Numerical Simulation and Experimental Studies of Gas Pressure Infiltration Al-356/SiC Composites

Abstract

In this study, the filling process, solidification parameters, temperature distribution, and residual stress distribution of gas pressure-infiltrated Al-356/SiC composites were investigated through simulation and experiment. In addition, a series of orthogonal tests was also carried out to precisely demonstrate the preheating temperature, infiltration temperature, and infiltration pressure. After a thorough analysis, the orthogonal tests revealed that the optimal process parameters are as follows: the SiC preheating temperature is 550 °C, the infiltration temperature of the Al-356 alloy is 620 °C, and the infiltration pressure is 8 MPa. The simulation results revealed that pressure had a sharp decrease of ~87% during filling, and the critical pressure was ~0.12 MPa. The velocity decreased with the increase in the filling time, and the average velocity was ~2.60 ms⁻¹. Feasible analysis suggested that critical pressure is ~0.11 MPa and average velocity is ~4.20 ms⁻¹; this difference is attributed to apparent velocity and the Kozeny constant. In the solidification process, shrinkage porosity appeared in the centers of the composites, which is evident with scanning electron microscopy. Moreover, the stress concentration of 171.3 MPa appeared in the composite region connected with the runner, which is the cause of the nucleation of the crack. However, based on the optimum orthogonal parameters and simulative results, the stress concentration was reduced, and crack-free and porosity-free composites were achieved.

1. Introduction

Aluminum-based ceramic-reinforced metal matrix composites (AMMCs) are an advanced class of materials [1,2] with excellent mechanical properties that cannot be achieved with traditional metals and alloys [3,4]. From a mechanistic point of view, the low cost of AMMCs as well as their high strength, high wear resistance, high thermal conductivity, and low thermal expansion have made them very attractive for various applications in the aerospace industry [5,6,7,8,9]. However, the reinforcement type, morphology, processing temperature, processing route, and final microstructure of the intended product are the key parameters in achieving high performance with AMMCs [10,11,12,13,14,15]. Hitherto, different routes were broadly employed for the fabrication of AMMCs, such as powder metallurgy [16], stir casting [17], and gas pressure infiltration [18]. Among them, powder metallurgy has the drawback of low-scale production and is cost-ineffective. In contrast, casting components are prone to low strength due to high porosity. Notably, gas pressure infiltration is a practical technique, especially for preparing different net shapes materials such as complex engine blocks, steering knuckles, etc. [19]. Therefore, in terms of complex shape components and large-scale production, the gas pressure infiltration route is far better than other conventional casting routes.

Previously, many attempts were made to fabricate Al-reinforced SiC composites by using the gas pressure infiltration technique, and simulations were conducted for predicting/acquiring perfect processing parameters. For example, Michaud and Jespersen [20] established an effective impregnation model for analyzing the infiltration phenomenon for a novel near-net shape performing method. A more precise study was conducted by Tong and Khan [21]: they designed models for the infiltration and re-melting zones and described the solidification phenomenon of metal in single-scale porous preforms. In addition, the slug flow approximation problem was tackled by using Darcy’s law. A similar piece of the study was conducted by Biswas [22], but their main focus revolved around the exchange of heat between the metal and the fiber. Pillai and Bo [23] constructed a three–dimensional model to simulate infiltration, solidification, and the transfer of heat during pressure infiltration while also focusing on the effects of the gate size and inlet pressure on the infiltration process.

Based on what we have discussed above, it can be summarized that some phenomena are more complex during the gas pressure infiltration process, such as changes in velocity, segregation, solidification, micro shrinkage, change in pressure, residual stress, and porosity formation. These are very significant parameters and must be considered in the numerical models and for the fabrication process. Most specifically, the pressure of the melt at different places in the SiC skeleton’s interior needs to be evaluated. Therefore, in the present study, the process of gas pressure infiltration was investigated through ProCAST to find out the characteristics of the filling and solidification processes. An approximate SiC skeleton model in micron scale and mold model were created to predict changes in pressure, velocity, filling time, stress concentration during solidification, shrinkage, and porosity. The orthogonal test was designed to find out the optimum temperature parameters. In addition, the microstructure of the final intended product was studied through scanning electron microscopy (SEM).

