage porosity due to rapid dendrite coalescence blocking liquid feeding channels. This underscores the importance of optimizing particle addition to avoid defects in high manganese steel casting.
The relationship between undercooling, nucleation density, and grain size can be further analyzed using the following equation derived from the nucleation model:
$$ \Delta T = \Delta T_{\text{max}} + \Delta T_{\sigma} \sqrt{-2 \ln\left(\frac{n \sqrt{2\pi} \Delta T_{\sigma}}{n_{\text{max}}}\right)} $$
This equation helps in predicting the undercooling required for effective nucleation in high manganese steel casting, which is vital for designing casting processes. Additionally, the growth velocity equation emphasizes the sensitivity of dendrite tips to undercooling, influencing the overall solidification morphology in high manganese steel casting.
In conclusion, the CA-FEM coupling approach provides a robust framework for simulating and controlling the solidification microstructure in high manganese steel casting. By adjusting pouring temperature and nucleation site density, we can tailor the grain structure to enhance mechanical properties and durability. The optimal nucleation site density for high manganese steel casting was found to be around \( 5 \times 10^7 \, \text{m}^{-3} \), beyond which porosity risks increase. This research demonstrates the potential of numerical simulations in advancing the quality and reliability of high manganese steel casting for demanding applications like railway components.