Journal of Alloys and Compounds 769 (2018) 848e857 Contents lists available at ScienceDirect Journal of Alloys and Compounds journal homepage: http://www.elsevier.com/locate/jalcom Variation of microstructure and mechanical properties with nano-SiCp levels in the nano-SiCp/AlCuMnTi composites Jianyu Li, Shulin Lü, Shusen Wu*, Wei Guo, Fei Li State Key Lab of Materials Processing and Die & Mould Technology, Huazhong University of Science and Technology, Wuhan 430074, PR China a r t i c l e i n f o a b s t r a c t Article history: Received 9 June 2018 Received in revised form 5 August 2018 Accepted 7 August 2018 Available online 10 August 2018 Nano-sized SiC particles (SiCnp for short) reinforced Al-5Cu-0.5Mn-0.15Ti composites have been successfully fabricated by a process consisting of high energy ball milling, mechanical stirring and ultrasonic treatment for composite slurry and squeeze casting. The variation of microstructure and mechanical properties with SiCnp levels in xSiCnp/AlCuMnTi (x ¼ 0.5, 1, 1.5, 2 wt%) composites is investigated for the ﬁrst time. The SiCnp are uniformly distributed in the SiCnp/Al-5Cu composites made with the process. With the increase of SiCnp content from 0.5 to 2.0 wt%, the primary a-Al and Al2Cu phases are reﬁned signiﬁcantly up to 1.5 wt% but then become coarser at 2.0 wt%. The optimal mechanical properties are obtained in 1.5 wt% SiCnp/AlCuMnTi composites, which exhibit 298 MPa in ultimate tensile strength (UTS), 178 MPa in yield strength (YS) and 12.9% in elongation. These properties are increased by 18.7%, 11.3% and 25.3%, respectively, compared with the AlCuMnTi matrix alloy. The enhancement of strength is attributed to four strengthening mechanisms, among which the DsCTE and DsOrowan are the most important contributors. It is noteworthy that with more SiCnp in the present composites, the strength increases while elongation increases as well. © 2018 Elsevier B.V. All rights reserved. Keywords: Aluminum matrix composites Nano-sized SiC particles Ultrasonic treatment SiCnp content Mechanical properties 1. Introduction Aluminum matrix composites (AMCs) have recently attracted much attention due to their high speciﬁc strength, high elastic modulus and good wear resistance, etc. [1e4]. In general, AMCs are reinforced by various ceramic particles such as SiC, Al2O3 and TiC, among which SiC particle is regarded as a suitable reinforcement in aluminum matrix for its unique physical and mechanical properties [1e5]. Recent studies reveal that the nano-sized ceramic particle is more favorable than micron-sized ceramic particle and it is becoming a research hotspot in the metal matrix composites [6,7]. However, the SiCnp tend to agglomerate in the molten Al alloys due to their poor wettability, attractive Van Der Waals interactions and large surface-to-volume ratio , which can reduce the mechanical properties of composites. Therefore, it is essential to develop new preparation processes to improve the distribution of nano-sized reinforcements. Up to now, AMCs have been fabricated using a variety of conventional fabrication methods including solid-state processing and * Corresponding author. E-mail address: [email protected] (S. Wu). https://doi.org/10.1016/j.jallcom.2018.08.066 0925-8388/© 2018 Elsevier B.V. All rights reserved. liquid-state solidiﬁcation processing, such as powder metallurgy and stir casting [8,9]. In molten-metal processes, ultrasonic treatment (UT) is a promising technology to prepare the SiCnp reinforced AMCs, for the ultrasonic vibration can give rise to great effects on the melt by the cavitation and acoustic streaming [10e12]. In order to facilitate more homogenous dispersion of SiCnp in the SiCnp/ AlCuMnTi composites, other processes including high energy ball milling (HEBM) and squeeze casting are also needed. However, it