2. Experimental Method

2.1. Three–Dimensional Modeling

The actual SiC skeleton/preform having dimensions $\varnothing =$ 100 mm × 100 mm × 10 mm complicated its three–dimensional interconnected pore shape and spatial distribution. The porosity of the SiC ceramic preform was ~21.88%, thus it was difficult to prepare the pre-processing model. Thus, an approximate ceramic skeleton model was considered for simulation. Based on the symmetry of the composite and mold, the size of the SiC skeleton model was set to 1 mm × 0.5 mm × 0.1 mm, as shown in Figure 1a. The models of composite and molds were comprised of the outer mold, inner mold, SiC, aluminum, and runner, as shown in Figure 1b.

2.2. Setting Parameters

The database of ProCAST (Ver. 13.5, Universal energy system, USA) contains predefined information on common materials that are extensively studied in casting. Thus, parameters of Al-356 alloy, SiC, graphite, and H13 steel dies were used as per the database of ProCAST. The heat transfer coefficients of Al-356/SiC, Al-356/graphite, Al-356/H13 steel, SiC/graphite, and graphite/H13 were assigned 500 W/(m²·K), 1400 W/(m²·K), 1500 W/(m²·K), 700 W/(m²·K), and 1100 W/(m²·K), respectively. Initial conditions and boundary conditions were assigned based on experiments. The infiltration pressure was set at 8 MPa while the filling and solidification processes were under the argon gas atmosphere. In addition, the mold preheating and SiC preheating temperatures were 650 °C and 780 °C, respectively. A gas infiltration pressure experiment with the same temperatures was carried out to verify simulation results. The orthogonal test is designed to study the effects of processing parameters on the Al-356/SiC composite. The experiment and simulation were conducted to choose the optimum processing parameters, including preheating temperature, infiltration temperature, and infiltration pressure. For microstructure evolution, the specimens were machined from the cross-sectional surface, grounded with SiC papers, and polished with colloidal silica. The observations were made on (SEM, Fei Quanta 450 F, Hillsboro, OR, USA).

3. Results and Discussions

3.1. Numerical Filling Process Analysis

The results of filling parameters, such as time required for filling of Al-356 melt and changes of pressure and velocity at different places in the composites, are presented in this section. In the filling process, the runner was filled with molten Al-356 alloy at a pressure of 8 MPa; later, the SiC skeleton started to fill from the region where it contacted the runner to the bottom. It was observed that the small volume of melt was entered into the runner at a time of 0.0048 s, while it was filled at a time of 0.0205 s. Furthermore, it started to fill SiC, and the complete filling time of 0.1351 s was required for Al-356/SiC composite as shown in Figure 2a. During the filling of the runner, a reduction in pressure was observed and mainly attributed to gravity and viscosity resistance. However, this region is not the key, as we were consistent with finding the pressure of the melt in the interior of the composite. The capillary pressure inside the SiC skeleton and the viscosity resistance of melt Al-356 alloy caused the decrease in pressure during the filling process. Four nodes in the model were selected to determine the changes in pressure inside the composite, as shown in Figure 2b. Point 1 was face-centered to the symmetry plane, Point 2 was the geometric center of the model, Point 3 was the furthest position from the runner, and Point 4 was the position in contact with the runner. Subsequently, the results of the pressure variations while filling in the composite are shown in Figure 2c. It was observed that the pressure increased significantly with the increased distance between the nodes and the runner. The pressure at Point 4 near the runner was high at each interval of time, while the pressure at Points 2 and 3 showed slight variations after a time of 0.06 s and increased up to 1.2 MPa at a filling time of 0.135 s. The pressures of Point 1 and Point 2 were similar at ~1.18 MPa and ~1.30 MPa, respectively.

