Numerical Simulation and Experimental Study on the Solidification Structure of Al – Mg – Si Alloy in Squeeze Casting

1. Introduction

Al – Mg – Si alloys have attracted extensive attention in recent years due to their high strength and good corrosion resistance. They are widely used in the automotive manufacturing and other industries. Squeeze casting, as an advanced casting method, combines the advantages of casting and forging. It can produce thick – walled aluminum alloy components with fewer defects, refine grains, and improve the microstructure. However, the final mechanical properties of squeeze – cast components are affected by many factors, such as pouring temperature, mold temperature, and casting pressure. In addition, the solidification process involves complex non – equilibrium heat and mass transfer, making it difficult to study the process and mechanism only through experiments. Therefore, numerical simulation technology has become an important means to optimize the squeeze casting process.

2. Numerical Simulation of Squeeze Casting

2.1 Macroscopic Numerical Simulation

2.1.1 Establishment of Physical and Geometric Models

A three – dimensional model was created using software, and then the solid files of the casting and mold were imported into the ProCAST software. The unit grid size of the finite – element model was set to 2mm for the casting and 10mm for the mold based on the model structure. The final number of volume grids generated was 306,422. The finite – element model is shown .

2.1.2 Calculation of Material Parameters

The chemical composition of the new Al – Mg – Si alloy is shown in Table 1. Since there is no relevant data of this alloy grade in the ProCAST software material library, new materials need to be established and their physical property parameters calculated. The squeeze casting simulation is applicable to the Scheil diffusion model based on a relatively fast cooling rate.

2.1.3 Setting of Interface Heat Transfer Coefficient

The interface heat transfer types are shown in Table 2. During the filling stage, the interface heat transfer coefficient between the mold and the molten metal was set to $2000 ~W \cdot(m^{2} \cdot K)^{-1}$ . During the solidification stage, the heat transfer coefficient was set differently according to the applied pressure, and the relationship between the heat transfer coefficient and the pressure is $h = 2000+98.4p$ (h is the interface heat transfer coefficient, $W \cdot(m^{2} \cdot K)^{-1}$ ; p is the punch pressure, MPa).

2.1.4 Pressure Application

The application of pressure is reflected in the form of the “mold – molten metal” interface heat transfer coefficient, mainly considering the thermodynamic effect of pressure parameters during the squeeze casting process.

2.1.5 Process Conditions

The punch speed was set to $10 ~mm \cdot s^{-1}$ , and the holding time was set to 30s. Different pouring temperatures, mold temperatures, and casting pressures were studied, and the specific parameters are shown .

2.1.6 Operation Parameters

The GATEFEED in the operation parameters was set to ON for pressure – maintaining and shrinkage – compensation settings, and PENETRATE was set to ON to activate the inter – penetrating grid algorithm.

2.2 Microstructure Numerical Simulation

2.2.1 CAFE Mathematical Model

The CAFE model was selected for the numerical simulation of the solidification microstructure of squeeze casting. The implementation of the CAFE model is the coupling of the macroscopic temperature field and the microscopic solidification microstructure, which combines the finite – element method (FE) and the cellular automaton method (CA). The nucleation distribution under this model can be described by the Gaussian distribution: $\frac{d n}{ d(\Delta T)}=\frac{n_{max }}{\Delta T_{\sigma}} exp \left[-\frac{1}{2}\left(\frac{\Delta T-\Delta T_{1}}{\Delta T_{\sigma}}\right)^{2}\right]$ ( $\Delta T$ is the transient supercooling degree during nucleation, K; $\Delta T_{1}$ is the average nucleation supercooling degree, K; $\Delta T_{\sigma}$ is the standard deviation of the nucleation supercooling degree, K; $n_{max }$ is the maximum nucleation density, $m^{-3}$ ).