has not been publically reported about the preparation of SiCnp/ AlCuMnTi composites using the similar processes. In general, SiCnp content also determines the mechanical properties of composites [8,13]. Wang et al. found that increasing SiCnp content in aluminum alloy led to an increase in UTS and YS of composites fabricated by semisolid stirring assisted with hot extrusion, but the elongation was decreased by 54.4% . Yao et al. found that increasing the volume fraction of SiCnp in AA6063 alloy caused an increase in UTS and YS of composites fabricated by powder metallurgy, but the elongation to fracture decreased from 10.0% to 2.3% . These researches indicated that more SiCnp led to higher UTS and YS, while lower elongation was obtained. Up to now, few researches have been carried out on the variation of microstructure and mechanical properties with SiCnp levels in SiCnp/AlCuMnTi composites fabricated by the similar processes. J. Li et al. / Journal of Alloys and Compounds 769 (2018) 848e857 Therefore, it is signiﬁcant to study the contributions of SiCnp content and various strengthening mechanisms to the increase of strength and elongation of SiCnp/AlCuMnTi composites. In this study, the microstructure and mechanical properties of SiCnp/AlCuMnTi (Hereafter, Al-5Cu for short of the AlCuMnTi alloy) composites were investigated, which were successfully fabricated by a process consisting of HEBM, UT for composite slurry and squeeze casting. The SiCnp content various from 0.5 wt% to 2 wt% in order to evaluate the effects of it. Signiﬁcantly, it is found that with more SiCnp in the present composites, the strength increases while elongation increases as well. Various strengthening mechanisms are discussed to analyze this behavior in detail. 2. Experimental procedure 2.1. Materials The preparation process for (0.5, 1, 1.5, 2) wt.% SiCnp/Al-5Cu composites is as following developed by the authors. Firstly, preoxidation treatment was used for SiCnp at 850 C for 2 h to form a SiO2 coating layer about 3.6 nm in thickness, which would prevent the reaction between Al and SiCnp. The details of pre-oxidation treatment could be found in our previous study . After preoxidation of SiCnp, the SiCnp/Al compound granules containing SiCnp were prepared by HEBM according to our previous work . The mass fractions of SiCnp was 6 wt% in compound granules with the size of 1~2 mm, as shown in Fig. 1(a). Furthermore, the SiCnp were uniformly distributed in compound granules, which would be beneﬁcial for the dispersion of SiCnp in the composites, as shown in Fig. 1(b). The chemical compositions of the Al-5Cu alloys used as matrix alloy were 5 wt% Cu, 0.5 wt% Mn, 0.15 wt% Ti, 0.3 wt% Mg, 0.15 wt% Fe, and balance Al. The raw materials including pure Al, pure Mg, pure Cu, Al-5%Ti-B master alloy and Al-10%Mn master alloy were melted in a graphite crucible at 750 C. Then the molten metal was degassed with pure argon gas for 10 min. After that, the SiCnp/Al compound granules were added to melts with mechanical stirring at 120 rmp for 10 min, which was beneﬁcial to accelerate the melting of SiCnp/Al compound granules and promote the dispersion of SiCnp in melt. The whole process was protected by argon gas. The SiCnp content in the composites was controlled to 0, 0.5, 1, 1.5, and 2 wt%, respectively. Then, the melt was treated by UT to assist the homogenous dispersion of SiCnp. The details of UT system could be found in our previous study . The ultrasonic horn preheated for 5 min at 720 C was inserted into the melt below surface at 10e15 mm, and the UT time was set at 5 min. The whole process of UT was 849 protected by argon gas. Finally, the composite ingots with diameter of 30 mm and height of 100 mm were fabricated by squeeze casting. The squeeze pressure was set at 50 MPa. 2.2. Characterization