In my research, I focus on the solidification behavior and microstructural evolution of Mg-6Al-xZn alloys, commonly referred to as AZ6x alloys (where x = 0, 2, 4, 6 wt.%), produced via sand casting. Sand casting is a widely used manufacturing process, particularly in aerospace applications, due to its ability to create complex shapes with relatively low cost. The microstructure of sand cast alloys, including grain size and second-phase formation, significantly influences their mechanical properties. Understanding these aspects is crucial for optimizing alloy design and processing conditions. In this work, I employ thermal analysis, microscopy, and thermodynamic modeling to investigate how zinc content affects the solidification pathways, grain refinement, and dendritic coherency in these magnesium alloys.
The experimental procedures involved preparing the AZ6x alloys using high-purity materials. I melted commercial pure Mg (99.9%), pure Al (99.9%), pure Zn (99.9%), and an Al-10%Mn master alloy in a low-carbon steel crucible under a protective atmosphere. The melts were poured at 735°C into cylindrical resin sand molds with an inner diameter of 100 mm and height of 120 mm. The sand casting process was chosen to simulate typical industrial conditions, where cooling rates are moderate. Chemical composition analysis was performed using inductively coupled plasma atomic emission spectroscopy (ICP-AES), and the results are summarized in Table 1.
| Alloy | Al | Zn | Mn | Mg |
|---|---|---|---|---|
| AZ60 | 5.74 | – | 0.22 | Bal. |
| AZ62 | 5.79 | 1.90 | 0.21 | Bal. |
| AZ64 | 5.69 | 3.70 | 0.24 | Bal. |
| AZ66 | 5.82 | 5.66 | 0.28 | Bal. |
To monitor the solidification process, I used a two-thermocouple thermal analysis technique. Two K-type thermocouples were placed at the mold wall and center, respectively, to record temperature-time curves. The temperature difference between these thermocouples helps identify the dendrite coherency point (DCP), a critical moment during solidification when dendrites begin to interconnect and form a continuous network. The solid fraction at DCP, denoted as $f_s^{DCP}$, was calculated from the cooling curves using Newtonian baseline methods. Additionally, I performed thermodynamic calculations using Pandat software with the Pan Mg database to simulate equilibrium phase diagrams, non-equilibrium solidification via the Scheil model, and growth restriction factors (Q values). Microstructural characterization was carried out using scanning electron microscopy (SEM) and electron backscatter diffraction (EBSD) to examine second-phase morphologies and quantify grain sizes.
The solidification behavior of sand cast AZ6x alloys revealed distinct phase transformations. For the AZ60 alloy, the cooling curve showed two main events: the nucleation of primary α-Mg and the formation of γ-Mg17Al12 phase through a non-equilibrium eutectic reaction. In alloys with zinc additions (AZ62 to AZ66), an additional phase, Φ-Mg21(Al,Zn)17, appeared. The amount of γ-Mg17Al12 decreased while Φ-Mg21(Al,Zn)17 increased with higher Zn content. Thermodynamic calculations confirmed that under equilibrium conditions, γ-Mg17Al12 and Φ-Mg21(Al,Zn)17 can dissolve into α-Mg at appropriate temperatures for AZ60 to AZ64 alloys, but for AZ66, Φ-Mg21(Al,Zn)17 remains stable at all temperatures, limiting heat treatment options. This highlights the importance of composition control in sand casting to achieve desired microstructures.
