Quantitative Evaluation of Eutectic Si Phase Distributions and Effect on Mechanical Properties in JIS AC4CH Aluminum Casting Alloy

Abstract

The effect of eutectic Si particles distributions on ductility in JIS AC4CH aluminum casting alloy (referred to as AC4CH alloy) was investigated through experimental approaches. AC4CH alloys were subjected to ECAP processing at the 2nd, 4th, 6th, and 8th-pass to prepare samples with various distributions of eutectic Si particles. First, a method for the quantitative evaluation of the distribution of the eutectic Si particles using an area grid was examined. The eutectic Si particle distributions were quantified for each plane in the test pieces for unprocessed and ECAP processed samples. These quantitative values corresponded with the eutectic Si particle distributions visually observed on the microstructural images. The relationship between the quantified values of eutectic Si particle distributions and the elongations obtained by the tensile test was investigated. The uniform, local, and fracture elongations determined by the tensile test were found to increase as the number of ECAP passes increased. After 8th-pass ECAP processing, each of these types elongation increased by 49%, 64%, and 54%, respectively, compared to the unprocessed sample. The correlation was found between the mean quantified values of eutectic Si particle distribution and elongations. The experiments confirmed that the homogenization of the three-dimensional distribution of eutectic Si particles leads to the elongation increasing.

1. Introduction

Al–Si eutectic alloys are widely used in industrial products such as automotive parts because of their excellent castability and high strength. However, Al–Si eutectic alloys are known to have poor ductility due to a non-uniform microstructure such as a dendritic primary Al phase, a network structure of eutectic region, and casting defects such as shrinkage porosity. In recent years, the application of severe plastic deformation (SPD) such as equal-channel angular pressing (ECAP), friction stir processing (FSP), and accumulative roll-bonding (ARB) to Al–Si eutectic alloys to improve the ductility has been investigated.^1–3) The application of the SPD process leads to microstructural improvement due to the introduction of significant deformation by a plastic flow. The microstructural change of Al–Si eutectic alloys by SPD processes has been examined in detail,^3–10) and is summarized as follows.

1. (I)    Improvement of the casting defects such as shrinkage porosity
2. (II)    Refinement of the primary α-Al phases and eutectic Si phases
3. (III)    Homogenization of the eutectic Si phases distributions

It is suggested that the synergetic effect of the above microstructural changes enhances the ductility of the Al–Si eutectic alloys.

Many researchers have reported investigations of the effects of the casting defect and the morphology of primary α-Al phases and eutectic Si particles on the mechanical properties of Al–Si eutectic alloys, particularly the Al–Si–Mg alloys. In some of these studies, by quantifying the interested microstructure and examining the correlation with the mechanical properties, particularly the ductility, useful knowledge was obtained.^11–17) For example, Gokhale et al. reported the relationships between the total area fraction of casting defects such as oxide film and shrinkage porosity in the corresponding tensile fracture surfaces and the tensile ductility were shown by a simple equation.¹³⁾ The relationships between the size of eutectic Si particles that were quantified from microstructural images and the mechanical properties for JIS AC4C casting alloy were investigated by Fuji et al.¹⁵⁾ They found the particle size of eutectic Si particles to influence the elongation, impact value, and fatigue strength. Furthermore, the ductility was shown to increase with the refinement and spheroidizing of Si particles. On the other hand, regarding the distribution of eutectic Si particles, there are a few reports¹⁸⁾ in which the eutectic Si particles were quantified and the correlation with mechanical properties discussed. In this regard, the quantification of the eutectic Si particle distributions seems to be an issue. In the Al–Si–Mg alloys, in addition to the above microstructural factor, the solid solution state and morphology of precipitation related to Mg addition and the morphology of Fe-rich intermetallics also slightly influence on the ductility.¹⁹⁾ Therefore, although a relationship between the microstructure and the ductility in the Al–Si–Mg alloys is not simple, an examination of the effect of eutectic Si particle distributions on ductility is expected to provide useful knowledge on the microstructural improvement using an SPD process.