Specimens for metallographic observation were cut from the top of casting ingots. The microstructure characterization was performed by using a XRD-7000S X-ray diffraction (XRD), a DMM480C optical microscopy (OM), a JSM-7600F scanning electron microscopy (SEM) equipped with an energy dispersive spectrometer (EDS) and a Tecnai G2-F30 telecom electron micrograph (TEM). The grain size of the primary a-Al phase was calculated by using a self-developed software system (Solidvf 3.0) with Heyn's linear intercept method . Then the room temperature tensile properties were measured by a SHIMADZU AG-IC machine under a constant rate of 1 mm/min. The average UTS, YS and elongation values were obtained from three samples for each speciﬁed condition. The size of sample was shown in Fig. 2. 3. Results and discussion 3.1. XRD analysis of SiCnp/Al-5Cu composites The XRD patterns of xSiCnp/Al-5Cu (x ¼ 0.5, 1, 1.5, 2 wt%) composites are shown in Fig. 3, which exhibits the peaks for SiCnp, a-Al and Al2Cu phase. It indicates that the SiCnp have been successfully introduced into the melt by the novel process of UT combined with HEBM and squeeze casting. No peak for Al4C3 can be found, which suggests that the pre-oxidation can efﬁciently prevent the reaction between SiCnp and Al-melt. In addition, the intensity of SiCnp peak is very low because of the low content (2 wt% for maximum). Fig. 2. The draft of the tensile test sample. Fig. 1. Composite granules after HEBM: (a) morphology, (b) SiCnp distribution in compound granules. 850 J. Li et al. / Journal of Alloys and Compounds 769 (2018) 848e857 Fig. 5. The average grain size of a-Al. Fig. 3. XRD spectrums of SiCnp/Al-5Cu composites: (a) 0.5 wt%, (b) 1.0 wt%, (c) 1.5 wt%, (d) 2.0 wt%. 3.2. Effects of SiCnp content on microstructure of SiCnp/Al-5Cu composites Fig. 4 shows the optical microstructures evolution of composites with different SiCnp content. As can be seen, the SiCnp content plays a signiﬁcant role in controlling the grain size of primary a-Al. The quantitative analysis result of grain size is shown in Fig. 5. Obviously, the grain size decreases with the increase of SiCnp content from 0 to 1.5 wt%. However, when the SiCnp content increases to 2.0 wt %, the grain size is no longer further reduced. That may be caused by the agglomeration of SiCnp, which can weaken the effect of SiCnp on hindering the growth of primary a-Al grains [10,15]. Although the UT can homogenize the distribution of SiCnp in the composites, the effects of ultrasonic vibration are weakened due to the high viscosity of the composite melt with 2 wt% SiCnp. Therefore, the optimal SiCnp content is found to be around 1.5 wt%, in which the grain size is decreased by 55% compared with the Al-5Cu Fig. 4. Optical microstructures of the SiCnp/Al-5Cu composites with different SiCnp content: (a) 0.5 wt%, (b) 1.0 wt%, (c) 1.5 wt%, (d) 2 wt%. J. Li et al. / Journal of Alloys and Compounds 769 (2018) 848e857 matrix alloy. The ﬁner grains generate more grain boundaries leading to the reﬁnement of phases in boundary, which will be beneﬁcial to improve the mechanical properties of composites . In this regard, the reﬁnement of a-Al grains in the composites can be attributed to the following reasons: (i) the pinning effect of SiCnp located at the grain boundaries, hindering the growth of grains; (ii) the cavitation and acoustic streaming of UT, creating nuclei and enhancing reﬁnement of grains . Fig. 6 represents the SEM microstructures of SiCnp/Al-5Cu composites with different SiCnp content. As shown in Fig. 6(a), coarse light-gray phases can be found at the grain boundaries in the composites containing 0.5 wt% SiCnp, and they are Al2Cu according to the EDS results shown in Fig. 7(e). These Al2Cu phases have two shapes including dot and bar, which are marked with orange rectangles and