Grain size is a key parameter influenced by alloying elements. I quantified the average grain size using EBSD analysis, as summarized in Table 2. The growth restriction factor (Q) was calculated to assess the effect of solute elements on grain refinement. For multicomponent alloys like AZ6x, the true Q value is derived from thermodynamic data rather than simple binary approximations. The relationship between grain size (d) and Q is often expressed as:
$$d = a + \frac{b}{Q}$$
where a and b are constants related to nucleation site density and potency. However, my results indicate that this linear relationship does not hold perfectly for the AZ6x system, suggesting complex interactions between solutes during sand casting solidification.
| Alloy | Average Grain Size (μm) | Q Value | $f_s^{DCP}$ from Thermal Analysis (%) | $f_s^{DCP}$ from Scheil Simulation (%) |
|---|---|---|---|---|
| AZ60 | 557 | 21 | 35 | 36 |
| AZ62 | 275 | 28 | 27 | 27 |
| AZ64 | 271 | 34 | 26 | 31 |
| AZ66 | 235 | 43 | 25 | 23 |
The dendrite coherency point is another critical aspect of solidification in sand casting. I determined $f_s^{DCP}$ from both thermal analysis and Scheil simulations, as shown in Table 2. As Zn content increased, $f_s^{DCP}$ generally decreased, indicating that higher solute levels promote earlier dendrite interaction. This trend correlates with reduced grain size and increased Q values. The lower $f_s^{DCP}$ in zinc-rich alloys implies that dendrites become coherent at a lower solid fraction, which can affect defect formation like microporosity and hot tearing during sand casting. The relationship between $f_s^{DCP}$, grain size, and Q can be explained by solute-driven growth restriction: higher Q values inhibit dendritic growth, leading to thinner dendrite arms that interconnect sooner.
To further analyze the solidification kinetics, I calculated the solid fraction ($f_s$) as a function of temperature using both experimental cooling curves and Scheil simulations. The Scheil model assumes no diffusion in the solid and complete mixing in the liquid, which approximates non-equilibrium conditions common in sand casting. The solid fraction is given by:
$$f_s = \frac{\int_0^t \left[ \left( \frac{dT}{dt} \right)_{cc} – \left( \frac{dT}{dt} \right)_{zc} \right] dt}{\int_0^{t_s} \left[ \left( \frac{dT}{dt} \right)_{cc} – \left( \frac{dT}{dt} \right)_{zc} \right] dt}$$
where $\left( \frac{dT}{dt} \right)_{cc}$ is the cooling rate from the curve, $\left( \frac{dT}{dt} \right)_{zc}$ is the baseline rate, and $t_s$ is the total solidification time. Comparisons between experimental and simulated $f_s$ curves showed good agreement in early solidification stages but deviations later, likely due to limited diffusion in the solid phase during sand casting.
The formation of second phases plays a vital role in the properties of sand cast Mg-Al-Zn alloys. In AZ60, only γ-Mg17Al12 is present, while in AZ62 to AZ66, both γ-Mg17Al12 and Φ-Mg21(Al,Zn)17 coexist. The transition between these phases depends on Zn content and solidification conditions. Thermodynamic calculations of vertical sections for Mg-5.76Al-0.24Mn-xZn systems revealed that for AZ60 and AZ62, a two-phase region of α-Mg and Al-Mn compounds exists, allowing complete dissolution of secondary phases upon heat treatment. For AZ64, careful temperature control is needed to avoid partial melting or incomplete dissolution. However, for AZ66, the stable Φ-Mg21(Al,Zn)17 phase cannot be fully dissolved, restricting the alloy’s processability. This underscores the need for precise composition design in sand casting applications.
The growth restriction factor (Q) is a key parameter for understanding grain refinement. For binary alloys, Q is defined as:
$$Q = m_L C_0 (k – 1)$$
where $m_L$ is the liquidus slope, $C_0$ is the solute concentration, and $k$ is the partition coefficient. For multicomponent systems like AZ6x, I used a more accurate thermodynamic approach to compute Q:
$$Q_{true} = \left( \frac{\partial (\Delta T_{cs})}{\partial f_s} \right)_{f_s \rightarrow 0}$$
where $\Delta T_{cs}$ is the constitutional undercooling. The calculated Q values increased with Zn content, as shown in Table 2, reflecting stronger growth restriction. This aligns with the observed grain size reduction from 557 μm in AZ60 to 235 μm in AZ66. However, the relationship between d and 1/Q was not strictly linear, indicating that other factors, such as nucleation site availability and solute interactions, also influence grain size in sand casting.
