j m a t e r r e s t e c h n o l .

2 0 2 0;9(x x):13646–13660

Available online at www.sciencedirect.com

https://www.journals.elsevier.com/journal-of-materials-research-and-technology

Original Article

Microstructure, mechanical properties and

machinability of particulate reinforced Al matrix
composites: a comparative study between SiC
particles and high-entropy alloy particles

Tiwen Lu a,b , Tianbing He b , Zixuan Li a , Hongyu Chen b , Xiaoliang Han b , Zhiqiang Fu a,∗ ,
Weiping Chen a,∗
a Guangdong Key Laboratory for Advanced Metallic Materials Processing, South China University of Technology, Guangzhou, Guangdong
510640, China
b IFW Dresden, Institute for Complex Materials, Helmholtzstraße 20, D-01069 Dresden, Germany

a r t i c l e i n f o a b s t r a c t

Article history: In this study, 2024Al matrix composites reinforced by SiC particles (SiC-2024Al) and
Received 4 July 2020 nanocrystalline high-entropy alloy particles (HEA-2024Al) fabricated by powder metal-
Accepted 7 September 2020 lurgy were systematically compared for the first time. There is a significant difference in
microstructure and mechanical properties as well as machinability between two kinds of
composites. In term of microstructure, when the volume fraction of reinforcements was 10%,
Keywords: both SiC-2024Al and HEA-2024Al composites showed a homogeneous particle distribution in
Al matrix composites the matrix. With the increase of reinforcement content, HEA-2024Al composites presented
SiC particle denser microstructure than that of SiC-2024Al composites. The composites with 10, 20 and
High-entropy alloy particle 30 vol.% HEA reinforcements all showed better plasticity than that of the SiC-2024Al com-
Microstructure posites with same volume fraction of reinforcements, which was related with better particle
Mechanical properties distribution and interface bonding. However, the strength showed the opposite tendency in
Machinability the two kinds of composites. Selecting 10SiC-2024Al and 10HEA-2024Al composites as exam-
ples to explore the difference in the yield strength of two kinds of composites, it is ascribed to
the dislocation punched zones around interface between the Al matrix and reinforcements,
which was analyzed in detail by a combination of calculation, nanoindentation tests and
finite element analysis. Additionally, HEA-2024Al composites showed better machinability
than those of SiC-2024Al composites. This work provides insight into the application of
particulate reinforced Al matrix composites.
© 2020 Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND
license (http://creativecommons.org/licenses/by-nc-nd/4.0/).

∗
Corresponding authors.
E-mails: kopyhit@163.com (Z. Fu), mewpchen@scut.edu.cn (W. Chen).
https://doi.org/10.1016/j.jmrt.2020.09.034
2238-7854/© 2020 Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/
licenses/by-nc-nd/4.0/).
 j m a t e r r e s t e c h n o l . 2 0 2 0;9(x x):13646–13660 13647