circles, respectively. The Al2Cu phases are mainly in the form of bar and the amount of dot shape is small. With the increase of SiCnp content, the Al2Cu phases are signiﬁcantly reﬁned. With addition of 1 wt% SiCnp, the Al2Cu phases with small size are obtained and they are mainly in the form of dot, as shown in Fig. 6(b). The amount of dot phases increases signiﬁcantly, but there are still a small amount of bar phases in the composites. When the SiCnp content increases to 1.5 wt%, the Al2Cu phases are further reﬁned, as shown in Fig. 6(c). All the bar phases are reﬁned to dot phases and the Al2Cu phases are uniformly distributed in the composites. However, increasing SiCnp to more than 1.5 wt% has no obvious effect on the reﬁnement of Al2Cu phases. The further increment in SiCnp content even leads to an increase in the size of Al2Cu phases. This increase might be attributed to SiCnp agglomeration existing in the composites at higher SiCnp content (beyond 1.5 wt%), as shown in Fig. 6(d). For the reﬁnement of Al2Cu phases, there are three main reasons: (i) The reﬁnement of a-Al phases leads to the thinning of Al2Cu phases in grain boundary. (ii) With 851 the application of UT, there are effects of the cavitation and the acoustic streaming left in the composites, leading to the reﬁnement of Al2Cu phases. (iii) The existence of SiCnp can hinder the growth of Al2Cu phases because Al2Cu phases precipitate at grain boundaries at the last stage. Fig. 7 shows the high-magniﬁcation SEM images of each composite to further conﬁrm the distribution of SiCnp. The EDS results shown in Fig. 7(f) indicate that the bright particles are SiCnp. Obviously, SiCnp show a tendency of distributing along grain boundaries in all SiCnp/Al-5Cu composites, as shown in Fig. 7(aed). Since the lattice misﬁt between SiCnp and a-Al is over 5%, SiCnp cannot be captured by a-Al during solidiﬁcation and are subsequently pushed to the solid/liquid interface or grain boundaries [16,17]. The SiCnp distributed at the grain boundaries are mixed with Al2Cu phases since intermetallic Al2Cu phases is formed at the last stage of solidiﬁcation . Meanwhile, a few SiCnp can be observed inside the a-Al, which may be attributed to the semi coherent relationship between them [19,20]. Fig. 8 shows the TEM images of 1.5 wt% SiCnp/Al-5Cu composites. Very clean interface between SiCnp and matrix can be observed and the interface has no any reaction products, which indicates that these dispersed SiCnp can act as suitable reinforcements for Orowan strengthening [10,15]. In addition, UT has been applied in this study to homogenize the distribution of SiCnp. Fig. 9 shows the illustration of how UT to affect the distribution of SiCnp in the composites. The application of ultrasonic vibration can greatly affect the melt by cavitation and acoustic streaming. A lot of cavitation bubbles, which are generated in the melt under cyclic high-intensity ultrasonic waves, will undergo the process of formation, growth and collapse, repeatedly. The collapse of bubbles forms transient high pressure of about 1000 atm and micro-ﬂows or injection of liquid (faster than 100 m/ Fig. 6. Low-times SEM image of the SiCnp/Al-5Cu composites with different SiCnp content: (a) 0.5 wt%, (b) 1.0 wt%, (c) 1.5 wt%, (d) 2 wt%. 