2. Materials and Methods
A mixture of commercial aluminum powders (with 99.9% purity and a particle size of <75 µm, Sigma-Aldrich, St. Louis, MO, USA) and silicon powders (with 99.9% purity and a particle size of <20 µm, Alfa Aesar, Haverhill, MA, USA) was milled in a planetary ball mill under an argon atmosphere to achieve an alloy with a composition of Al–5 wt %Si. The following parameters were used in the milling process: a ball-to-powder mass ratio of 12:1, a ball diameter of 10 mm, and a speed of 250 rpm. We added 1.5 wt % of stearic acid (CH3(CH2)16COOH) to the mixture as a process control agent (PCA) to moderate the cold welding process and to prevent the adhesion of the powder to the balls and interior surface of the milling tank. The mixture of powders and PCA was milled for up to 50 h, and samples were collected after 5, 20, and 50 h of milling. Mechanically alloyed powders were uniaxially cold-pressed in a cylindrical steel die (a diameter of 10 mm) for 30 s at 500 MPa pressure and then hot-pressed for 60 min at 450 °C under 500 MPa to achieve the consolidated samples. Crystallite size, lattice strain, and hardness of the consolidated samples were compared to their powder counterparts.
X-ray diffraction measurements were performed with a wide angle diffractometer in the θ–2θ step scan mode using CuKα radiation. Scans were collected over a 2θ range of 20–90° with a step of 0.01°. The crystallite size (d
) and the equivalent lattice strain (ε
) were determined using the Williamson-Hall (W-H) line broadening analysis method [14
] according to the following:
is the full width at half maximum (FWHM) of the diffraction peak, θ
is the diffraction angle, k is a constant whose values is approximately 0.9, and λ is the incident X-ray wavelength (1.54060 Å).
It is clear from Equation (1) that β cosθ vs. sinθ curve exhibits a straight line with the slope of (2ε) and intercept of kλ/d. Thus, the crystallite size (d) and the lattice strain (ε) can be achieved.
Scanning electron microscope (SEM) equipped with an energy dispersive X-ray spectrometer (EDX, VEGA/TESCAN, Kohoutovice, Czech Republic) was used in characterizing the microstructure of the samples as well as analyzing the elemental distribution. Additionally, to evaluate the hardness of MA-prepared powders, a micro-hardness test was utilized using a Vickers indenter under the load of 10 gf and dwell time of 15 s (MVK-H21, Akashi Co., Hyogo, Japan). For harness measurements of consolidated samples, a load of 25 gf and a dwell time of 15 s was used. Measuring the hardness of powders is a challenging issue and usually results in scattered values dependent on the method of the measurement, on the properties of the mounting resin, and on the “depth” of the particle under indenter [15
]. In this study, we used hot epoxy resin to mount the cold-pressed powders prior the micro-hardness measurements. The excess of the resin was removed from the top layer, exposing the surface of powders to air. The micro-hardness of the powders was estimated by the evaluation of the depth of the indentation. The choice of low load for the hardness measurements on powdered samples was to reduce the indenter impression and its depth, allowing reasonable hardness measurements on small powders. A similar method was employed earlier by Abdoli et al. to evaluate the micro-hardness of nanostructured composite powders produced by MA [16
]. For better statistics of the hardness values, the hardness tests were performed on three individual points on each consolidated sample and on 10 points for each powdered sample. Moreover, compression testing on consolidated sample was carried out at a cross-head speed of 0.2 mm/min and a strain rate of 0.005 min−1
. Prior to the compression tests, both ends of the specimens were polished to make them parallel to each other.
shows the changes in the morphology of mixed powders with increasing milling times. As Figure 1
shows, increasing milling time gradually changes the powder morphology from flat flakes to semi-globular structures. As is expected, in preliminary stages of milling (Figure 1
a) aluminum particles are still soft, and they undergo plastic deformation while brittle silicon particles are fragmented. The Si phase is believed to accelerate the deformation of the powder through a second hard phase formation in the mill [17
]. Increasing the milling time results in the sequential welding of the aluminum particles and the distribution of silicon particles within the aluminum matrix. With the continuation of milling (Figure 1
b), various factors such as deformation, welding, and solid dispersion lead to work hardening of aluminum particles, which in turn enhances their tendency to fracture. Thus, aluminum particles break down, and the average particle size slightly decreases. Finally, welding and fracture processes reach an equilibrium (Figure 1
c) where randomly orientated boundaries are formed. In this stage, particles becomes semi-globular in shape, and their size distribution becomes more uniform.
shows an individual semi-globular particle formed after 50 h of milling. As can be seen from this figure, the surface of the semi-globular particle exhibits some moderate roughness as a consequence of simultaneous competition of welding and fracturing mechanisms. Although Figure 2
exhibits a typical particle shape after 50 h of milling, as can be observed in Figure 1
c, there are instances of a transverse crack through the whole particle is observed, most likely due to the increase in the internal stresses and lattice strains within the semi-globular particles, which will be discussed in the following.