In this study, to understand the correlation between the distribution of eutectic Si particles and mechanical properties, the following investigation was conducted for a JIS AC4CH aluminum casting alloy (referred to as AC4CH alloy). First, the proposed method for the quantitative evaluation of the distribution of eutectic Si particles in AC4CH alloy is described. Samples with various distributions of eutectic Si particles were prepared by an ECAP process. The distribution of the eutectic Si particles was evaluated quantitatively using an area grid. Then, the quantitative evaluation of the relationship between the distributions of eutectic Si particles by the method based on the area grid is presented, after which the mechanical properties, particularly the ductility, that were obtained by the tensile tests, is discussed.

2. Experimental Details

An AC4CH alloy was used in this study. The melt at 700°C was poured into a JIS-type permanent mold held maintained at 150°C under atmospheric pressure. The analyzed chemical composition (mass%) of the ingot is provided in Table 1. The degassing was performed to the alloy by Ar gas to prepare for casting and Sr was added with the aim of modifying the morphology of the eutectic Si particles. Test pieces with a rectangular shape (40 mm × 15 mm × 5 mm) were machined from an ingot for the ECAP process. The shape and sampling position of the test piece are illustrated in Fig. 1.

Table 1 Analyzed chemical composition of sample (mass%).

Fig. 1

Schematic illustration showing specimen preparation.

The ECAP process was carried out at room temperature. In order to enhance the formability of the test piece during the ECAP process at room temperature, test pieces were preheated to 550°C for 2 hours and then the furnace was allowed to cool.²⁰⁾ The ECAP process was conducted using a die with a channel cross-sectional shape of 15 mm × 5 mm, a channel angle Φ of 120° and an outer corner angle Ψ of 16°. In this study, Route A for the processing of the ECAP process was selected and test pieces were pressed at the 2nd, 4th, 6th, or 8th pass. In the case of route A, a test piece is not rotated and pressed same direction between pressings. The test pieces for the ECAP process were lubricated by MoS₂ grease. The two sets of these test pieces were prepared and subjected to microstructure observation and a tensile test, respectively. An equivalent strain accumulated after N-pass through the die, ε_N, is calculated by the following equation.²¹⁾   

\begin{equation} \varepsilon_{N} = \frac{N}{\sqrt{3}}\left[2cot\left(\frac{\Phi}{2} + \frac{\psi}{2}\right) + \psi \text{cosec}\left(\frac{\Phi}{2} + \frac{\psi}{2}\right)\right] \end{equation} (1)

In the ECAP process, the strain hardening behavior of a test piece depends on the number of ECAP passes. However, it is desirable that the strength of the matrix phase is equivalent regardless of the number of ECAP passes since we are mainly interested in relationship between the distribution of the eutectic Si particles and the mechanical properties in this work. Therefore, in order to cancel the effect of strain hardening due to the ECAP process, the test pieces were heat-treated (annealed) for 30 min at 350°C after the ECAP process. The Vickers micro-hardness (measuring load is 1.96N) for before and after annealing of the test pieces subjected to the ECAP process are shown in Fig. 2 for reference. Before annealing the samples, the hardness of the test pieces after the ECAP process for the 2nd-pass substantially increases and then the hardness increases with an increasing number of ECAP passes. In contrast, the hardness of the test pieces after heat treatment is almost constant regardless of the number of ECAP processes. The heat treatment conditions of the test piece are summarized in Fig. 3. The tensile test and microstructure observation were carried out after annealing.

Fig. 2

Effect of annealing on Vickers hardness of unprocessed and ECAP processed test pieces.

Fig. 3

Summary of conditions for heat treatment of test piece.

An optical microscope was used for microstructure observation to evaluate the distribution of the eutectic Si particles. The position of the center of gravity of the eutectic Si particles was measured using image analysis to evaluate the distribution states of the eutectic Si particles. The method for evaluating the distributions of eutectic Si particles is described in detail in section 3.2.