852 J. Li et al. / Journal of Alloys and Compounds 769 (2018) 848e857 Fig. 7. High-times SEM image of the SiCnp/Al-5Cu composites with different SiCnp content: (a) 0.5 wt%, (b) 1.0 wt%, (c) 1.5 wt%, (d) 2 wt%, (e) Spectrum of the gray Al2Cu phases, (f) Spectrum of the bright SiCp phases. Fig. 8. TEM images of 1.5 wt% SiCnp/Al-5Cu composites: (a) a higher magniﬁcation of SiCnp, (b) a HRFEM image of the interface between SiCnp and a-Al matrix, with the corresponding SAED pattern obtained from black particle in (a). J. Li et al. / Journal of Alloys and Compounds 769 (2018) 848e857 853 Fig. 9. Sketch of effects of ultrasonic cavitation and acoustic streaming on particles distribution. s) [21e23]. Meanwhile, the turbulent whirlpools are generated in the melt with the higher speed than that of heat convection, which can strongly stir the melt to accelerate the transfer of solute and SiCnp . Therefore, the cavitation and acoustic streaming can cooperatively promote the uniform distribution of SiCnp in the composites. Additionally, the reﬁnement of a-Al grains is also enhanced in the composites with UT owing to the effects of the cavitation and acoustic streaming [24,25]. 3.3. Density analysis of SiCnp/Al-5Cu composites Table 1 shows the theoretical and measured density of the matrix alloy and xSiCnp/Al-5Cu (x ¼ 0.5, 1, 1.5, 2 wt%) composites. The density of specimens was measured by Archimedes' method . As shown in Table 1, when SiCnp content increases to 2.0 wt%, the relative density of composites decreases due to the SiCnp agglomeration existing in the composites at higher SiCnp content. The entrapped air inside the SiCnp agglomeration can prevent metal ﬂowing into them and contribute to the reduction of measured density . However, the relative densities of both matrix and SiCnp/Al-5Cu composites with UT are above 99% due to the ultrasonic degassing effect . Therefore, it indicates that an ideal degassing of composites can be obtained by the processing of UT combined with Ar-gas bubbling. 3.4. Mechanical properties of SiCnp/Al-5Cu composites Fig. 10 shows the variation of mechanical properties of SiCnp/Al5Cu composites with SiCnp content. For comparison, the mechanical properties of unreinforced matrix alloy are also included in Fig. 10. It is clearly seen that the added SiCnp have a great inﬂuence on the deformation behavior of the composites. Compared with Al-5Cu matrix alloy, the SiCnp/Al-5Cu composites exhibit certain improvement in UTS, YS and elongation, respectively. Among these composites, the 1.5 wt% SiCnp/Al-5Cu composites show the highest UTS, YS and elongation, which are increased by 18.7%, 11.3% and 25.3%, respectively, compared with Al-5Cu matrix alloy. However, the further increase of SiCnp content leads to the degradation of mechanical properties due to the SiCnp agglomeration. Interestingly, with more SiCnp in the present composites, the strength increases while elongation increases as well, until to reach optimal at Table 1 Theoretical and measured density of the matrix alloy and xSiCnp/Al-5Cu (x ¼ 0.5, 1, 1.5, 2 wt%) composites. Specimens Theory density Measured density Relative density Al-5Cu matrix alloy 0.5SiCnp/Al-5Cu 1.0SiCnp/Al-5Cu 1.5SiCnp/Al-5Cu 2.0SiCnp/Al-5Cu 2.812 g 2.814 g 2.815 g 2.817 g 2.819 g 2.798 g 2.804 g 2.808 g 2.812 g 2.801 g 99.50% 99.64% 99.75% 99.82% 99.37% 1.5 wt%. The increased elongation of SiCnp/Al-5Cu composites is mainly attributed to ɑ-Al grains reﬁnement, uniform distribution of SiCnp and Al2Cu reﬁnement [28e36]: (i) In referring to Fig. 4, ɑ-Al grains reﬁnement of composites increases the number of crystal grains obviously. Thus the stress scatters in more ﬁne grains during tensile deformation, which leads to the less stress concentration and more uniform plastic deformation [28e30]. Besides, the more grain boundaries associated with ﬁner a-Al grains are beneﬁcial to hinder crack propagation and result in better ductility , which can be conﬁrmed by the increased dimples in the fracture surface, as shown in Fig. 12. (ii) As shown in Fig. 8, a clean interface between the SiCnp and matrix can enhance the bonding strength of interface and interfacial load transfer between SiCnp and the matrix, which are helpful to prolong the shear slide deformation to improve the fracture elongation of the composites [32,33]. Additionally, uniform distribution of SiCnp is also