2.2.2 Establishment of the CA Model

Setting of Growth Kinetics Parameters: After inputting the chemical composition of the alloy in Table 1 into the ProCAST software, the physical property parameters were calculated, and then the CAFE parameters were calculated. The Gibbs – Thompson coefficient $\Gamma$ was $2 × 10^{-7}$ , $D_{t, max }$ was 10, $D_{t, max }$ (fit) was 10, and the growth kinetics parameters $a_{2}$ was $6.0932 × 10^{-7} ~m \cdot(s \cdot K^{2})^{-1}$ , $a_{3}$ was $1.6855 × 10^{-6} ~m \cdot(s \cdot K^{2})^{-1}$ .
Setting of Nucleation Parameters: For $n_{s, max }$ and $n_{v, max }$ , they were obtained through experimental metallographic structures and corresponding mathematical relationships ( $N_{V}=0.8 ~N_{A}^{3 / 2}=0.5659N_{L}^{3}$ , where $N_{v}$ , $N_{A}$ , and $N_{L}$ represent the number of grains per unit volume, per unit area, and per unit length, respectively). The other parameters were determined by referring to the experimental micro – structure diagrams. The final nucleation parameters used for the simulation are shown.

3. Experimental Methods

3.1 Alloy Melting and Microstructure Analysis Methods

The alloy was melted in a graphite – clay crucible in an electric resistance furnace at a melting temperature of 750℃. After pure aluminum melted, Al – 10Mn, Si, Fe, Zn, Cu, Al – 5Ti, Al – 4Zr, and Al – 2Sc were added in sequence. Finally, the magnesium block wrapped in aluminum foil was added to the melt. After melting for 30min, hexachloroethane was added for melt degassing. When the melt temperature reached the pouring temperature, it was held for 10min and then poured into a pre – heated metal mold for squeeze casting. After the casting was completely cooled, a 10mm parallel section sample was taken through the center of the bottom for rough grinding, fine grinding, and polishing. It was corroded with a 0.5% hydrofluoric acid (HF) aqueous solution. The microstructure of the alloy was observed using a Zeiss Axio Observer.Z1m optical microscope (Carl Zeiss AG, Germany), a JSM – 6301F scanning electron microscope (JEOL Ltd., Japan), and an EBSD back – scattered electron detector equipped on a Zeiss Merlin Compact field – emission scanning electron microscope (Carl Zeiss AG, Germany).

3.2 Shrinkage Porosity Statistics Method in Squeeze – Casting Castings

Grid Processing: The sample was gridded with a grid size of 5mm×5mm. A two – dimensional coordinate system was established, with the origin set at the intersection of the extension lines of the outer wall and the bottom, as shown in Figure 2(a). The grid coordinates were the center points of each grid.
Area Statistics: The shrinkage porosity in each grid was photographed by the BSE mode of the scanning electron microscope. The ImageProPlus 6.0 software was used to count the total shrinkage porosity area and its proportion in each grid, and the shrinkage porosity area fraction in each grid was calculated.
Data Set Creation: Origin software was used to create a data set of the shrinkage porosity area fraction and its distribution. The data was converted into a matrix in a gridded manner, as shown in Figure 2(b), and then a contour cloud map of the shrinkage porosity area fraction and its distribution in the casting was drawn.

4. Results and Discussion

4.1 Filling Process Analysis

The simulation results of the squeeze casting at different filling times with a pouring temperature of 700℃, a mold temperature of 500℃, and a pressure of 100MPa are shown in Figure 3. The solidus (Tliq) and liquidus (Tsol) temperatures of the alloy are 585.1℃ and 622.7℃, respectively. The results show that the melt filling process under these process conditions is relatively stable, and no defects such as gas entrapment occur, indicating that the design of the filling speed is reasonable.