Figure 4 illustrates the preparation of a tensile specimen. Two kinds of tensile specimens were sectioned from the test pieces using a wire-electrical discharge machine. The first specimen was prepared by ensuring that the loading direction during the tensile test and the extruded direction during the ECAP process are parallel (A) and the other was prepared by using perpendicular directions (B). The tensile test was carried out under conditions of an initial strain rate of 2.1 × 10⁻³ s⁻¹ at room temperature and tensile properties as 0.2% proof stress, and the fracture strength, uniform elongation, local elongation and fracture elongation were measured.

Fig. 4

Schematic illustration showing tensile specimen preparation.

3. Results and Discussion

3.1 Microstructural changes with ECAP process

The optical micrographs of the unprocessed and ECAP processed test pieces are shown in Fig. 5. The microstructure was observed for three planes in the test piece as illustrated in Fig. 5. In the unprocessed test piece, a typical cast microstructure composed of a primary α-Al dendritic structure and network structure of the eutectic cell containing Si particles was observed in all planes (Fig. 5(a)). The microstructure of the test piece after the 2nd-pass ECAP process, was deformed due to plastic deformation during the ECAP process (Fig. 5(b)). The primary α-Al dendritic structure and eutectic constituents were elongated along the Z-axis in the X-plane. In the Y-plane, the microstructure was elongated along the extruded direction. The microstructure in the Z-plane was similar to that of the unprocessed sample and has the appearance of a typical cast microstructure. In the Y-plane, the extent of structure elongation was increased as the number of ECAP passes increased. Moreover, the eutectic Si particles are arranged linearly along the extruded direction (Fig. 5(c)–(e)). On the other hand, the microstructures in the X-plane of the test pieces after the 4th of the ECAP process did not changed clearly compared with after the 2nd-pass ECAP process. The microstructure as the typical cast microstructure in the Z-plane has been maintained without relation to the number of ECAP passes.

Fig. 5

Optical micrographs showing effect of ECAP pass on microstructures a) Unprocessed, and ECAP processed at b) 2nd-pass, c) 4th-pass, d) 6th-pass and e) 8th-pass.

In order to understand the microstructural changes during the ECAP process, the shear deformation behavior of the test piece during the ECAP process has been surveyed.^22,23) For example, according to Furukawa et al.,²³⁾ when subjected to the ECAP process using Route A, the microstructure of the X-plane is elongated in the Z-axis direction. The microstructure in the Y-plane elongates to an angle with the extruded direction, and the inclination angle shifts to become parallel to the extruded direction by increasing the number of ECAP passes. On the other hand, the microstructure does not change significantly in the Z-plane. The microstructural changes shown in Fig. 5 were in close agreement with those reported by Furukawa et al.

3.2 Quantification of eutectic Si particle distributions

The mean free path and nearest particle distance are known as a parameter that can be used to evaluate the distribution of the secondary-phase particles in materials.²⁴⁾ The mean free path, f, is defined as the mean distance between two adjacent particles on a line drawing on the micrographs. For simplicity, assuming that the secondary-phase particles has constant radius, r, it is calculated as follows.   

\begin{equation} f = \frac{4r(1 - V_{V})}{3V_{V}} \end{equation} (2)

where V_v is the volume function of the secondary-phase particles. The nearest particle distance, δ, is calculated by the following equation.   

\begin{equation} \delta = \sqrt{\frac{2}{3}} r_{m}\left\{\left(\frac{\pi}{V_{V}}\right)^{1/2} {}-{} 2\right\} \end{equation} (3)

where r_m is the mean radius of secondary-phase particles. As can be understood from eq. (2) and (3), these parameters are determined by the particle radius and their volume fraction. In other words, these parameters are effective for evaluating the distribution when the particle radius and volume fraction of the secondary-phase particles are clearly changed. On the other hand, it is inferred that significant changes of the volume fraction of the eutectic Si particles does not occur by the introduction of plastic deformation. Therefore, it is difficult to evaluate the microstructural changes from a typical cast structure to a uniform structure in which the eutectic Si particles are uniformly distributed in the matrix from the view point of the mean free path and/or nearest particle distance. In recent year, it has been proposed that a method of evaluating a planar/spatial dispersion of secondary-phase particles from the number of particles contained in the measuring circle/sphere when a measuring circle/sphere is placed a gravity center of a noticed particle.²⁵⁾

