Effect of flow characteristics in the inner gate on the microstructure of a semi-solid processed hypereutectic Al–Si alloy

ABSTRACT This study aims to prepare hypereutectic Al–Si alloy cylindrical samples with hard inner surfaces by changing the depth of inner gates in squeeze casting molds. The characteristic parameters (volume fraction and equivalent diameter) of the hard phases (primary Si and Fe-rich phases) and the hardness of the samples were observed. When the depth of the inner gate is 5 mm, there is a radial gradient in the characteristic parameters of the hard phases in the members resulting in surface strengthening of the structural member. However, when the depth of the inner gate is 9 mm, there is no radial gradient. The mechanisms of achieving surface strengthening in hypereutectic Al–Si alloy cylindrical samples by changing the rheological characteristics of semi-solid Al–Si alloy slurries were revealed.


Introduction
Hypereutectic Al-Si alloy (12 ≤ Si% ≤ 26%) is an ideal material for cylinder liners in all-aluminum engines because of its good wear resistance, thermal stability, and low coefficient of thermal expansion [1][2][3].The primary Si particles in the hypereutectic Al-Si alloy are hard and brittle, and their microhardness can exceed 1000 Hv, which can ensure good abrasion resistance for the hypereutectic Al-Si alloy.The primary Si particles in the hypereutectic Al-Si alloy are coarse and irregular due to their complex nucleation, resulting in hypereutectic Al-Si alloy with poor processing properties and low tensile strength, therefore, the coarse Si particles in the hypereutectic Al-Si alloy need to be modified by using melt-controlled cooling technology to improve the homogeneity [4][5][6].
Hypereutectic Al-Si alloy has a wide solid-liquid phase range (around 200°C), which is prone to severe air rolls and the formation of defects such as holes during the conventional casting process [7].These unique material properties and poor material formability greatly limit the commercial applications of high-silicon aluminum alloys.The performance of high-silicon aluminum alloys cast in semi-solid casting is better than those formed using traditional die-casting, sand-casting, and metal mold-casting [8].
The semi-solid technology achieves spheroidal microstructures formed by shear force at the temperature range where the solid and liquid phases coexist [9,10].The process can change the morphology and distribution of Si particles, and the semi-solid compactness is good, which is conducive to the improvement of the mechanical properties of Al-Si alloy [11].
Squeeze casting is a special casting technique combining high-pressure die-casting and melt filling technology, which is suitable for semi-solid melting forming alloys with high apparent viscosity [12].The filling pressure directly acts on the surface of the entire metal melt to improve the mechanical properties of the material.The semi-solid alloys melts with high apparent viscosities and demonstrate good molding properties at high pressures.At the same time, the defects caused by solidification during the filling process of liquid metal can be avoided [13,14].
The key parameters of the process include filling rate, final molding pressure, and mold temperature.In addition, the squeeze casting process can avoid the solidification caused by the molten metal filling process defects [15].
Lin et al. [16] pointed out that high viscosity slurries and low molding temperature were conducive to the production of hypereutectic Al-Si alloy castings with fewer defects.However, the study on the microstructure of hypereutectic Al-Si alloy shows that the semisolid slurry of hypereutectic Al-Si alloy is a nearliquid structure composed of a large number of suspended primary Si particles and liquid Al melts [17][18][19][20].The agglomeration of small primary Si in liquid will significantly affect the performance of subsequent castings.In addition, the distribution of primary Si phases is random and difficult to control.The purpose of this research is to regulate the distribution of hard phases in the rheological squeeze casting hypereutectic Al-Si alloy cylinder liner by changing the inner gate structure of the mold.The influence of the in-gate structure of extrusion die on the distribution of hard phases was studied by combining the difference of hard-phase characteristic parameters and macro-hardness in different regions of the cylinder liner.The discussion focuses on the influence mechanism of the inner gate structure on the flow field and stress field during the filling process.