composites reinforced with ceramic particles and composites

1. Introduction reinforced with metallic particles have rarely been stud-
ied. Secondly, whether the machinability of composites can
Aluminum matrix composites (AMCs) are widely used in var-
be improved by using metallic particles to replace ceramic
ious engineering applications [1,2]. Among them, ceramic
particles. To explore the two aspects, silicon carbide (SiC)
particulate reinforced aluminum matrix composites are one
and nanocrystalline high-entropy alloy (NC-HEA) particles
of the most promising engineering materials, ascribed to
were selected as the reinforcements. 10 vol.%, 20 vol.% and
their outstanding thermal properties, mechanical properties
30 vol.% particulate reinforced 2024Al matrix composites have
and cost effectiveness [3–5]. Due to the addition of hard
been well prepared to study the difference in microstructure,
ceramic particles, the modulus and strength are significantly
mechanical properties and machinability of two kinds of com-
improved. However, ceramic particles are harder and stiffer
posites.
than the matrix, resulting in results that machining becomes
more difficult than that of conventional materials [3]. Some
researchers [6] mentioned that adding high-stiffness ceramic 2. Experimental
whiskers into the ductile matrix resulted in thermal mismatch
and plastic incompatibility, severely decreased the forgeabil- 2.1. Materials preparation
ity of the composite. How to improve the machinability of
composites is always a tough task [6–9]. Due to the high cost The irregular SiC particle with an average diameter of 14.7 mm,
of polycrystalline diamond (PCD) tools [7], generally, there is supplied by ChaoXing Nanotechnology Co., Ltd. China, was
a clear and tangible demand for cheaper and efficient turn- selected as the reinforcement. The other reinforcement,
ing of MMCs. Additionally, many researches show that both nanocrystalline CoNiFeCrAl0.6 Ti0.4 high-entropy alloy particle,
extrinsic (cutting speed, feed rate, depth of cut and type of was synthesized by ball milling mixtures of Al, Fe, Ni, Co, Cr
cutting tools) and intrinsic (particulate size, volume fraction) and Ti elemental powders (purity higher than 99.6 wt.%) for
parameters affects machinability of particle reinforced MMCs 60 h at a speed of 320 rpm with process control agent of cyclo-
[7–9]. Among all parameters, the type of reinforcement is one hexane, which were dried in vacuum chamber. The size of final
of important factors. NC-HEA powders was controlled less than 25 ␮m (500 mesh).
Good mechanical properties can also be achieved when The specific preparation process has been widely described
metallic particles are introduced into Al matrices. This is in our previous work [15]. The reinforcement shape was not
mainly due to the stable metallurgical bonding formed at taken in account in this work. Gas-atomized 2024Al powders
the interface between the Al matrix and metallic particles, were used as the matrix. The morphology and size distribution
and adequate load bearing ability of metallic particles [10–12]. of the above powders are shown in Fig. 1. The two reinforce-
High-entropy alloy (HEA) particles have been one of the ments were mixed with the 2024Al powders and subsequently
most frequently used reinforcements by powder metallurgy, milled at a speed of 200 rpm with a ball-to-powder weight
squeeze casting and friction stir processing (FSP) [13–16]. Meng ratio of 10:1 for 10 h, respectively. The ball milling experi-
et al. [13] prepared AlCoCrCuFeNi high-entropy alloy parti- ments were carried out in a high energy planetary ball mill
cle reinforced magnesium matrix composite by laser melt (QM-3SP2 Planetary Ball Mill). A vacuum hot pressing device
injection on the surface of magnesium alloy, and the wear (HP-12 × 12 × 12, Centorr Vacuum Industries, USA) was utilized
of 40 vol.% high entropy alloy particle reinforced magnesium to sinter the composite powders at 480 ◦ C for 0.5 h with a
matrix composite coating was only one seventh of the matrix. pressure of 150 MPa, and subsequently the bulk composites
Yang et al. [14] added AlCoCrCuFeNi high-entropy particles were hot extruded at 450 ◦ C with an extrusion ratio of 10:1.
to the AA5083 matrix by using the underwater friction stir Al matrix composites with 10 vol.%, 20 vol.% and 30 vol.% of
processing, and found that the grains of the prepared com- reinforcements were successfully fabricated (hereafter named
posite were obviously refined to obtain better strength and 10HEA-2024Al, 20HEA-2024Al, 30HEA-2024Al; 10SiC-2024Al,
plasticity. Nanocrystalline high-entropy alloy (NC-HEA) parti- 20SiC-2024Al, 30SiC-2024Al, respectively).
cles can be prepared by mechanical alloying, and keep good
thermal stability during the processing [15,16]. More impor-
2.2. Materials characterization
tantly, due to the addition of NC-HEAp , the composites show
improved strength and balanced plasticity [15]. However, it is a
The size distribution of the powders was characterized using
short time when researchers used the metallic particles as the
a Mastersizer 2000 laser diffraction system (Malvern, UK). The
reinforcements, and no work has been done about the machin-
microstructure of the composites was characterized by scan-
ability of composites reinforced by metallic particles. There
ning electron microscopy (SEM, NOVA NanoSEM 430, USA) and
is also few literature about comparisons between composites
high resolution transmission electron microscopy (HRTEM,
reinforced with ceramic particles and composites reinforced
G2 F20) equipped with an energy dispersive spectrometer
metallic particles. Selvakumar et al. compared Mop /6082Al
(EDS). TEM specimens were firstly polished to ∼50 ␮m and
[17] and TiCp /6082Al [18] composites in mechanical properties,
then thinned to electron transparency by ion milling. The
Mop /6082Al composites displaying better elongation than that
nanoindentation measurements were done by a Hysitron TI-
of TiCp /6082Al composites. However, the comparison men-
950 equipped with a standard Berkovich diamond indenter
tioned above is based on the different process methods.
tip, in the load-controlled mode. Prior to the nanoindentation
In view of above discussions, this work was motivated
tests, the specimen surface was carefully prepared using met-
by the following factors: firstly, the comparisons between
allographic techniques and subsequently polished with firstly
13648 j m a t e r r e s t e c h n o l . 2 0 2 0;9(x x):13646–13660

Fig. 1 – Morphology and size distribution of powders: (a) SiC; (b) HEA; (c) 2024Al; (d) particle size distribution.

2.5 ␮m, 1 ␮m and then 0.25 ␮m diamond abrasive for 20 min. respectively. Then, the inserts were cleaned in alcohol (99%) by
A 1 ␮m spacing between neighboring indents and a 500 ␮N ultrasonic. The wear morphologies of inserts and chips were
maximum indentation load were used in the following exper- observed by SEM.
iments. Both tensile tests (the cylindrical specimens with a
15 mm gauge length and 3 mm in diameter) and compression
tests (3 mm × 4.5 mm) were conducted using an INSTRON 3. Results
5869 testing system (USA, INSTRON) at a speed of 5 × 10−3 s−1
and 1 × 10−3 s−1 , respectively. 3.1. Microstructure

Fig. 2 is the polished cross-section of the as-extruded compos-

2.3. Machinability tests ites as a function of the volume fraction of reinforcements. In
term of SiC-2024Al composites, the 10SiC-2024Al composite
The turning experiments were carried out on a CNC shows a uniform distribution of SiC particles in the matrix.
CA6150i lathe (DMTG Co., China) under dry condition with- However, some particle agglomeration (red dashed circles)
out lubricant and coolant. The cemented carbide inserts and pores (black arrows) can be found in the 20SiC-2024Al
(YBC251/TNMG160404-PM) were embedded in the tool sub- and 30SiC-2024Al composite. Experiments show that non-
strate. Turning experiments was conducted with a cutting uniformity arises when the content of the SiC particulate
speed of 10 m/min, feed rate of 0.1 mm/r and a depth-of-cut reinforcement increases. All HEA-2024Al composites show
of 0.05 mm. For the same cutting parameter, every insert was denser microstructure than those of SiC-2024Al composites,
taken down from the tool substrate after the cylindrical sam- which is related with the characteristic that HEA particles
ples (ϕ20 mm × 30 mm) machined with a length of 10 mm, have more deformability than that of SiC particles during the
 j m a t e r r e s t e c h n o l . 2 0 2 0;9(x x):13646–13660 13649

Fig. 2 – SEM images of hot extruded composites along the extrusion direction (ED): (a, c, e) 10SiC-2024Al, 20SiC-2024Al and
30SiC-2024Al composites, respectively; (b, d, f) 10HEA-2024Al, 20HEA-2024Al and 30HEA-2024Al composites, respectively.