In this study, the method using an area grid was examined to evaluate the changes of eutectic Si particle distributions by the ECAP process. A summary of this method is shown in Fig. 6. Firstly, micrographs were obtained by optical microscope and an area grid was drawn in these micrographs. Then, the number of eutectic Si particles that present inside in the grid was counted. In fact, the position of the center of gravity coordinates of the eutectic Si particles on the image was measured using image analysis. The number of eutectic Si particles contained in each grid was calculated by collating the center of gravity coordinate of eutectic Si particles and the occupied coordinate of each grid. Then, a histogram that shows the relationship between the number of eutectic Si particles located in the grid and the frequency of that grid was generated. When we applied the above process to a test piece with a typical cast structure, it was inferred that the grid, which does not include the eutectic Si particles are present in a relatively large number due to a typical cast structure in the form of a coarse dendritic primary Al phase. On the other hand, it is inferred that in the case of the test pieces with a uniform distribution of eutectic Si particles, the histogram shows a normal distribution because the number of eutectic Si particles located in the grid are close to a constant value. The arithmetic average (average value of the number of eutectic Si particles located in a grid), μ, and standard deviation (dispersion of the number of eutectic Si particles located in the grid), σ, were calculated from the histogram. In this study, the coefficient of variation (referred to as the CV value), which is obtained by the following eq. (4) was defined as a characteristic value representing the distribution of the eutectic Si particles.   

\begin{equation} \text{CV} = \frac{\sigma}{\mu} \end{equation} (4)

Since the CV value represents relative dispersion, it is suitable for comparing dispersion of groups with different average values. In fact, by using CV value, the dispersion of the number of eutectic Si particles located in the grid can be relatively compared even when the arithmetic average of the number of Si particles located in the grid, i.e., the number density of eutectic Si particles is different.

Fig. 6

Schematic illustration showing quantitative evaluation method of dispersed state of eutectic Si particles.

To evaluate the distribution of eutectic Si particles using the above method, it is important to select the size of the grid properly with respect to the micrograph size. Therefore, a preliminary investigation to determine the capturing image size and grid size was carried out. As a result, the image size was determined to be 102 µm × 136 µm in the case of present study. This image contains several dendritic primary Al phases and enables the eutectic Si particles to be correctly identified by using image analysis. In order to determine the grid size, the histograms of the unprocessed test piece and the test piece after 8th-pass ECAP processing were generated with several grid sizes in respects of the above image size. For a grid size of 20 µm × 20 µm, the shape of the histogram and the CV value were distinctly changed due to the changes in the distribution of the eutectic Si particles. From this result, the appropriate grid size to estimate the distribution of eutectic Si particles was determined as 20 µm × 20 µm in this study.

The distribution of the eutectic Si particles was evaluated for the unprocessed and ECAP processed test piece using the above method. Figure 7 shows the relationship between the number of eutectic Si particles located in a grid and the frequency of that grid for each plane in the unprocessed test piece and that after 2nd- and 8th-pass ECAP processing. The grids were described as four vertical and five horizontal blocks on a microstructural image that randomly captured the position in each plane of each test piece (see Fig. 6). These histograms were generated from 10 images per plane, in fact, the total number of grids was 200. In addition, the standard deviation, σ, and arithmetic average, μ, are also shown in the figure. First, the X-plane is explained. In the case of the unprocessed test piece, with an increasing number of eutectic Si particles located in a grid, the frequency of that grid was decreased and the shape of the histogram is downward sloping. After 2nd-pass ECAP, the number of grids containing a small number and no eutectic Si particles decreased. The shape of the histogram is that of a normal distribution and it shows the grid containing approximately 8 eutectic Si particles is most. This seems to be caused by a deformation of the coarse dendritic primary Al phase and eutectic regions by plastic flow during the ECAP process. The number of grids containing a small number and no eutectic Si particles is large when the microstructure containing coarse dendritic primary Al phase. The deformation of the coarse dendritic primary Al phase and eutectic regions resulted in decrease in spacing of eutectic regions. Therefore, the number of grids containing either a small number or no eutectic Si particles was decreased. The standard deviation, σ, was decreased by approximately 32% compared to the unprocessed test piece. After the 8th-pass ECAP process, the shape of the histogram becomes sharper and the standard deviation, σ, decreased by approximately 49% compared to the unprocessed test piece. In the Y-plane, the shape of the histogram for the unprocessed test piece was similar to the X-plane. The shape of the histogram changes to that of a normal distribution when increasing the number of ECAP passes. In the Z-plane, the shape of the histogram for the unprocessed test piece was similar to that in the other planes. On the other hand, the difference in the histogram shapes between unprocessed and ECAP processed pieces was not significant; furthermore, the histograms were consistently downward sloping.