Preparation of semi-solid alloy slurry
The composition of the hypereutectic Al-Si alloy cylindrical samples is displayed in Table 1.Pure Al (99.9%) and instant silicon (98%, the rest being combustion accelerants) were added in proportion.Fe (75%, the rest being combustion accelerants) and Mn (75%, the rest being combustion accelerants) were smelted in a resistance furnace.Figure 1 shows the preparation process of hypereutectic Al-Si alloy cylindrical samples.The high-silicon aluminum alloy semi-solid slurry was prepared by the rotating rod induction nucleation device, in which the diameter of the hollow rotating rod was 75 mm, and the inner diameter was 75 mm; this system had a continuous drainage system for cooling and a maximum speed of 2500 rpm.A semi-solid squeeze casting method was used with the filling direction being vertically upwards.Process parameters included smelting temperature at 790 ± 20°C, the forming pressure of 160 MPa, and the extrusion rate of 16 mm/s.
Figure 2 shows hypereutectic Al-Si alloy cylindrical samples with a curved inner gate of 5 mm depth   (D5) and a curved inner gate of 9 mm depth (D9), respectively.The surface quality of hypereutectic Al-Si alloy samples is intact and the slurry fills the mold completely.As shown in Figure 2(b, d), the radial distribution of the hard phase was observed by taking the samples from three parts (upper, middle, and lower, respectively).

Microstructural analysis
Sampling was made from inside to outside in different parts of the cylindrical samples (upside, middle, and lower).The sample selection area is shown in Figure 3.The samples were etched in 0.5% HF solution and the microstructures were observed by optical microscopy (Eclipse MA200, Nikon).The equivalent diameter (ED) and volume fraction (VF) of hard phases of different samples were analyzed by using Image-pro Plus software, the ED was calculated by use of the following formula [21]: where N represents the total number of the hard phases and A i is the area of hard phase i.The macrohardness of the cylinder liner is measured using an HR-150A Rockwell hardness tester, the test load is 980 kN, and the loading time is 5 s.

Results and discussion
Microstructures of D5 samples     Figure 5(a, b) displays the variations in hard-phase characteristic parameters of the radial section microstructure of D5 samples.The ED and VF of the hard phases in D5-B and D5-M samples are gradually reduced from the inside to the outside owing to the fact that the inner gate accelerates the flow rate inside the mold filling process, resulting in the aggregation of solid hard phases on the faster flowing side; the ED and VF of the hard phases in D5-U samples increase from the inside out and then decrease, which is due to the fact that the semi-solid slurries on the inside are chilled by the core, causing the slurries to solidify and shrink, so that the liquid phase continuously flows to the inside, thus the ED and VF of hard phases are reduced.The degree of change in the characteristic parameters of the hard phases in the D5-U samples is lower due to the gradual decrease in speed during filling, resulting in the changes in characteristic parameters of hard phases being negligible.Figure 5(c) shows the trend in the hardness of the radial microstructure of D5 samples; the change in hardness is similar to the characteristic parameters of the hard phase, showing that the macro-hardness of the samples is related to the hard phase, the higher the VF and average grain diameter of the hard phase, the stronger the specimens.microstructure of D9-U sample, the distribution of primary Si particles is relatively uniform, Figure 6(e) exhibits the outside microstructure of D9-U sample, and the eutectic Si phases are shown to be fibrous.Figure 6(f-j) shows the radial section microstructure of D9-M samples from inside to outside. Figure 6(h) shows the middle microstructure of D9-M sample, there are a few flaws in this part.Figure 6(k-o) illustrates the radial section microstructure of D9-B samples from inside to outside. Figure 6(k) shows the inside microstructure of D9-B sample, there is significant agglomeration in the hard phases.Figure 6(m) displays the middle microstructure of D9-B sample and part of the primary Si particles grows into long rods.