4.2 Solidification Process Analysis

The simulation results of the solid – phase fraction change at different solidification times with a pouring temperature of 700℃, a mold temperature of 500℃, and a pressure of 100MPa are shown in Figure 4. At 2.28s when solidification starts, the solid – phase fraction in the middle – lower part (thin – walled area) is relatively high, followed by the middle – upper part, and the solid – phase fractions at the bottom and top are the smallest, corresponding to different solidification sequences. At 4.31s, the solid – phase fractions on the inner and outer surfaces of the middle – lower part have reached 0.85, indicating that the first solidifying part is the surface of the middle – lower part, and the corner between the middle – lower part and the middle – upper part has solidified, while the core has not solidified, indicating that the transition area between the thin and thick walls is an area prone to cracks. At 8.15s, the middle – lower part is completely solidified, and at this time, the solid – phase fractions on the inner and outer surfaces of the middle – upper part and the top have reached 0.9, indicating that the second solidifying part is the surface of the entire side wall. At 13.25s, the side wall is completely solidified, and the bottom starts to solidify. At 15.06s, the top is completely solidified, and then the bottom is completely solidified. At 21.94s, the bottom finally solidifies. As the last solidifying parts, the top and bottom have a greater tendency to form shrinkage porosity defects. So the solidification sequence of the variable – cross – section casting is: middle – lower part – middle – upper part – top – bottom.

4.3 Effects of Different Process Parameters on Shrinkage Porosity Volume Fraction and Distribution

4.3.1 Macro Model Verification

The shrinkage porosity distribution cloud maps of the casting under the conditions of a mold temperature of 475℃, a pouring temperature of 660℃, and a pressure of 100MPa are shown in Figure 5. Figure 5(a) is the simulation result, and Figure 5(b) is the experimental result (contour cloud map). The results show that the positions of shrinkage porosity in the simulation and experimental results (at the top) are relatively consistent, indicating that the macroscopic finite – element model of squeeze casting established in this paper is reasonable.

4.3.2 Effect of Pouring Temperature on Shrinkage Porosity Volume Fraction and Distribution

The simulation results of the shrinkage porosity volume fraction and distribution under different pouring temperatures with a mold temperature of 475℃ and a pressure of 100MPa are shown in Figure 6, and Figure 7 shows the shrinkage porosity volume fraction at different pouring temperatures. The results show that the shrinkage porosity is mainly distributed at the bottom and top of the casting. With the increase of the pouring temperature, the shrinkage porosity volume fraction first decreases and then increases, and the shrinkage porosity volume fraction is the smallest at a pouring temperature of 700℃.

The formation of shrinkage porosity is affected by two main factors: the feeding ability of the melt and the density difference between the liquid and solid phases. When the pouring temperature is 660℃, due to the relatively large viscosity of the alloy melt, the feeding ability of the melt between the dendrites is relatively weak, so the shrinkage porosity volume fraction of the casting is relatively large. When the pouring temperature is 740℃, although the melt viscosity decreases, the melt density increases significantly with the increase of temperature, and the density difference between the two phases is more significant, resulting in an obvious tendency to form shrinkage porosity, so the shrinkage porosity volume fraction in the casting is relatively large. When the pouring temperature is 700℃, the melt has a good feeding ability, and the density difference between the solid – liquid phases is not too large, so the shrinkage porosity volume fraction is the smallest.

4.3.3 Effect of Mold Temperature on Shrinkage Porosity Volume Fraction and Distribution

The simulation results of the shrinkage porosity volume fraction and distribution under different mold temperatures with a pouring temperature of 700℃ and a pressure of 100MPa are shown in Figure 8, and Figure 9 shows the shrinkage porosity volume fraction at different mold temperatures. The results show that with the increase of the mold temperature, the shrinkage porosity volume fraction first decreases and then increases, and the shrinkage porosity volume fraction is the smallest at a mold temperature of 475℃.

5. Conclusions

A finite – element physical and geometric model for squeeze casting of variable – cross – section cup – shaped parts and a CA model for grain growth were established, realizing the simulation of the filling process, shrinkage porosity distribution, and grain structure of the new Al – Mg – Si alloy during squeeze casting.
The solidification sequence of the variable – cross – section casting of the squeeze – cast Al – Mg – Si aluminum alloy is: middle – lower part – middle – upper part – top – bottom.
As the pouring temperature and mold temperature increase, the shrinkage porosity volume fraction first decreases and then increases; as the pressure increases, the shrinkage porosity volume fraction decreases.
The structures at the bottom and middle – upper part of the casting are relatively coarse, while those at the middle – lower part and the top are relatively fine, which is related to the wall thickness and supercooling degree.