Fig. 7

Effect of ECAP on dispersed state of eutectic Si particles in each plane of specimen.

The relationship between the number of ECAP passes and the CV value calculated from the standard deviation, σ, and the arithmetic average, μ, is shown in Fig. 8. The CV value of each plane for the unprocessed test piece is shows a relatively high, from 0.73 to 0.84. In the case of the X-plane, the CV value of the test piece after 2nd-pass ECAP processing decreased by approximately 40% compared to the unprocessed test piece, after which it became saturated. In the Y-plane, the CV value decreased monotonically by increasing the number of ECAP passes. The CV value after the 8th-pass ECAP process was reduced to approximately 56% of that of the unprocessed test piece. On the other hand, the effect of ECAP processing on the CV value in the Z-plane was not significant, and the CV values after ECAP processing were maintained at a high level.

Fig. 8

Effect of number of ECAP pass on CV value in each plane of specimen.

The CV value that shows a distribution of the eutectic Si particles defined in this study, because of its characteristic, is interpreted as indicating that, when this value is large, the variation of the number of eutectic Si particles located in the grid is large, in fact, the distribution is non-uniform. On the other hand, when the CV value is small, the distribution is interpreted to be uniform. The changes in the CV value with the number of ECAP passes (Fig. 8) closely coincides with the distribution changes of the eutectic Si particles as confirmed visually from the microstructural image shown in Fig. 5. In other words, the CV value defined in this study appropriately estimates the distribution of eutectic Si particles as a numerical value.

A point to notice is that the CV value also contains information on the distribution density of eutectic Si particles because it defined by the average value of the number of eutectic Si particles located in a grid, μ, and the dispersion of the number of eutectic Si particles located in the grid, σ. However, in this experimental result, since the difference in μ between each plane and number of ECAP passes is small, the CV value mainly represents the distribution of eutectic Si particles. It is assumed that the difference in μ was small in this experiment due to investigate under conditions such as composition and morphology of eutectic Si particles was limited.

3.3 Tensile properties

In this section, the situation where the loading direction during the tensile test and the extruded direction during ECAP processing are parallel (see Fig. 4(A)) is described. Figure 9 shows 0.2% proof stress and the ultimate tensile strength determined by the tensile test for unprocessed and ECAP processed specimens. In the 0.2% proof stress, although the value of the specimen subjected to 6th-pass ECAP was a little higher than that of other specimens, the difference between the unprocessed specimen and those after ECAP was not significant, with a mean value of approximately 78 MPa. The tensile strength is approximately 128 MPa at mean value under each condition and the difference between the unprocessed specimen and that after ECAP did not change significantly. Additionally, the behavior of 0.2% proof stress and tensile strength with the above-mentioned ECAP process were consistent with their hardness of after annealing shown in Fig. 2.

Fig. 9

Plots of 0.2% proof stress and tensile strength of after annealing against number of ECAP pass.

The uniform elongation, local elongation, and fracture elongation for the unprocessed and ECAP processed specimens are shown in Fig. 10. It is clear that each type of elongation increased by increasing the number of ECAP passes, and the test piece after ECAP processing at the 8th-pass increased by approximately 49%, 64% and 54%, respectively, than before ECAP. The uniform and fracture elongation of the specimen after 6th-pass ECAP processing were a little lower. This seems because the 0.2% proof stress at 6th-pass ECAP processing is a little higher than for the other specimens.