Fig. 3 – The distribution of reinforcement size in as-extruded composites: (a) 10SiC-2024Al composite, (b) 10HEA-2024Al
composite.

hot pressing and hot extrusion. The 30HEA-2024Al composite structure might be finally transformed into layered structures
presents some layered structures made up of sub-micron HEA after the hot pressing and hot extrusion. Notably, particle size
particles. Our former results reported that the hierarchical statistics based on SEM images, the average particle sizes of
microstructure has been found in the HEA-2024Al compos- 10SiC-2024Al and 10HEA-2024Al composites are 6.96 ␮m and
ite powder, which displayed a core consisting of relatively 6.04 ␮m (Fig. 3), respectively, which indicates that there is still
large HEA particles embedded in the Al matrix and a shell no obvious difference in two reinforcement particle size after
composed of many fine HEA particles [15]. When the frac- ball milling. Fig. 4 shows XRD patterns of the composites. It can
tion volume of HEA particles reaches 30 vol.%, the core–shell be seen that there are the Al, Al2 Cu and Al7 Cu2 Fe in both com-
13650 j m a t e r r e s t e c h n o l . 2 0 2 0;9(x x):13646–13660

Fig. 4 – XRD patterns of as-extruded composites: (a) SiC particle reinforced Al matrix composites, (b) HEA particle reinforced
Al matrix composites.

Fig. 5 – Mechanical properties of two kinds of hot extruded composites: (a) engineering tensile stress-strain curves; (b)
engineering compressive stress-strain curves; (c) comparisons of ultimate tensile strength (UTS), yield strength (YS) and
fracture strain; (d) comparisons of compressive yield strength and strain.

posites. The Al2 Cu phase was normally formed during the hot 3.2. Mechanical properties
pressing [19]. Notably, the intensity of Al2 Cu in the SiC-2024Al
composites is significantly higher than that of the HEA-2024Al Figs. 5(a) and (c) show the engineering tensile properties
composites. of two kinds of composites. With the increase of rein-
forcement content, the yield strength and fracture strain of
 j m a t e r r e s t e c h n o l . 2 0 2 0;9(x x):13646–13660 13651

Fig. 6 – SEM images showing the fracture surfaces of two kinds of composites with the different reinforcement content: (a)
10SiC-2024Al composite, (b) 20SiC-2024Al composite, (c) 30SiC-2024Al composite, (d) 10HEA-2024Al composite, (e)
20HEA-2024Al composite, (f) 30HEA-2024Al composite.

two kinds of composites display increased and decreased ratio of fractured HEA particles (marked by red arrows in
tendency, respectively. It is worth mentioning that all SiC- Fig. 6(e) and (f)), which is attributed to nonhomogeneous par-
2024Al composites show higher tensile strength than those ticle distribution resulting in stress concentration around HEA
of HEA-2024Al composites with same volume fraction of rein- reinforcements [11]. Fig. 7 shows the lateral surface of the
forcement, whereas the fracture strain shows the opposite compressed composite samples after polishing. Because of the
tendency. The ultimate tensile strength in SiC-2024Al compos- good ductility, both the 10SiC-2024Al and 10HEA-2024Al com-
ites decreases with the increase of the reinforcement content, posites have no cracks in the samples after a 50% deformation.
which might be ascribed to defects and particle aggregations. There are two cracks in the 20SiC-2024Al composite along the
Tension–compression (T/C) asymmetry is controversial in Al diagonal direction, while only one crack near the surface of
matrix composites reinforced by metallic particles [12,13]. 20HEA-2024Al sample can be found, which can illustrate that
Therefore, the compressive properties of composites are mea- the ductility of the latter is better than the former. There are
sured, shown in Fig. 5(b) and (d). The overall tendency in the two long cracks in the 30SiC-2024Al composite along the diag-
yield strength is consistent with tensile properties, except for onal direction, and there are also three short cracks present
two composites reinforced with 30 vol.% reinforcements. The in the 30HEA-2024Al composite. It is worth mentioning that
layered structure in the 30HEA-2024Al composite is the rea- the cracks in the 30HEA-2024Al composite seem to originate
son for higher compressive yield strength than that of the from the layered structure (inset in Fig. 7(f)), which indicates
30SiC-2024Al composite. that the layered structure in the 30HEA-2024Al composite is
harmful for the ductility of composites.