Notably, Point 4 constantly shows ~0.10 MPa pressure. In the end, the pressures of the four nodes abruptly reach 7.0 MPa at 0.140 s because of pressure compensation, which is lower than the applied pressure of 8 MPa, while the abrupt increase was found after 0.135 s, suggesting that the overall pressure inside the composite was quite low.

Theoretically, the capillary pressure P_c and the pressure caused by the viscosity resistance P_v can be calculated with Equation (1):

$$P_c=\frac{4\left(1-\epsilon \right){\sigma }_{LV}\cos \theta }{d_e\epsilon }$$

(1)

where ε is the porosity, ${\sigma }_{LV}$ is liquid surface tension, θ is the wetting angle, and d_e is the average pore diameter.

The assumptions according to technological practice are shown as follows:

(1)

The molten metal liquid cannot be compressed.

(2)

The preform will not produce deformation under applied pressure.

(3)

The metal liquid is melting in the filling process.

(4)

When the metal liquid is in the preform, heat exchange can occur immediately.

The ${\sigma }_{LV}$ and $\theta$ can be calculated by using Equations (2) and (3):

$${\sigma }_{LV}=868-0.152\left(T-T_m\right)$$

(2)

$$\theta =155-0.34T$$

(3)

The Reynolds number Re in the porous body can be calculated using Equation (4):

$$Re=\frac{\rho }{\eta }\mathrm{·}\frac{v}{S}$$

(4)

where Tm is melting temperature, ρ is the density of aluminum alloy, η is the viscosity of molten aluminum alloy, v is apparent velocity, and S is the specific surface area of the porous body. When the temperature of aluminum melt was 1053 K, ρ was 2.3 g cm⁻³, v was 1.2 ms⁻¹, η was 0.15 mN·s/m², S was 3 × 10⁴ m²/m³, and Re as calculated with Equation (4) was 613. The critical Re is 2, thus the fluid flow was turbulent. Thus, the P_v is calculated using the Ergun equation [24]:

$$\frac{\mathrm{\Delta }P}{H}=\frac{150{\left(1-\epsilon \right)}^2}{{\epsilon }^2}\mathrm{·}\frac{\eta v}{D_p^2}+\frac{1.75\left(1-\epsilon \right)}{{\epsilon }^3}\mathrm{·}\frac{\rho v^2}{D_p}$$

(5)

$$D_p=d_e\mathrm{·}{\phi }_s$$

(6)

where H is the distance of molten aluminum alloy, D_p is the feature size of particles, d_e is the mean equivalent diameter of particles, and φ_s is the average sphere factor.

The following equation can calculate the mean velocity of molten aluminum alloy:

$$\overline{v}=\frac{v}{\epsilon }\sqrt{\frac{\alpha }{2}}$$

(7)

where α is the Kozeny constant that is generally between 3.5 and 6.0. In this study, 6.0 was taken into consideration.

The P_c and P_v were calculated based on the above equations and the physical function parameters of materials as shown in

Table 1. The parameters and calculated results.

ε	de/µm	v/(m/s)	H/mm	η/(mN·s/m2)	β/(×10−12 m2)	$\overline{\mathit{v}}$/(m/s)	Pc/MPa	Pv/MPa
0.35	26.35	1.20	1.12	0.15	8.15	4.40	0.07	0.04

. The theoretical, critical infiltration pressure P₀ is 0.11 MPa. Meanwhile, P₀ as simulated with ProCAST was found to be 0.12 MPa. The pressure of Point 3 was found to be 3.24 MPa; it was much larger than the simulated result of 0.10 MPa. The main reasons for the difference may be described with two aspects: (1) the formula of P_c is just for the single channel condition, but the model has three–dimensional channels, and the crossing regions cause the capillary pressure to become more complicated, and (2) the flow state of alloy liquid is not lamellar flow or turbulent flow during the infiltration process. The state of the boundary layer is more complicated, and it causes the calculated result of P_v to be more inaccurate. Thus, there is a difference between calculated results and simulated results.