The dendrite coherency point (DCP) is crucial for understanding casting defects. I determined DCP temperatures ($T_{DCP}$) from thermal analysis by identifying the maximum temperature difference between wall and center thermocouples. The corresponding solid fractions $f_s^{DCP}$ decreased with higher Zn content, as summarized in Table 2. This decrease correlates with increased Q values and finer grain sizes. In low-solute alloys like AZ60, dendrites grow rapidly in all directions, leading to higher $f_s^{DCP}$ and coarser grains. In contrast, high-solute alloys like AZ66 exhibit inhibited dendritic growth, resulting in thinner arms that interconnect at lower $f_s^{DCP}$. This behavior impacts the mechanical integrity of sand cast components, as earlier coherency can reduce hot tearing susceptibility but may promote microporosity if not controlled.
To illustrate the microstructural evolution, EBSD maps revealed distinct grain morphologies. AZ60 showed large, equiaxed grains with an average size of 557 μm, while AZ62 to AZ66 exhibited progressively finer grains. The grain refinement with Zn addition is attributed to increased growth restriction and potentially enhanced nucleation during sand casting solidification. Additionally, the distribution of second phases influenced grain boundary pinning, further affecting grain size. The use of sand casting, with its moderate cooling rates, allowed for detailed observation of these microstructural features.
Thermal analysis provided insights into phase transformation temperatures. Key temperatures such as the onset of primary α-Mg nucleation ($T_{onset}$), peak temperatures for phase reactions ($T_{peak}$), and DCP temperatures ($T_{DCP}$) are listed in Table 3. These data help construct cooling curves and validate thermodynamic models. For example, in AZ64 and AZ66, a third peak corresponding to Φ-Mg21(Al,Zn)17 formation was detected, consistent with microstructural observations.
| Alloy | Peak A: $T_{onset}$ / $T_{peak}$ | Peak B: $T_{onset}$ / $T_{peak}$ | Peak C: $T_{onset}$ / $T_{peak}$ | $T_{DCP}$ |
|---|---|---|---|---|
| AZ60 | 617 / 614 | 439 / 436 | – / – | 610 |
| AZ62 | 610 / 608 | 407 / 404 | – / – | 604 |
| AZ64 | 602 / 599 | 389 / 385 | 361 / 359 | 593 |
| AZ66 | 598 / 594 | 372 / 369 | 362 / 360 | 587 |
The solidification curves derived from Scheil simulations and thermal analysis were compared. In early stages, both methods showed similar trends, but deviations occurred near the end of solidification due to assumptions in the Scheil model and baseline errors in experimental data. Despite these discrepancies, the overall solidification behavior in sand casting was well-captured, providing a basis for process optimization.
In conclusion, my investigation into sand cast Mg-6Al-xZn alloys demonstrates that zinc content significantly influences solidification behavior, grain size, and dendrite coherency. Higher Zn levels increase the growth restriction factor (Q), leading to finer grains and lower solid fractions at the dendrite coherency point. The formation of second phases, particularly Φ-Mg21(Al,Zn)17, becomes more prominent with Zn addition, affecting heat treatment responses. Sand casting, as a versatile manufacturing method, allows for controlled microstructural development through composition and process adjustments. These findings contribute to the design of high-performance magnesium alloys for lightweight applications, emphasizing the importance of integrated experimental and thermodynamic approaches in materials science.
Future work could explore the effects of other alloying elements or varying sand casting parameters, such as mold geometry and cooling rate, on solidification characteristics. Additionally, mechanical property testing could correlate microstructural features with performance, further optimizing these alloys for industrial use. The insights gained from this study underscore the complexity of multicomponent solidification and the value of sand casting as a tool for producing tailored microstructures.