3.3. Fracture morphology

3.4. Crack propagation
For the tensile fracture of SiC-2024Al composites (Fig. 6(a)–(c)),
all of them show fractured SiC particles, marked by blue In order to investigate concrete factors affecting the mechan-
arrows, similar with reports in former work [20,21]. Frac- ical behavior in the SiC-2024Al and HEA-2024Al composites,
tured SiC particles are mainly ascribed to residual stress in crack propagation paths were obtained in the compressive
the SiC particle due to its coefficient of thermal expansion tests of 20SiC-2024Al and 20HEA-2024Al composites, shown
(CTE) far different from that of the matrix [21]. However, in in the SEM images (Fig. 8). Evidently, the crack propagation
the 10HEA-2024Al composite, interface debonding between paths exhibit significant difference in the two composites. In
the Al matrix and HEA particles is the main characteristic the SiC-2024Al composite, SiC particles show obvious inter-
(Fig. 6(d)). Besides, the surfaces of HEA particles have some facial debonding with the matrix along the crack (Fig. 13(a)),
matrix dimples, which is related with good interface bonding and there is no matrix attached to the surfaces of SiC par-
and high-strength nanocrystalline HEA particles [15]. Increas- ticles. Besides, many fractured SiC particles (marked by blue
ing the HEA content, however, the composites present a larger arrows) can also be seen on both sides of the crack. In con-
13652 j m a t e r r e s t e c h n o l . 2 0 2 0;9(x x):13646–13660

Fig. 7 – The cross-section of composites after compression tests: (a) 10SiC-2024Al composite, (b) 10HEA-2024Al composite,
(c) 20SiC-2024Al composite, (d) 20HEA-2024Al composite, (e) 30SiC-2024Al composite, (f) 30HEA-2024Al composite.

trast, the crack deflects when it encounters HEA particles in the increase of volume fraction, the condition become worse.
the 20HEA-2024Al composite, shown in Fig.13(b), which indi- Because the properties of the SiC particle have a huge differ-
cates that the HEA particles can hinder the crack propagation ence with that of the Al matrix, the hot deformation behavior
and raise the damage tolerance of HEA/2024Al composites. of the SiC/Al composite is different from that of traditional
Along the cracks, fractured HEA particles are hardly seen, and metallic materials. Adding high-hardness SiC particles into
some matrix can be seen on the surface of HEA particles. The the ductile matrix increases the resistance to hot plastic defor-
further observation reveals that the crack propagation path mation and then brings the risk of damage during hot working.
in the 20HEA-2024Al composite has a more winding shape The higher the volume fraction of SiC particles is, the worse
in comparison to that in the 20SiC-2024Al composite. Based the deformability of composites is. Fig. 10 shows the com-
on the above observations, it can be concluded that the inter- parison in machinability between SiC-2024Al and HEA-2024Al
face structure in HEA/2024Al composites is better than that composites. Fig. 10(a) presents the schematic diagram of dry
of SiC/2024Al composites, supporting the better plasticity of turning experiments. Reinforcement types have a significant
HEA/2024Al composites. influence on the machined surface of materials. As shown
in Fig. 10(b), with increasing the volume fraction, compos-
3.5. Machinability ites reinforced by SiC particles break off at the middle and
edge of the workpieces due to the presence of hard abra-
Fig. 9 presents the surfaces of hot extruded composite bars. sive particles and the high pressure generated at the cutting
When the volume fraction of reinforcements reaches 10%, zone. By contrast, the surfaces of three HEA-2024Al compos-
both surfaces of the two kinds of composites are smooth. ites are flat and smooth. From the images of flank faces of
20HEA-2024Al and 30HEA-2024Al composites also show good tools after turning experiments in Fig. 10(c), the tool wear
surfaces. However, there are many cracks on the surfaces of of SiC-2024Al composites is significantly severer than that
both 20SiC-2024Al and 30SiC-2024Al composites, and with of HEA-2024Al composites. Besides, worn areas of tools used
 j m a t e r r e s t e c h n o l . 2 0 2 0;9(x x):13646–13660 13653

Fig. 8 – Crack propagation paths in the two composites based on compressive tests: (a) the 20SiC-2024Al composite, (b) the
20HEA-2024Al composite.

in the 10HEA-2024Al, and short ear-type chips were pro-

duced during the turning of 20HEA-2024Al and 30HEA-2024Al
composites. Prior studies discussed the effect of machining
conditions on chip formation for metal-matrix composites
[22], which demonstrated that the increased volume fraction
of hard reinforcements reduced ductility of the composites,
and the composites favored chip breakage when compared to
the machining of the matrix. Generally speaking, HEA-2024Al
composites have better machinability than SiC-2024Al com-
posites.

3.6. TEM observations

To explore the difference in mechanical properties between

SiC-2024Al and HEA-2024Al composites, grain size and inter-
face structure were observed using TEM (Figs. 11 and 12).
Fig. 11(a)–(d) show the shape and size distribution of grains in
Fig. 9 – The surfaces of hot extruded composite bars. 10SiC-2024Al, 10HEA-2024Al, 20SiC-2024Al and 20HEA-2024Al
composites, respectively. Based on about 300 grains, it is
found that the grain size of composites with the same rein-
in SiC-2024Al composites increase with the increase of the forcement content is close. When the volume fraction of
reinforcement content. There are limited tool wear in the reinforcements increases from 10 vol.% to 20 vol.%, the grain
HEA-2024Al composites with different reinforcement con- size of composites almost falls by half. Fig. 11(e)–(h) display
tent. Shown in Fig. 10(d), short C-type chips were formed microstructure around reinforcements in these four compos-
as observed in the 10SiC-2024Al composite and 20SiC-2024Al ites, and dislocation zones can be easily observed around
composite. Many bulk chips are found in the 30SiC-2024Al the SiCp /Al interface in the SiC-2024Al composites (Fig. 11(e)
composite, which ascribed to the 30SiC-2024Al composite con- and (g)). Besides, the length of dislocation zones increase
taining higher amount of hard and wear-resistant particles, with increasing the reinforcement content. As the micro-
difficult for turning. Long spring-type chips were obtained zone of the HEA-2024Al composite shown in Fig. 10(f) and
13654 j m a t e r r e s t e c h n o l . 2 0 2 0;9(x x):13646–13660

Fig. 10 – The comparison in machinability between SiC-2024Al and HEA-2024Al composites: (a) schematic diagram of dry
turning experiments, (b) the surface condition of composites after machining, (c) SEM images of flank faces of cutting tools
after the turning experiments (inset is original condition), (d) chips of different composites.