Based on the simulated results, the average velocity $\overline{v}$ of liquid Al alloy is 4.40 ms⁻¹ as listed in Table 1. Conversely, the theoretically calculated average velocity was 2.60 ms⁻¹. The difference is due to the values of apparent velocity and the Kozeny constant. On inspection of Figure 2d, it can be observed that the velocity decreases with increasing distance from the runner. The velocity of Point 4 was nearly a straight line and became a maximum of ~11.5 ms⁻¹ until 0.025 s elapsed, after which it decreased gradually after 0.03 s and reached ~8.3 ms⁻¹. On the other hand, the velocity of Points 2 and 4 show the same behavior with slight variations, but the velocity increases up to ~5 ms⁻¹ at 0.135 s. Moreover, the velocity at Point 3 was maintained at 0 ms⁻¹ before the filling time reached 0.135 s. This means the alloy liquid does not flow in that region. The reason is the pressure of the region 0.10 MPa is lower than P₀ ~ 0.11 MPa. The velocity of Point 3 reaches 2.3 ms⁻¹ at 0.137 s. The pressure compensation in the infiltration process is the main factor. The velocities of the four nodes were reduced to zero at the end of the filling process as shown in Figure 2d.

3.2. Experimental Solidification Process Analysis

The results of the solidification process, such as temperature and stress distribution, are presented in Figure 3 and Figure 4. During solidification, the runner is solidified first while the center area of the composite is solidified last as shown in Figure 3a(1–4). It is worth noting that the Al-356 alloy liquid in the SiC skeleton was not solidified while the runner was almost solidified. It hindered the liquid metal flow, leading to shrinkage porosity in the last solidified region as shown in Figure 3b. The maximum shrinkage porosity was found to be 81.5%, and the volume of shrinkage porosity was 16.49 × 10⁻³ mm³. For further verification, the signatures of the shrinkage porosity were also evident in the SEM micrograph and labeled with red arrows, as shown in Figure 3c.

Generally, it is believed that the temperature distribution in the solidification process also influences stress distribution. Figure 4 shows the stress distribution in Al-356/SiC composites. The stress distribution σ_x in the Y–Z section was the highest at ~171.3 MPa in the region connected to the runner, as shown in Figure 4a. In the X–Z section, the stress σ_y in the connection along the thickness direction is about 103 MPa, as shown in Figure 4b, whereas in the X–Y section, the stress σ_z is about 138 MPa, as shown in Figure 4c. According to the values of stress distribution obtained in this study, it can be deduced that the bottom of the composite material tends to nucleate the crack. Notably, in the beginning, the crack nucleates in the thickness direction of the composite and eventually leads to delamination in real A356/SiC composite as evident in Figure 4d. The porous structure of SiC complicates the ruleless pore channel, and the section shape usually changes. Therefore, shape-changing regions such as the junction between thick and thin or the junction of two intersecting surfaces are prone to stress concentration as displayed in Figure 4.

3.3. Temperature Distribution

In the previous section, it was described that an argon gas atmosphere was used for the fabrication of composites. Therefore, temperature change in the cooling stage was not as affected as it could be affected under an air atmosphere. Figure 5 shows the simulated evidence of temperature distribution at different times during the solidification process. As the temperature of aluminum alloy melt is higher than that of the preform, heat transfer will occur between the following: (1) the heat transfer path between the aluminum melt in the liquid guide pipe and the external environment of the A356-H13 steel die, and (2) the heat transfer path between the aluminum melt in the composite material and the external environment of the A356-SiC-H13 steel die. In addition, the melting chamber (the lower part of the furnace body) is set with specific cooling environments. Thus, the cooling rate of the melting chamber is higher than that of the infiltration chamber, and the aluminum melt in the melting chamber solidifies early during solidification. Figure 5 shows that the first solidified part of the composite is the outer surface, and the last solidified part is the center of the composite. From the temperature distribution of the composite, it is also observed that the temperature of the aluminum phase is lower than that of the SiC phase at the same time.