Fig. 10

Plots of elongation of after annealing against number of ECAP pass.

3.4 Effect of eutectic Si particles distributions on ductility

The relationship between the CV value for unprocessed and ECAP processed test pieces and the elongations (mean value) is shown in Fig. 11. This figure shows the case when the tensile direction and the extruded direction during ECAP process are parallel. In this case, the distribution of the eutectic Si particles in the X-plane was expected to dominate the ductility because the X-plane is perpendicular to the fracture surface. In the X-plane, although the CV value in this plane, denoted as CV_X, for ECAP processed specimens was significantly decreased compared with the unprocessed specimen, the changes in CV_X with respect to the number of ECAP passes was not clear. Corresponding to that behavior, the elongations of the ECAP processed specimens remarkably increased compared with the unprocessed specimen; however, a clear correlation between CV_X and elongations was not confirmed among specimens after ECAP processing specimens. The coefficient of correlation between each elongation and CV_X were 0.93, 0.89, and 0.90, respectively. On the other hand, in the Y-plane, each elongation was increased with decreasing the CV value in the Y-plane, CV_Y. The coefficient of correlation between each elongation and CV_Y was the highest among the three planes, which were 0.94, 0.89, and 0.92, respectively. It is suggested that there is a good correlation between the CV_Y and elongations in the Y-plane. In the Z-plane, the changes in the CV value in the Z-plane, CV_Z, of the ECAP processed specimen, also including the unprocessed specimen, were very small, therefore the relationship between each elongation and CV_Z could not confirmed. The coefficient of correlation between each elongation and CV_Z was the lowest among the three planes, which were 0.33, 0.15, and 0.55, respectively.

Fig. 11

Relationship between CV value in each plane and elongation.

It was expected that the distribution of eutectic Si particles in the X-plane dominate the ductility because the X-plane is perpendicular to the fracture surface. However, the result indicates that the highest correlation exists between the distribution of eutectic Si particles in the Y-plane parallel to the tensile direction and the elongation. To understand that issue, the fracture path of the tensile specimen was observed. The fractured specimens after the tensile test were observed for the Y and Z-plane parallel to the tensile direction using an optical microscopy. The result showed that a fracture path was formed intricately; in fact, a crack was propagating three-dimensionally with tensile deformation. Therefore, it is important that the distribution is not considered to be on a particular plane but three-dimensionally in order to discuss the effects of eutectic Si particle distributions on ductility. To understand the relationship between ductility and the three-dimensional distribution of eutectic Si particles, the mean CV value, CV_m, was calculated as the mean value of CV in the X-, Y- and Z-plane. Figure 12 shows each elongation against CV_m. It is clear from Fig. 12 that there is excellent correlation between each elongation and CV_m. The coefficient of correlation between each elongation and CV_m was 0.99, 0.98, and 0.92, respectively. This result indicated that the reduction of CV_m, in other words, homogenization of the three-dimensional distribution of eutectic Si particles, leads to an increase in the elongation.

Fig. 12

Relationship between mean CV value and elongations. Mean CV value CV_m was calculated as mean value of CV in X, Y and Z-plane.

In the presented alloy, it is suggested that the failure of materials and eutectic Si particles are closely related.^26–29) It is inferred from these studies that the fracture process develops as follows. First, when the material deforms, the stress concentrates in the eutectic regions in which eutectic Si particles with a complex morphology densely exist. The stress concentration leads to the nucleation of voids by the fracture of eutectic Si particles and/or the interfacial debonding between the eutectic α-Al and Si particles. As the deformation of materials develops further, their voids grow and then the micro-crack originated in the voids propagates in eutectic α-Al to interlink them with each other. Finally, the materials experience failure. Assuming that the fracture developed according to the above process, the reason for the elongation increasing with homogenization of the three-dimensional distribution of eutectic Si particles was interpreted as follows: the homogenization of eutectic Si particle distributions leads to the relaxation of stress concentration during tensile deformation. The relaxation of stress concentration reduces the nucleation of voids by fracturing the eutectic Si particles. Therefore, the uniform elongation was increased. In addition, the contribution of the aluminum matrix with large ductility was increased when propagation of the micro-crack resulting from the density of eutectic Si particles such as in the formation of eutectic regions was eliminated. Therefore, local elongation was increased.