Fig. 11 – TEM images about grains and interface conditions of composites: (a) and (e) 10SiC-2024Al composite, (b) and (f)
10HEA-2024Al composite, (c) and (g) 20SiC-2024Al composite, (d) and (h) 20HEA-2024Al composite.
 j m a t e r r e s t e c h n o l . 2 0 2 0;9(x x):13646–13660 13655

Fig. 12 – The interface between the Al matrix and reinforcements in 10SiC-2024Al and 10HEA-2024Al composites: the
bright-filed image of the interface in the 10SiC-2024Al (a) and 10SiC-2024Al (c) composites, element distribution along the
linear scanning of SiC/Al (b) and HEA/Al (d) interfaces.

(h), the dislocation density around the interface between the 3.7. Hardness variation in the vicinity of interface
Al matrix and HEA particles is relatively low. This behavior
is related with the smaller CTE difference between the Al In the former literature [23,24], the measured hardness is
matrix and HEA particles than that between the Al matrix dependent on the grain size and grain orientations to some
and SiC particles. Additionally, HEA particles possess the bet- extent. Therefore, several arrays of indentation points in
ter damage tolerance and fracture toughness than those of random direction were normally conducted. Selecting 10HEA-
SiC particles [14]. Selecting 10SiC-2024Al and 10HEA-2024Al 2024Al and 10SiC-2024Al composites as examples, at least
composites as examples to investigate the interface in two 3 arrays of indentation across interface were operated, and
kinds of composites, the interface length and interface struc- 20 individual indents for each indentation array were per-
ture are also completely different. Observed from the line formed, shown in Fig. 13(a) and (d). Fig. 13(b) and (e) show
scanning in Fig.12, only a thin oxide layer can be seen in the the distribution of hardness and reduced modulus across the
interface between the Al matrix and SiCp, while a 200 nm inter- interface of composites. The red dashed lines corresponds to
facial layer contains Al, Cu, Co and Ni aggregation between the approximate position of the interface in Fig. 12(b) and
the Al matrix and HEA. Interface consumes some Cu, lead- (d), the right of which corresponds to the Al matrix, and the
ing to the result that the density of the Al2 Cu phase in left of which represents the reinforcements. In the Fig. 12(b),
HEA-2024Al composites is relatively lower than that of SiC- in the left of interface, the plateau has the magnitude of
2024Al composites, which were verified by the XRD patterns ∼28 GPa and ∼10 GPa, corresponding to the hardness of the
(Fig. 4). SiCp and HEAp , respectively. In the left regime, there are
13656 j m a t e r r e s t e c h n o l . 2 0 2 0;9(x x):13646–13660

Fig. 13 – Variation of indentation hardness and reduced modulus along the particle-matrix interfaces in the 10SiC-2024Al
and 10HEA-2024Al composites: (a) and (d) optical images of 10SiC-2024Al and 10HEA-2024Al composites, (b) variation of
hardness, (e) variation of reduced modulus, (c) and (f) typical P–h curves of the reinforcement, transition region and matrix
in 10SiC-2024Al and 10HEA-2024Al composites.

two distinct regions, which shows different hardness values strength between 10HEA-2024Al and 10HEA-2024Al compos-
reduction from the interface to the Al matrix in the two com- ites. According to the calculation formula for strengthening
posites. Regarding the 10SiC-2024Al composite, the hardness mechanisms [4], when the reinforcement particle size, volume
generally shows a gradual reduction from the interface to the fraction and grain size are nearly equal, the strength incre-
Al matrix, and the width of this transition region is about ments from Hall–Petch strengthening, load transfer effect and
3 ␮m. When the indentations are more than 3 ␮m away from Orowan strengthening should show no obvious difference.
the SiC/Al interface, the hardness values reach a plateau of Therefore, the difference in strength could only come from dis-
∼2 GPa. However, for the 10HEA-2024Al composite, the transi- location (GND) strengthening. It has been proposed by Ashby
tion zone is not obvious, and the width may be smaller than [26] that there are two possible sources of the GND strength-
2 ␮m. The variation of reduced modulus is similar with that ening. The first category is the CTE mismatch between the
of hardness values, shown in Fig. 12(e). The fluctuations of Al matrix and reinforcements induced GNDs when the com-
the hardness around the interface is due to the presence of posite cools down from the processing temperature. Another
dislocations, demonstrating that the larger punched zones category is the geometrical mismatch, and it is a result of the
exist in SiC/2024A composites, which is one of reasons for deformation-induced plastic strain gradient that arises when
the higher yield strength of 10SiC-2024Al composite than that the composite is plastically deformed. Based on the equations
of the 10HEA-2024Al composite [25]. Fig. 13(c) and (f) display in the literature [16], the strengthening values from the geo-
typical load-displacement curves of the reinforcement, tran- metrical mismatch for two composites are also nearly equal.
sition region and matrix in 10SiC-2024Al and 10HEA-2024Al Therefore, for 10HEA-2024Al and 10HEA-2024Al composites,
composites. the difference in strength is mainly reflected in the CTE mis-
match between the Al matrix and reinforcements, calculated
as [27]:
4. Discussion
√
CTE = ωGM bM GND (1)
The strength increment in these composites can be explained
by several contributions: Hall-Petch strengthening (␴H–P ), load 12T · ˛ · f
CTE = (2)
transfer effect (␴L–T ), Orowan strengthening (␴Or ), as well bM d(1 − f )
as geometrically necessary dislocation strengthening (␴GND )
[11,24]. Because of the relatively homogeneous microstruc- where ω is the a constant of 1.25 [28], and GND is the density
ture, we select 10HEA-2024Al and 10HEA-2024Al composites of GNDs; GM and bM are the shear modulus and Burgers
as targets to study the difference in the strengthening vector of the Al matrix, respectively; d and f are the radius
mechanism. There is a difference of 29.8 MPa in the yield and volume fraction of reinforcements. ˛ is the difference
 j m a t e r r e s t e c h n o l . 2 0 2 0;9(x x):13646–13660 13657