For clearer analysis, the temperature distribution of composite materials during the solidification is shown in Figure 5b. The first solidification position inside the composite includes the position contacting the liquid pipe and the right angle of the outer contour. The final solidification position is the lower part, but not the core of the composite because the core is located in the SiC phase and is connected with other parts of aluminum through a small channel. The structure of the small channel is similar to an island and located near the core of the composite; this makes the lower part less disposed to heat transfer, and hence becomes the final solidification part of the composite. The solidification time difference of the aluminum phase in the composite is 54 s.

3.4. Residual Stress Distribution

After the solidification of the SiC/Al composite, the stress in the composite is retained due to inconsistent shrinkage, which is further attributed to the large difference in thermal expansion coefficients between the SiC phase and the aluminum phase. However, in this study (simulation), the SiC phase was set as a rigid body when setting stress conditions; this simulation mainly calculates and investigates the stress distribution of the aluminum phase. Figure 6a–c displays the residual stress distribution on the X–Y, Y–Z, and X–Z planes, respectively. It can be seen that the small parts of the aluminum phase, especially the part of the pore connecting to the neck, are susceptible to stress concentration, resulting in high residual stress values in local points and parts. To better observe the residual stresses in these parts, the SiC phase is neglected and the aluminum phase is magnified as shown in Figure 6d. The maximum residual stress in these parts was observed at ~268 MPa, which was relieved with the nucleation of micro cracks. It is worth mentioning that it was difficult to observe the change rule of residual stresses in the composite, and it was mainly attributed to the irregularity of the aluminum phase in the CT model. Therefore, the residual stresses of the four nodes (Figure 2b) were also measured as shown in Figure 6e. The residual stress value of Point 4 (the part directly in contact with aluminum in the liquid drain pipe) was the highest at ~112 MPa, and the residual stress value of Point 1 (composite center) was the lowest at ~13 MPa. The main reason for the lower residual stresses is macro shrinkage porosity in the center, leading to incomplete contact between the aluminum and SiC phases. On the other hand, shrinkage porosity is a source of space for the aluminum phase, which also brings residual stress. It can be seen that Point 2 is closer to the liquid introduction pipe than Point 3. Therefore, the residual stress of Point 2 (~51 MPa) is significantly higher than that of Point 3 (~38 MPa).

In short, we can say that:

• The residual stress of the Al-356/SiC composite is affected by the distance from the liquid pipe. Therefore, the closer to the liquid pipe, the greater the residual stress value.
• The residual stress is higher in a position which has a low temperature.
• The residual stress is higher in the position where solidification occurs first.

3.5. Orthogonal Test Analysis

Filling time is associated with the reaction time between the Al-356 alloy liquid and the SiC skeleton; the filling time must be as short as possible to reduce the intensity of the reaction. After solidification of the Al-356/SiC composite, the shrinkage porosity may cause the reduction of mechanical and thermophysical properties. Therefore, the volume of shrinkage porosity needs to be eliminated or reduced. Moreover, residual stress concentration leads to the nucleation of cracks and seriously affects composite quality. Therefore, residual stress concentration should be as low as possible.

This paper studied three aspects, including the shrinkage porosity, residual stress, and filling time, to optimize temperature through an orthogonal test, as shown in

Table 2. The orthogonal test design.

Test (No.)	Preheating Temp. (°C)	Infiltration Temp. (°C)	Infiltration Press. (MPa)
	Factor A	Factor B	Factor C
1	A1(650)	B1(780)	C1(8)
2	A1(650)	B2(700)	C2(4)
3	A1(650)	B3(620)	C3(1)
4	A2(600)	B1(780)	C2(4)
5	A2(600)	B2(700)	C3(1)
6	A2(600)	B3(620)	C1(8)
7	A3(550)	B1(780)	C3(1)
8	A3(550)	B2(700)	C1(8)
9	A3(550)	B3(620)	C2(4)

. It is found that experiments 7, 8, and 9 had the minimum volume of shrinkage porosity with stresses of 170, 196, and 194 MPa respectively at filling times 0.41, 0.41, and 0.20, respectively. All other testing values are shown in

Table 3. The orthogonal experiment results.