beneﬁcial to the improvement of elongation [34,35]. The SiCnp in composites can change the direction of crack growth, leading to the formation of crack bridging, branching and deﬂection in composites. During this process, a lot of energy is absorbed and the resistance of crack propagation increases, which will improve the elongation to fracture of composites . (iii) The Al2Cu reﬁnement is also beneﬁcial to the increase of elongation. Usually, the coarse phases can act as stress concentrators, which will provide sites for micro-cracking and reduce the ductility . In combination with Fig. 6, the ﬁner Al2Cu phases precipitate uniformly in composites, which can even act as reinforcement phases and hinder the dislocations motions during deformation. In addition, the high relative densities may also have contributed to the increase of elongation, as shown in Table 1. For the variation of UTS and YS, it is reported that there are four main reasons, including load-bearing strengthening (DsLoad), CTE (Coefﬁcient of Thermal Expansion) mismatch strengthening (DsCTE), Orowan strengthening (DsOrowan) and gain reﬁnement strengthening [37e43]. Firstly, according to load-bearing strengthening mechanism, the dispersed and well-bonded SiCnp can directly shear the load to strengthen the composites [37,38]. The DsLoad can be described as the following equation: 1 2 DsLoad ¼ Vp sm (1) where sm is the YS of Al-5Cu matrix which is 160 MPa; Vp is the volume fraction of reinforcement, and the volume fraction of x wt% SiCnp/Al-5Cu (x ¼ 0.5, 1.0, 1.5, 2.0) composites is 0.44%, 0.88%, 1.32% and 1.77%, respectively. Secondly, the difference of CTE between SiCnp and the matrix leads to the generation of geometrically necessary dislocations during solidiﬁcation, which can also strengthen the composites. This strengthening mechanism can be described as the following equation [13,39]. 854 J. Li et al. / Journal of Alloys and Compounds 769 (2018) 848e857 Fig. 10. The variation of mechanical properties of the SiCnp/Al-5Cu composites with different SiCnp content: (a) UTS and YS strength, (b) elongation. DsCTE sﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃ 12DaDTVp ¼ bGm b bdp 1 Vp (2) Vp* ¼ εVp where b is strengthening coefﬁcient equal to 1.25; Gm is shear modulus of Al matrix equal to 25.8 GPa; b is Burgers vector of Al matrix which is 0.286 nm; Da is the CTE difference between SiCnp and matrix which is 19.9 106 K1; DT is the difference between the pouring temperature and test temperature which is 690 K; dp and Vp are the mean size and volume fraction of SiCnp, respectively. Thirdly, the addition of SiCnp reﬁnes ɑ-Al grains and generates more grain boundaries, which can effectively hinder the dislocation movement during deformation. Moreover, the dispersed SiCnp themselves are beneﬁcial to improve the UTS and YS of the composites, by pinning the dislocations initiated from the matrix and forming a large number of dislocation loops around SiCnp. Thus SiCnp can serve as strong obstacles to the dislocation movement, leading to strength enhancement . Overall, the strength enhancement resulting from gain reﬁnement and dispersed SiCnp can be explained by the Hall-Petch relationship (Dshp ) and Orowan strengthening (DsOrowan ), respectively. These two main strengthening mechanisms can be calculated by using the following equations, respectively [41e43]. DsOrowan ¼ dp 0:13Gm b 1 ln 2b 1 3 1 dp 2V p Dshp ¼ khp .pﬃﬃﬃﬃﬃﬃ dm strengthening mechanisms of composites, namely the modiﬁed method . (6) where Vp* is the effective SiCnp content and ε is a coefﬁcient related to microstructural features of the composites. In order to obtain ε value, the load-bearing strengthening, CTE strengthening and Orowan strengthening are taken into consideration because they are related with volume fraction of SiCnp. Lloyd et