Fig. 14 – Finite element analysis used to evaluate the residual thermal stress of composites: (a) stress field of the
10SiC-2024Al composites, (b) stress field of the 10HEA-2024Al composites, (c) the model used in the finite element analysis,
(d) the stress distributed along the matrix-reinforcement interface, the line marked in (c).

in the coefficient of thermal expansion (CTE) between the respectively. Next, the calculation and the simulation are used
Al matrix and the reinforcements, CTEAl = 22.8 × 10–6 /◦ C, to compare the length of dislocation punched zones between
CTESiC = 4.7 × 10–6 /◦ C and CTEHEA = 14.7 × 10–6 /◦ C [16,20]. both composites. The Eqs. (3)–(5) are usually utilized to calcu-
Finally, the strength of the 10HEA-2024Al and 10HEA-2024Al late the length of dislocation punched zones [11,30].
composites from the the CTE mismatch are 15.3 MPa and
⎧ ⎫
21.4 MPa, respectively, which is the biggest difference in the ⎪⎛  ⎞ 13 ⎪
⎪
⎨ B (1 − 2Pf ) + B2 (1 − 2Pf )2 + 16 ⁄
ym G PB
⎪
⎬
strength increment in the two composites in comparison m

with other strength mechanisms [26–28]. a=d ⎝ ⎠ −1

⎪
⎪ 4( ym ⁄Gm ) ⎪
⎪
Taya et al. [29] proposing when the thermal stress exceeded ⎩ ⎭
the yield strength of the matrix, the dislocation loops near the (3)
interface are punched out into the matrix with a punching
distance to relax the thermal stress and then form a disloca-
tion punched zone in the matrix. The nanoindentation results
show that for the 10SiC-2024Al and 10HEA-2024Al composites,
(1 + m ) · CTE · T
the width of hardness transition region is ∼3 ␮m and ∼2 ␮m, B= (4)
(1 − m)
13658 j m a t e r r e s t e c h n o l . 2 0 2 0;9(x x):13646–13660

from the interface, which is the reason that the length of the
dislocation punched zone measured by nanoindentation tests
is obviously lower than theoretical and numerical simulated
values. Nonetheless, the tendency is consistent in three meth-
ods. It is worth mentioning that the residual thermal stress
in SiC particles is significantly higher than that of HEA parti-
cles, as shown in Fig. 14(c) and (d), which is also related with
the condition that the SiC particles were mostly fractured in
the 10SiC-2024Al composite, and the HEA particles only show
interface debonding in the 10HEA-2024Al composite. Shortly,
the difference in the dislocation punched zones is the decisive
reason for the fracture mechanism and strengthening effect
of the two composites.
To describe the comprehensive comparisons between SiC-
2024Al and HEA-2024Al composites, a schematic radar chart
Fig. 15 – Radar chart of comprehensive comparisons is presented in Fig. 15. We take the 10SiC-2024Al and 10HEA-
between two kinds of composites, taking the 10SiC-2024Al 2024Al composites for example. The mechanical properties
and 10HEA-2024Al composites for example. of two composites are presented numerically, while the
value of microstructure homogeneity, interfacial bonding and

2(1 − m )(3 + 2G)

P=  1+
 (5)
(1 − m) (1 − f ) (3 + 2G) + 3[f (3 p + 2Gp ) ( 1− m
m
) + (1 − f ) (3 m + 2Gm )]