According to Fig. 11 and Fig. 12, evaluation of the effect of eutectic Si particle distributions on the ductility makes it necessary to consider their three-dimensional distribution in material. To examine the validity of the above consideration, a tensile test was carried out with the tensile direction vertical to the extruded direction during the ECAP process, i.e., the fractured surface is formed on the Y-plane (See Fig. 4(B)). The relationship between the mean CV value, CV_m, of each plane in the specimen and the elongation is showed in Fig. 13. In this examination, the maximum number of ECAP passes was six. Figure 13 also shows the result of the tensile test with the tensile direction parallel to the extruded direction during the ECAP process. It is clear from Fig. 13 that each elongation was increased with decreasing CV_m regardless of the tensile direction, and the plot is approximately distributed on a straight line. This result, experimentally verified that the elongation and the CV_m, i.e., the three-dimensional distribution of eutectic Si particles, are closely related.

Fig. 13

Relationship between mean CV value and elongations for each tensile directions.

In this study, we focused on the effect of the distribution of eutectic Si particles on the tensile properties. However, it is inferred that a plastic flow by the ECAP process eliminates the casting defects such as shrinkage porosity and changes the morphology of eutectic Si particles.^5,6) The shrinkage porosity and morphology of eutectic Si particles has an influence on the tensile properties. Therefore, the effect of eutectic Si particle distribution on tensile properties with consideration of these influences needs to be examined. Additionally, a survey that verifies the superiority of the presented method that estimates the distribution of eutectic Si particles compared to other methods such as the mean free path and nearest particle distance will be implemented in the future.

4. Summary

In this work, the correlation between eutectic Si particle distributions and tensile properties, particularly the ductility, in JIS AC4CH aluminum alloy was investigated experimentally. The distribution of eutectic Si particles in a test piece was variously changed by ECAP processing (Route A). As a first step of this work, a method for quantitatively evaluating the eutectic Si particle distributions using an area grid was investigated. Then, the relationship between the eutectic Si particle distributions and tensile properties was discussed. The conclusions are as follows.

1. (1)    In this study, the coefficient of variation, CV, was defined as a characteristic value representing the distribution of the eutectic Si particles. In the X-plane, the CV_X of the test piece after 2nd-pass ECAP processing decreased by approximately 40% compared to the unprocessed test piece before becoming saturated. In the Y-plane, CV_Y decreased monotonically as the number of ECAP passes increased. On the other hand, in the Z-plane, the effect of ECAP processing on CV_Z was not significant. The change in the CV value with the EACP processing was consistent with the change in the distribution of the eutectic Si particles that visually confirmed by optical micrographs.
2. (2)    The uniform, local, and fracture elongation were increased by increasing the number of ECAP passes. After 8th-pass ECAP processing, each type of elongation increased by 49%, 64%, and 54%, respectively, compared to the unprocessed specimen. The relationship between the mean CV value, CV_m, i.e., the three-dimensional distribution of eutectic Si particles and elongations, was surveyed. The result showed that each type of elongation increased with decreasing CV_m.

Acknowledgments

The present study was financially supported in part by The Light Metal Educational Foundation, Inc.