al. reported that only the grain sizes smaller than 10 mm would obviously inﬂuence the yield strength in AMCs due to a low kh-p value of Al alloys . However, the average grain size of the matrix and composites is from ~60 mm to ~30 mm in this work. Therefore, the effect of gain reﬁnement strengthening can be ignored in the process of obtaining the ε value . Fig. 11 shows the comparison of calculated YS using the arithmetic summation method and measured YS, in which the calculated values deviate far from the experimental ones. It is because that the calculated YS by Eq. (5) is based on the simplifying and ideal hypothesis. For example, the SiCnp are assumed to be uniformly distributed in the whole matrix and the shape of them is assumed to be a perfect sphere. However, the SiCnp are not uniformly distributed in the matrix and most of SiCnp are distributed at (3) (4) where khp is the Hall-Petch constant and dm is the average grain diameter, respectively. The yield strength of the composites is estimated by considering multiple strengthening mechanisms, namely the arithmetic summation method. Thus, the predicted YS (sc ) of the composites can be calculated by the following equation according to the arithmetic summation method [42,44]: sc ¼ sm þ DsLoad þ DsCTE þ DsOrowan (5) However, recent studies have reported that the predicted values are much higher than the experimental values by using this method. Since the strengthening effects are related to the nominal SiCnp content, Wang et al. suggested that a coefﬁcient ε should be introduced to account for the effective SiCnp content (Vp*) in the Fig. 11. Calculated YS using the arithmetic summation method and the modiﬁed method (ε ¼ 0.01) without regard for grain reﬁnement strengthening. J. Li et al. / Journal of Alloys and Compounds 769 (2018) 848e857 855 Fig. 12. SEM images of the tensile fracture surface of SiCnp/Al-5Cu composites with different SiCnp content: (a) 0.5 wt%, (b) 1.0 wt%, (c) 1.5 wt%, (d) 2 wt%. the grain boundaries, which can signiﬁcantly reduce the strengthening effects. In order to calculate YS more accurately, we replace the Vp with the Vp* in Eqs. (1)e(3). The results shown in Fig. 11 suggest that the calculated yield strength using the modiﬁed method can ﬁt well with the measured data when choosing 0.01 as the value of ε, which implies the rationality of this modiﬁed method. However, when the content is beyond 1.5 wt%, the measured YS again deviates from the calculated value, which may result from the SiCnp agglomeration. Therefore, the theoretical contributions of the four strengthening mechanisms to the YS of the composites can be accurately calculated according to the modiﬁed method. The theoretical contribution values of four strengthening mechanisms to YS of the composites are shown in Table 2. Obviously, increasing SiCnp leads to higher theoretical contributions to YS of the composites, which is consistent with variation of the measured YS, and both of them increase until to reach optimal at 1.5 wt%. It can be also seen that the effect of load-bearing strengthening on the YS of composites is very limited due to the low SiCnp content (2 wt% for maximum). Among the four strengthening mechanisms, the DsCTE is the most important contributor due to the large CTE difference between the SiCnp and matrix. The contribution of DsOrowan is also important. Fig. 12 shows the SEM images of the tensile fracture surface of the SiCnp/Al-5Cu composites with different SiCnp content. Obviously, the SEM analysis shows that the composites exhibit ductile fracture with dimples. As shown in Fig. 12(a), there are a few dimples on the fracture surface. With the increase of SiCnp content, the number of dimples increases and the dimples are uniformly distributed on the fracture surface. However, the number of dimples decreases when the SiCnp content increases to 2.0 wt%. That may be attributed to the SiCnp agglomeration existing in the composites at higher SiCnp content. As shown in Fig. 12(d), the hole and SiCnp agglomeration marked with red circles can be observed on the fracture surface. Commonly the cracks are easier to form in the pore and agglomeration regions, leading to the fracture failure of composites. Thus the 2.0 wt% SiCnp/Al-5Cu composites have low strength and poor elongation, as shown in Fig. 10. 