From XRD analysis, the crystallite size and lattice strain were measured, and the obtained values are summarized in Table 1
and Table 2
for the powder and consolidated samples, respectively. In all cases a decreasing trend in the crystallite size is observed upon increasing the milling time. As proposed by Miraghaei et al. [18
], the formation of new defects, especially dislocations, is responsible for the reduction of the crystallite size. Multiple mechanisms were proposed for the accumulation of dislocations. For instance, dense regions of these dislocations can be formed in sub grains, dislocations might pile up at the grain boundaries, or clusters can be accommodated within the crystallites. Overall, a severe plastic deformation during mechanical alloying and consequently the reduction of the crystallite size can contribute to the generation of extra dislocations. Moreover, the results in Table 1
and Table 2
indicate that the lattice strain is generally enhanced as the milling time increases. As mentioned before, the severe plastic deformation during the milling process introduces dislocations, vacancies, impurities, and other lattice defects, which, in turn, increase the stress field in the alloys. Similar trends were observed for mechanically produced Al–Mg/Al2
nanocomposites, Al/Fe alloys, and alumina-reinforced nanocrystals [19
In the current study, the initial growth of the average particle diameter due to a flattening of the particles, similar to the observations by Milligan et al. [17
], is neglected as it is out of the scope of this work. However, a partial crystallite growth is observed for the consolidated samples during the hot pressing stage. Nevertheless, the extensive crystal growth is still retarded due to the pinning effects at the grain boundaries [22
]. Explaining such trends for consolidated samples requires a better understanding of the elemental composition distribution, which will be discussed later. As is depicted in Figure 3
, the variation of the crystallite size with the increase in milling time exhibits very similar trends for powder and consolidated samples, showing that, after almost 20 h of milling, the size of crystals do not vary any more for either sample.
To understand how the distribution of Si in the aluminum matrix affects the crystallite size and the growth rate by the pinning mechanism, EDX measurements are performed on selected consolidated samples. As can be seen in Figure 4
, Si distribution in the Al matrix is more uniform after a prolonged milling time (i.e., 20 h), causing a better pinning of the grain boundaries, which in turn prevents the crystalline growth.
The distribution of Si in the Al matrix as well as the residual internal stress within the crystals affect the hardness of the powders and consolidated samples. Variation in the hardness of the Al–5 wt %Si powders and consolidated samples as a function of milling time is shown in Figure 5
. Both crystallite size reduction and elemental dispersion strengthening affect the hardness of samples. Based on these results, the measured hardness for both powdered and consolidated samples is almost monotonically enhanced when the milling time is increased, reaching its maximum value at around 50 h of milling time. However, the consolidated samples show greater absolute hardness values compared to their powder counterparts most likely due to the distribution of the internal stress within their structure (see Table 1
and Table 2
). It is worth noting that microhardness measurements performed on powder particles are inherently prone to systematic errors. Therefore, microhardness value comparisons between the powder specimens and the consolidated specimens should be considered approximations at best. Despite almost a constant grain size after 20 h of milling (results from Figure 3
), overall hardness values increase when the milling time is increased. These phenomena point to the fact that, in addition to grain reinforcement, certain mechanisms, such as work hardening of the fragile fragments in the initial stages of milling, the homogeneous distribution of Si atoms in the Al matrix, and its possible oversaturation, lattice microstrains, as well as the formation and distribution of defects within newly formed crystals, affect the final mechanical properties of the alloys. However, quantification of the fractal contributions of these mechanisms is out of the scope of the current manuscript.
presents yield and compressive strengths of consolidated samples for various milling time. Similar to hardness test results (results from Figure 5
), strength values increase with milling time reaching a maximum yield strength after 50 h of milling. It is known that the grain boundaries result in a higher yield stress than the matrix itself, since grain boundaries are able to prevent dislocation movement [21
]. In other words, dislocations pile up behind grain boundaries resulting in a stress concentration. Since increasing the milling time reduces the grain size, it will also strengthen the material, resulting in an increase in its yield stress. Furthermore, a better distribution of silicon in the aluminum matrix as well as work hardening can contribute to the increased yield strength of consolidated samples as the milling time increases [23