Test	Indicator I	Indicator II	Indicator III
	Filling Time (s)	The Volume of Shrinkage Porosity (×10−3 mm3)	Maximum Equivalent Stress (MPa)
1	0.14	16.49	206.70
2	0.20	16.38	198.70
3	0.41	16.46	200.00
4	0.20	10.74	195.10
5	0.41	10.03	211.40
6	0.14	9.57	195.20
7	0.41	3.68	170.00
8	0.14	4.83	196.00
9	0.20	3.05	196.40

.

However, the best process parameters cannot be measured directly. These two parameters, the range of factors R and the average value of results (such as filling time, the volume of shrinkage porosity, and maximum equivalent stress) K, are required to evaluate the rationality of processing parameters. The R values are listed in

Table 4. R values for three factors.

R-Value	Factor A	Factor B	Factor C
	Preheating Temperature	Infiltration Temperature	Infiltration Pressure
IR	0.00	0.00	0.27
IIR	12.59	0.72	0.24
IIIR	14.40	11.20	17.30

, and the K values are shown in Figure 7.

From R-value, it can be deduced that the influence on filling time decreases in this order: C > A = B. Infiltration pressure is the most significant determinant of filling time. R values of factor A, preheating temperature, and factor B, infiltration temperature, were found to be 0. It means factors A and B do not affect the filling time; moreover, to reduce the filling time, C₁ (8 MPa) and C₂ (4 MPa) can be selected for factor C as shown in Figure 7b.

Preheating temperature is the most important factor for the volume of shrinkage porosity due to its high R-value. It can be deduced that the decreasing preheating temperature causes a distinctly reducing volume of shrinkage porosity. Thus, A₃ (550 °C) is the best choice. In contrast, R values of infiltration temperature and pressure are much smaller, as shown in Figure 7a. In addition, infiltration temperature B₁ (780 °C) can be chosen. C₃ (1 MPa) can be considered, but this pressure tends to increase fill time; therefore, C₂ (4 MPa) is the best choice.

The infiltration pressure is much more important; taking the R-values into account, the effect of residual stress reduces in this order: C > A > B. The R-value of factor A was 14.40. It means preheating temperature plays the second critical role in maximum equivalent stress. To reduce maximum equivalent stress factors, factor C and factor A are the main factors; therefore, we have obtained C₃ (1 MPa), C₁ (8 MPa), A₃ (550 °C), and A₂ (600 °C) as shown in Figure 7c. Moreover, by considering filling time and volume of shrinkage porosity, C₁ (8 MPa) and A₃ (550 °C) are the best choices. Although factor B is not as important as factors C and A, B₃ (620 °C) is the best choice.

Based on what we have discussed above, the optimum process parameters are A₃B₃C_1, and it has a better microstructure of the core of the SiC/Al composite as shown in Figure 7d. It can be seen that the interfacial bonding of the two phases in the composite is improved compared with the composite prepared with the original parameters. It is worth noting that the representative SEM micrograph displayed no obvious shrinkage defect and, therefore, porosity was only 1.0%. Based on these facts, it can be concluded that great reliability in simulation and experimental results has been achieved. Thus, preheating temperature, infiltration temperature, and infiltration pressure are 550 °C, 620 °C, and 8 MPa, respectively.

4. Further Discussion

Pouring temperature is also one of the key parameters that can define the size of the contact angle between the incorporated particles and the melted aluminum, which also significantly influences the surface tension of the melted aluminum [25]. For example, Narciso et al. [26] evaluated the temperature (1073, 1173, and 1273 K) effect on the surface tension value and size of the contact angle in an Al-Si alloy. Their experimental results show that the surface tension values and the size of the contact angle between Al and Si decrease as temperature increases in the Al-Si alloy. Tian et al. [27] also fabricated Al/SiC and Al–12%Si/SiC composites with nitrogen pressure infiltration in the temperature range of 923–1273 K. They claimed that the contact angle of infiltration and threshold pressure decrease with an increase in temperature. According to the study, the contact angle of infiltration and threshold pressure decreased slowly for temperatures in the range of 923–1173 K. Once the temperature reached 1173 K, both factors decreased rapidly. With a further temperature increase, the gaseous product Al₂0₃ content improves the wettability of aluminum melted and reinforced SiC particles. For the temperature above 1173 K, the infiltration threshold pressure and contact angle also decreased significantly. Thus, in short, it can be stated that pouring temperature has a substantial effect on the fabrication of Al/SiC composites.