Microstructures of D9 samples
Figure 7(a-b) shows the variation trend of hard phases' characteristic parameters of the radial section microstructure of D9 samples.The VF of the hard phases in D9-M and D9-U samples varies irregularly, because the deeper inner gate makes the filling speed too fast, causing turbulence at the front end of the filling, and thus the distribution of the hard phase cannot be controlled.Meanwhile, the difference in ED of hard phases is small, which indicates that turbulence has little effect on the size of hard phases.Figure 7(c) presents the change in the hardness of the radial microstructure of D9 samples.The variations in the hardness of D9-M and D9-U samples are similar to that of the VF of the hard phase, but that of D9-B sample is not, which is because the excessive filling speed and pressure at the inner gate leads to stress concentration, so the main strengthening mechanism of the D9-B sample is strain hardening.filling velocity (V 1 ) is 30.14 mm/s, the filling mode is laminar, and the filling front end is stable.The VF of hard phases in D5-B and D5-M samples shows a gradient distribution in the radial direction, which is because the inner gate changes the flow field and stress field of semi-solid slurries in the filling process.On the one hand, the inner gate accelerates the filling speed of the inside of the die, and the solid hard phases flow to the faster side due to Magnus effect [22], causing a gradient distribution of the VF of the hard phases in the radial direction.On the other hand, the inner gate changes the stress field in the filling process, so that the inside of semi-solid slurries is subjected to axial extrusion.The liquid phases begin to flow to the outside before the solid phase due to its good fluidity, resulting in solid-liquid separation (Figure 8(c)).For the D5-B sample, due to its slow rate of solidification, the liquid phase has enough time to move outward, resulting in an extremely low hardphase content outside the radial direction of the sample [23]; for the D5-M sample, some of the liquid phases are wrapped by solid phases because of the decreases in pressure and the obstruction of solidified solid phases, forming liquid phase segregation.Meanwhile, for the D5-U sample, due to the fact that the semi-solid slurries on the inside are chilled by the core, causing the slurries to solidify and shrink, so that the liquid phase continuously flows to the inside, thus the VF of hard phases in the radial direction is reduced.The VF and ED of primary Si affect the macro-hardness of D5 samples.Zhao C et al. [24] pointed out that the hardness of the material increased with the increase of the volume fraction of the primary Si.The primary SI strengthens the hypereutectic Al-Si alloy by transfer load.

Mechanism
Figure 8(b) shows the filling schematic of D9 samples.In the filling process, the maximum vertical average filling velocity (V 2 ) is 43.7 mm/s.An excessive filling speed results in the filling mode not advancing in laminar flow [25].Combined with the previous analysis, except for the D9-B samples, there is no obvious radial gradient distribution in the hard-phase characteristic parameters of other samples due to the presence of the inner gate that the VF of the hard phases in the D9-B samples shows a gradient distribution in the radial direction.However, the deeper inner gate leads to greater filling speed, resulting in turbulence in the front end of the filling, which cannot control the distribution of solid hard phases.Meanwhile, due to the short filling time, the VF of the liquid phase in the semi-solid slurries increases, thus the pressure has little effect on the distribution of solid hard phases.For D9-B samples, the excessive filling speed and pressure of the 9 mm inner gate caused the stress concentration.Therefore, the strengthening mechanisms of D9-B samples include strain hardening and load strengthening.

Conclusions
The effects of different inner gate structures on the microstructures and hypereutectic Al-Si alloy cylinder liners were investigated.The main conclusions are drawn as follows: (1) The inner gates change the stress field and flow field of hypereutectic Al-Si alloy semi-solid slurry during the filling process, which leads to the deviation of the liquid phase and solid hard phases in speed or direction, thus forming a unidirectional gradient distribution of hard phases and inner surface strengthen.(2) For the D5-U samples, due to the solidification and shrinkage of semi-solid slurries by the cooling effect of the core, the liquid phase continuously flows into the inside, which causes a decrease outward in the VF of the hard phase in the radial direction.
(3) For the D9-M and D9-U samples, the deep inner gate increases the velocity, leading to turbulence in the front end of the filling type, so the distribution of solid hard phases cannot be controlled.(4) The VF and ED of the hard phases greatly affect the macro-hardness of the samples.However, the excessive filling speed and pressure of the inner gate caused the stress concentration.Therefore, the main strengthening mechanism of the D9-B samples which near the inner gates is strain hardening.

Disclosure statement
No potential conflict of interest was reported by the author(s).

Funding
The

Figure 1 .
Figure 1.Schematic drawings showing the semi-solid processing process for the preparation of the hypereutectic Al-Si alloy cylindrical samples.

Figure 2 .
Figure 2. The schematic illustrations of hypereutectic Al-Si alloy cylindrical samples with different inner gate structures: (a, b) the depth of the curved inner gate is 5 mm (D5) and (c, d) the depth of the curved inner gate is 9 mm (D9).