REFERENCES

• 1)   Z.Y.  Ma,  R.S.  Mishra and  M.W.  Mahoney: Scr. Mater. 50 (2004) 931–935.
• 2)   J.M.  García-Infanta,  A.P.  Zhilyaev,  C.M.  Cepeda-Jimenez,  O.A.  Ruano and  F.  Carreno: Scr. Mater. 58 (2008) 138–141.
• 3)   R.  Jamaati,  S.  Amirkhanlou,  M.R.T.  Toroghinejad and  B.  Niroumand: Mater. Sci. Eng. A 528 (2011) 2495–2501.
• 4)   J.C.  Kim,  Y.  Nishida,  H.  Arima and  T.  Ando: Mater. Lett. 57 (2003) 1689–1695.
• 5)   A.  Ma,  N.  Saito,  M.  Takagi,  Y.  Nishida,  H.  Iwata,  K.  Suzuki,  I.  Shigematsu and  A.  Watazu: Mater. Sci. Eng. A 395 (2005) 70–76.
• 6)   A.  Ma,  K.  Suzuki,  Y.  Nishida,  N.  Saito,  I.  Shigematsu,  M.  Takagi,  H.  Iwata,  A.  Watazu and  T.  Imura: Acta Mater. 53 (2005) 211–220.
• 7)   M.L.  Santella,  T.  Engstrom,  D.  Storjohann and  T.Y.  Pan: Scr. Mater. 53 (2005) 201–206.
• 8)   Z.Y.  Ma,  A.R.  Sharma and  R.S.  Mishra: Metall. Mater. Trans. A 37 (2006) 3323–3336.
• 9)   Z.Y.  Ma,  A.R.  Sharma and  R.S.  Mishra: Mater. Sci. Eng. A 433 (2006) 269–278.
• 10)   I.  Gutierrez-Urrutia,  M.A.  Munoz-Morris and  D.G.  Morris: Acta Mater. 55 (2007) 1319–1330.
• 11)   C.H.  Cáceres and  B.I.  Selling: Mater. Sci. Eng. A 220 (1996) 109–116.
• 12)   J.A.  Francis and  G.M.D.  Cantin: Mater. Sci. Eng. A 407 (2005) 322–329.
• 13)   A.M.  Gokhale and  G.R.  Patel: Scr. Mater. 52 (2005) 237–241.
• 14)   C.D.  Lee: Mater. Sci. Eng. A 464 (2007) 249–254.
• 15)   N.  Fujii,  M.  Fujii,  S.  Morimoto and  S.  Okada: Imono 56 (1984) 387–392.
• 16)   Y.  Awano,  H.  Kawahara and  S.  Tamura: J. JFS 72 (2000) 465–470.
• 17)   S.  Iwasawa,  T.  Yamaguchi,  S.  Saikawa,  K.  Hayashi,  S.  Kamado and  Y.  Kojima: J. JFS 74 (2002) 291–297.
• 18)   Y.  Harada,  S.  Tamura and  S.  Kumai: J. JILM 61 (2011) 213–219.
• 19)   Q.G.  Wang: Metall. Mater. Trans. A 34 (2003) 2887–2899.
• 20)   Y.  Nakayama and  T.  Miyazaki: Mater. Trans. 51 (2010) 918–924.
• 21)   Y.  Iwahashi,  J.  Wang,  Z.  Horita,  M.  Nemoto and  T.G.  Langdon: Scr. Mater. 35 (1996) 143–146.
• 22)   Y.  Wu and  I.  Baker: Scr. Mater. 37 (1997) 437–442.
• 23)   M.  Furukawa,  Y.  Iwahashi,  Z.  Horita,  M.  Nemoto and  T.G.  Langdon: Mater. Sci. Eng. A 257 (1998) 328–332.
• 24)  The Japan Institute of Metals and Materials: Kinzoku Binran 6th edition, (Maruzen, Tokyo, 2000) p. 265.
• 25)   K.  Sugio,  Y.  Momota,  D.  Zhang,  H.  Fukushima and  O.  Yanagisawa: Mater. Trans. 48 (2007) 2762–2767.
• 26)   K.  Kobayashi,  H.  Shingu and  R.  Ozaki: J. JILM 20 (1970) 247–255.
• 27)   Q.G.  Wang and  C.H.  Caceres: Mater. Sci. Eng. A 241 (1998) 72–82.
• 28)   T.  Hidaka,  H.  Toda,  M.  Kobayashi,  K.  Uesugi and  T.  Kobayashi: J. JILM 58 (2008) 58–64.
• 29)   K.  Lee,  Y.N.  Kwon and  S.  Lee: J. Alloys Compd. 461 (2008) 532–541.