Here, subscripts ‘‘m” and ‘‘p” stand for the matrix and particle,
respectively. In Eqs. (3)–(5), CTE is the difference in the CTE machinability are not quantitative and mainly deduced from
between the matrix and the particle, T is the temperature SEM and TEM observation and turning experiments. Seen from
change, ␶ym is the shear yield strength of the matrix, which Fig. 15, the 10 HEA-2024Al composite shows better overall per-
can be identified as the frictional stress for the glide motion formance than that of the 10SiC-2024Al composite. When the
of dislocations, and is assumed to be constant without con- volume fraction of reinforcements increase, the advantage in
sidering the work hardening effect [9], ␭ and G is the Young microstructure homogeneity and machinability will be more
modulus and shear modulus, ␥m is the Poisson’s ratio, and obvious. The parameters of NC-HEA reinforcements (particle
␭ = ␭p −␭m and G = Gp −Gm are the mismatch of the Lame size, the density, etc.) and interface diffusion degree are not
constants [30]. Finally, the length of dislocation punched zones optimized yet, and it is believed that the mechanical prop-
for 10SiC-2024Al and 10HEA-2024Al composites are calculated erties of the HEA/Al composites could be further improved
to ∼8 ␮m and ∼5 ␮m, which are notably larger than results of when these parameters reach optimal value. Another problem
the nanoindentation tests. is the high cost and complex preparation process of nanocrys-
Finite element model (FEM) has been a common method talline high-entropy alloy particles, which will be improved
to evaluate the deformation mechanism and residual thermal in the future work. In other filed, such as wear-resistant and
stress [31–33]. FEM was used here to evaluate the dislocation electronic packaging industries, the SiC-2024Al composites
punched zone in the thermal residual stress of two compos- maybe show better properties. In a word, above results and
ites during the cooling process using ANSYS Workbench 15 discussions provide some experience to design composites for
software and a thermo-mechanical coupled model. The simu- different structural applications.
lation results in Fig. 14(a) and (b) qualitatively show the stress
distribution surrounding various particulates. We assume that
both composites have the homogenous microstructure and 5. Conclusion
same particle size, and the simulation models of 10SiC-2024Al
and 10HEA-2024Al composites are simplified as Fig. 14(c). In this study, microstructure, mechanical properties and
Three features in two composites are directly observed as fol- machinability of SiC particulate reinforced 2024Al and
lows: (i) the thermal stress concentration takes place at the NC-HEA particulate reinforced 2024Al composites are system-
interface between the Al matrix and reinforcements, (ii) the atically studied. In terms of microstructure, when the volume
thickness of the stress zone surrounding a circle particle is fraction of reinforcements is 10%, the SiC-2024Al and HEA-
proportional to the particle size, (iii) most important of all, 2024Al composites show a uniform particle distribution in
the stress value in the 10SiC-2024Al composite is near two the matrix. With increasing the reinforcement content, HEA-
times than that of the 10HEA-2024Al composite. A line show- 2024Al composites show denser microstructure than those
ing stress distribution across the interface from the Al matrix of SiC-2024Al composites. The HEA-2024Al composites show
towards the reinforcements are presented in Fig. 14(d), the better plasticity than SiC-2024Al composites, which is related
length with the large stress fluctuation which can be regarded with interface bonding condition of two kinds of compos-
as the dislocation punched zones in 10SiC-2024Al and 10SiC- ites. However, the strength show the opposite results. The
2024Al composites are ∼5 ␮m and ∼4 ␮m. The variation of CTE mismatch between the Al matrix and reinforcements
hardness is not obvious in the dislocation punched zone away is the decisive factor for the difference in the strength of
 j m a t e r r e s t e c h n o l . 2 0 2 0;9(x x):13646–13660 13659