Figure 4 (
Figure4(a-e) exhibits the radial section microstructure of D5-U samples from inside to outside.The hard phase particles (primary Si particles and iron-rich phase particles) are partially agglomerated, and the α-Al dendrites surround the hard phases.Figure4(c) displays the middle microstructure of a D5-U sample and the distribution of primary Si particles is relatively uniform.Figure4(e) shows the outside microstructure of D5-U sample, where the VF of the primary Si particles and the iron-rich phase is lower, and the dendritic α-Al is finer.Figure 4(f-j) demonstrates the radial section microstructure of D5-M samples from inside to outside. Figure 4(f) shows the inside microstructure of the D5-M sample, the α-Al surrounds the hard phase, and the eutectic Si particles are fine and fibrous.Figure 4(j) presents the outside microstructure of the D5-M sample; the VF of primary Si particles and iron-rich phases in it is lower than that in the middle microstructure of the D5-M sample.The dendritic α-Al is coarser.Figure4(k-o) shows the radial section microstructure of D5-B samples from inside to outside.There is no obvious agglomeration in the radial section microstructure.Figure4(k) demonstrates the inside microstructure of the D5-B sample, part of the massive iron-rich phases grow at the concave corner of Figure4(a-e) exhibits the radial section microstructure of D5-U samples from inside to outside.The hard phase particles (primary Si particles and iron-rich phase particles) are partially agglomerated, and the α-Al dendrites surround the hard phases.Figure4(c) displays the middle microstructure of a D5-U sample and the distribution of primary Si particles is relatively uniform.Figure4(e) shows the outside microstructure of D5-U sample, where the VF of the primary Si particles and the iron-rich phase is lower, and the dendritic α-Al is finer.Figure 4(f-j) demonstrates the radial section microstructure of D5-M samples from inside to outside. Figure 4(f) shows the inside microstructure of the D5-M sample, the α-Al surrounds the hard phase, and the eutectic Si particles are fine and fibrous.Figure 4(j) presents the outside microstructure of the D5-M sample; the VF of primary Si particles and iron-rich phases in it is lower than that in the middle microstructure of the D5-M sample.The dendritic α-Al is coarser.Figure4(k-o) shows the radial section microstructure of D5-B samples from inside to outside.There is no obvious agglomeration in the radial section microstructure.Figure4(k) demonstrates the inside microstructure of the D5-B sample, part of the massive iron-rich phases grow at the concave corner of

Figure 3 .
Figure 3.The schematic illustrations of the sample selection area.
Figure 4(o) shows the outside microstructure of the D5-B sample, where the VF of the hard phases drops significantly.Most of the microstructure consists of fine granular eutectic Si and fine near-spherical Al.

Figure 6 (
Figure 6(a-e) depicts the radial section microstructure of D9-U samples from inside to outside.The hard phases of the agglomeration in the radial microstructure can be divided.Figure 6(c) illustrates the middle

Figure 5 .
Figure 5. Characteristic parameters (equivalent diameter (a) and VF (b)) of hard phases and macro-hardness (c) in different regions of D5 samples.

Figure 8 (
Figure 8(a) shows the filling schematic of D5 samples.In the filling process, the maximum vertical average

Figure 7 .
Figure 7. Characteristic parameters (equivalent diameter (a) and VF (b)) of hard phases and macro-hardness (c) in different regions of D9 samples.
authors acknowledge funding for this research from National Natural Science Foundation of China [grant number 52065032], Basic Research Project of Yunnan Province [grant number 202101AT070123], Key Research and Development Project of Yunnan Province and International Science and Technology Cooperation Project [grant number 202103AF140004], Science and Technology Major Ten Thousand Talent Program of Yunnan Province [grant number YNWR-QNBJ-2019-106], Science and Technology Innovation Base Construction Project Fund [grant number 202207AB110003].This work is supported by the National and Local Joint Engineering Laboratory of Advanced Metal Solidification Forming and Equipment Technology, and Analytic and Testing Research Centre of Yunnan, Kunming University of Science and Technology, Kunming, China.