At the start of the infiltration process of the composites, the flow rate of aluminum melt moving through the liquid pipe attained a speed of 20 ms⁻¹. It is well-accepted that small pores need a larger infiltration pressure, which leads to stagnation, and the horizontal pores need a longer infiltration time. In the current study, it has already been stated that the simulation results revealed that the infiltration process takes 0.0529 s. During the filling process of aluminum melt, the pressure values in the SiC preform decrease significantly. The pressure was recorded at ~2.95 MPa mm⁻¹ when the melt flowed in the complex, three–dimensional, connected micro-pores. The potential reason is the influence of capillary resistance and viscous resistance. However, the critical infiltration pressure of the Al/SiC composite is 4.07 MPa. During infiltration, the velocity of aluminum melt decreases gradually, and the change rule is the same as that of infiltration pressure. The maximum velocity reached 11.5 ms^−1, and the average velocity was recorded at 1.6 ms⁻¹ during the infiltration process. After the process, the impurities and oxides easily accumulation at the position where the final infiltration is completed. The simulation results of the temperature and solidification distribution reveal that the core of the composite is the last place to solidify. The shrinkage porosity is easily nucleated in the core of the composites, which is consistent with the porosity observed in the actual experimental core of fabricated composite. Conversely, the liquid inlet pipe is the first point of solidification. On the microscopic level, the residual stress in the aluminum phase connection neck of the composite is the highest at ~268 MPa, which causes microcracks. On the macroscopic level, the residual stresses at the position where the composite plate is closer to the liquid pipe are greater. More specifically, there is a concentration of tensile stress (~100 MPa) in the thickness direction, which can easily lead to delamination and cracking of the composite plate near the connection position of the liquid pipe. This result is consistent with the observed cracking failure phenomenon of Al/SiC composite plates. It is found that the change rule of pressure and velocity in the infiltration process as well as the valid stress distribution characteristics in the solidification process can be obtained by simplifying the micropore model compared to the numerical simulation results of the finite element model. Moreover, the regular characteristics of melt flow during infiltration, shrinkage defects, and stress concentration during solidification can be obtained through the finite element model based on CT scanning.

5. Conclusions

A numerical simulative and experimental study was conducted for A356/SiC composite. The feasible approximate ceramic skeleton model was modeled with ProCAST software. The following important outcomes were observed during the filling and solidification process.

(1)

The filling time of the composite was observed at 0.1356 s at a pressure of 8 MPa. The infiltration pressure has a sharp reduction of 87% of that applied pressure during the filling process. The average velocity during the simulation was 2.60 ms⁻¹, and the theoretical velocity was calculated to be 4.40 ms⁻¹; the difference is attributed to apparent velocity and the Kozeny constant.

(2)

In the solidification process, shrinkage porosity emerges at the center of the SiC/Al composite material, which is the last cooled place mainly due to lack of feeding. This result was evident with SEM analysis as well.

(3)

The stress concentration at the region connected with the runner is 171.3 MPa and, in the thickness direction, 138 MPa, resulting in the delamination fracture of composites and nucleation of cracks.

(4)

The orthogonal experiment shows infiltration pressure is the most important factor for filling time and maximum equivalent stress; preheating temperature is the second most significant factor for the volume of shrinkage porosity. The optimum temperature parameters are a preheating temperature of 550 °C, an infiltration temperature of 620 °C, and an infiltration pressure of 8 MPa. We have successfully achieved that result in better filling and solidification of the composite as shown by SEM analysis.