the two composites. Last but not least, the HEA-2024Al com- [14] Yang X, Dong P, Yan Z, Cheng B, Zhai X, Chen H, et al.
posites show better hot-workability and turning performance AlCoCrFeNi high-entropy alloy particle reinforced 5083Al
than those of the SiC-2024Al composites. Above all, the rein- matrix composites with fine grain structure fabricated by
forcement type is an important parameters to design the submerged friction stir processing. J Alloy Compd
2020;836:155411.
composites for structural applications.
[15] Lu T, Chen W, Li Z, He T, Li B, Li R, et al. Processing and
mechanical properties of fine grained Al matrix composites
reinforced with a uniform dispersion of nanocrystalline
Conflicts of interest high-entropy alloy particles. J Alloy Compd 2019;801:
473–7.
The author declares no conflicts of interest. [16] Chen W, Li Z, Lu T, He T, Li R, Li B, et al. Effect of ball milling
on microstructure and mechanical properties of 6061Al
matrix composites reinforced with high-entropy alloy
particles. Mater Sci Eng A 2019;762:138116.
Acknowledgments
[17] Selvakumar S, Dinaharan I, Palanivel R, Ganesh Babu B.
Characterization of molybdenum particles reinforced Al6082
The authors would like to acknowledge the Guangzhou Scien- aluminum matrix composites with improved ductility
tific Research Foundation (No. 2018100210). Tiwen Lu, Tianbing produced using friction stir processing. Mater Charact
He and Xiaoliang Han thank the support of the China Schol- 2017;125:13–22.
arship Council. [18] Thangarasu A, Murugan N, Dinaharan I, Vijay S. Synthesis
and characterization of titanium carbide particulate
reinforced AA6082 aluminium alloy composites via friction
references stir processing. Arch Civ Mech Eng 2015;15:324–34.
[19] Jin P, Xiao BL, Wang QZ, Ma ZY, Liu Y, Li S. Effect of solution
temperature on aging behavior and properties of
SiCp /Al-Cu-Mg composites. Mater Sci Eng A
[1] Ni D, Wang J, Zhou Z, Ma Z. Fabrication and mechanical
2011;528(3):1504–11.
properties of bulk NiTip /Al composites prepared by friction
[20] Lu T, Scudino S, Chen W, Wang W, Li D, Mao M, et al. The
stir processing. J Alloy Compd 2014;586:368–74.
influence of nanocrystalline CoNiFeAl0.4 Ti0.6 Cr0.5
[2] Suryanarayana C, Al-Aqeeli N. Mechanically alloyed
high-entropy alloy particles addition on microstructure and
nanocomposites. Prog Mater Sci 2013;58(4):383–502.
mechanical properties of SiCp /7075Al composites. Mater Sci
[3] Ma X, Zhao Y, Tian W, Qian Z, Chen H, Wu Y, et al. A novel Al
Eng A 2018;726:126–36.
matrix composite reinforced by nano-AlNp network. Sci Rep
[21] Liu Y, Yang C, Chen W, Zhu D, Li Y. Effects of particle size and
2016;6:34919.
properties on the microstructures, mechanical properties,
[4] Li Y, Liu W, Ortalan V, Li W, Zhang Z, Vogt R, et al. HRTEM and
and fracture mechanisms of 7075Al hybrid composites
EELS study of aluminum nitride in nanostructured Al
prepared by squeeze casting. J Mater Sci 2014;49(22):
5083/B4 C processed via cryomilling. Acta Mater
7855–63.
2010;58(5):1732–40.
[22] Dabade U, Joshi S. Analysis of chip formation mechanism in
[5] Yang H, Jiang L, Balog M, Krizik P, Schoenung J.
machining of Al/SiCp metal matrix composites. J Mater
Reinforcement size dependence of load bearing capacity in
Process Technol 2009;209:4704–10.
ultrafine-grained metal matrix composites. Metall Mater
[23] Oliver W, Pharr G. An improved technique for determining
Trans A 2017;48(9):4385–92.
hardness and elastic modulus using load and displacement
[6] Bai W, Roy A, Sun R, Silberschmidt V. Enhanced
sensing indentation experiments. J Mater Res
machinability of SiC-reinforced metal-matrix composite
2011;7(6):1564–83.
with hybrid turning. J Mater Process Technol 2019;268:
[24] Guo X, Guo Q, Nie J, Liu Z, Li Z, Fan G, et al. Particle size effect
149–61.
on the interfacial properties of SiC particle-reinforced
[7] Bhushan R, Kumar S, Das S. Effect of machining parameters
Al-Cu-Mg composites. Mater Sci Eng A 2018;711:
on surface roughness and tool wear for 7075 Al alloy SiC
643–9.
composite. Int J Adv Manuf Technol 2010;50(5–8):459–69.
[25] Song J, Guo Q, Ouyang Q, Su Y, Zhang J, Lavernia E, et al.
[8] Ozben T, Kilickap E, Cakır O. Investigation of mechanical and
Influence of interfaces on the mechanical behavior of SiC
machinability properties of SiC particle reinforced Al-MMC. J
particulate-reinforced Al-Zn-Mg-Cu composites. Mater Sci
Mater Process Technol 2008;198:220–5.
Eng A 2015;644:79–84.
[9] Sekhar R, Singh T. Mechanisms in turning of metal matrix
[26] Ashby M. The deformation of plasticity non-homogeneous
composites: a review. J Mater Res Technol 2015;4:
materials. Philos Mag 1970;21:399–424.
197–207.
[27] Zhang Z, Topping T, Li Y, Vogt R, Zhou Y, Haines C, et al.
[10] Dong S, Zhou J, Hui D, Wang Y, Zhang S. Size dependent
Mechanical behavior of ultrafine-grained Al composites
strengthening mechanisms in carbon nanotube reinforced
reinforced with B4C nanoparticles. Scr Mater
metal matrix composites. Compos A Appl Sci Manuf
2011;65(8):652–5.
2015;68:356–64.
[28] Jiang L, Yang H, Yee J, Mo X, Topping T, Lavernia E, et al.
[11] Shao J, Xiao B, Wang Q, Ma Z, Yang K. An enhanced FEM
Toughening of aluminum matrix nanocomposites via spatial
model for particle size dependent flow strengthening and
arrays of boron carbide spherical nanoparticles. Acta Mater
interface damage in particle reinforced metal matrix
2016;103:128–40.
composites. Compos Sci Technol 2011;71(1):39–45.
[29] Taya M, Lulay K, Lloyd D. Strengthening of a particulate
[12] Wang G, Zuo L. Strengthening mechanism of load sharing of
metal matrix composite by quenching. Acta Metall Mater
particulate reinforcements in a metal matrix composite. J
1991;39:73–87.
Mater Sci 2007;42(12):4215–26.
[30] Lin K, Pang S. The influence of thermal residual stresses and
[13] Meng G, Lin X, Xie H, Yue T, Ding X, Sun L, et al. The effect of
thermal generated dislocation on the mechanical response
Cu rejection in laser forming of AlCoCrCuFeNi/Mg composite
of particulate-reinforced metal matrix nanocomposites.
coating. Mater Des 2016;108:157–67.
Compos B Eng 2015;83:105–16.
13660 j m a t e r r e s t e c h n o l . 2 0 2 0;9(x x):13646–13660

[31] Liu L, Yang C, Zhang W, Xiao Z, Zhang L. Correlation between [33] Bao Y, Liu C, Huang J. Effects of residual stresses on strength
microstructure and deformation mechanism in and toughness of particle-reinforced TiN/Si3 N4 composite:
Ti66 Nb13 Cu8 Ni6.8 Al6.2 composites at ambient and elevated theoretical investigation and FEM simulation. Mater Sci Eng
temperatures. Mater Sci Eng A 2019;767:138448. A 2006;434(1–2):250–8.
˛
[32] Weglewski W, Michał B, Marcin C, Katarzyna P. Modeling of
thermally induced damage in the processing of Cr-Al2 O3
composites. Compos B Eng 2012;43(2):255–64.