4. Conclusions (1) The SiCnp are uniformly distributed in the SiCnp/Al-5Cu composites by the processes of UT combined with mechanical stirring of the melt. (2) The primary a-Al and Al2Cu phases are reﬁned in the composites due to the addition of SiCnp, compared with Al-5Cu matrix alloy. The sizes of primary a-Al and Al2Cu phases Table 2 Theoretical contributions of four strengthening mechanisms for YS of SiCnp/Al-5Cu composites. Samples DsLoad (MPa) DsCTE (MPa) DsOrowan (MPa) Dshp (MPa) sm (MPa) 0.5SiCnp/Al-5Cu 1.0SiCnp/Al-5Cu 1.5SiCnp/Al-5Cu 2.0SiCnp/Al-5Cu 0.004 0.007 0.011 0.014 6.5 9.3 11.4 13.2 5.1 6.3 7.3 8.0 0.4 2.3 2.6 1.5 160 160 160 160 856 J. Li et al. / Journal of Alloys and Compounds 769 (2018) 848e857 decrease, when the SiCnp content increases from 0 wt% to 1.5 wt%. (3) The composites with SiCnp content from 0 to 1.5 wt% possess high relative densities, but the further increment in SiCnp content leads to a decline of density due to the SiCnp agglomeration. (4) Compared to Al-5Cu matrix alloy, the SiCnp/Al-5Cu composites exhibit signiﬁcant enhancement in UTS, YS and elongation when the SiCnp content increases from 0 wt% to 2.0 wt%. Among them, the SiCnp/Al-5Cu composites having 1.5 wt% SiCnp exhibit the best UTS, YS and elongation, which are increased by 18.7%, 11.3% and 25.3%, respectively, compared with Al-5Cu matrix alloy. (5) A modiﬁed method is applied to predict the theoretical YS, which is in good agreement with experimental value. The enhancement of strength is attributed to four strengthening mechanisms, among which the DsCTE and DsOrowan are the most important contributors. The increase in elongation is mainly attributed to a-Al grains reﬁnement, Al2Cu reﬁnement and uniform distribution of SiCnp. Acknowledgments This work was funded by the Project 51574129 supported by National Natural Science Foundation of China, and by the project JCKY 2016209A001. The authors would also express their appreciation to the Analytical and Testing Centre, HUST. References  T.B. He, H.Q. Li, P.J. Tang, X.L. He, P.Y. Li, Inﬂuence of quenching agent on microstructure, properties and thermal stress of SiCp/2009 composites, Mater. Char. 118 (2016) 547e552.  S.A. Sajjadi, H.R. Ezatpour, M.T. Parizi, Comparison of microstructure and mechanical properties of A356 aluminum alloy/Al2O3 composites fabricated by stir and compo-casting processes, Mater. Des. 34 (2012) 106e111.  D. Mandal, S. Viswanathan, Effect of heat treatment on microstructure and interface of SiC particle reinforced 2124 Al matrix composite, Mater. Char. 85 (2013) 73e81.  J.M. Molina, R. Prieto, J. Narciso, E. Louis, The effect of porosity on the thermal conductivity of Al-12 wt.% Si/SiC composites, Scripta Mater. 60 (2009) 582e585.  N. Nemati, R. Khosroshahi, M. Emamy, A. Zolriasatein, Investigation of microstructure, hardness and wear properties of Al-4.5 wt.% Cu-TiC composites produced by mechanical milling, Mater. Des. 32 (2011) 3718e3729.  Y.C. Kang, S.L.I. Chan, Tensile properties of nanometric Al2O3 particulate reinforced aluminum matrix composites, Mater. Chem. Phys. 85 (2004) 438e443.  B. Xiong, Z.F. Xu, Q.S. Yan, C.C. Cai, Y.Z. Zheng, B.P. Lu, Fabrication of SiC nanoparticulates reinforced Al matrix composites by combining pressureless inﬁltration with ball-milling and cold-pressing technology, J. Alloys Compd. 497 (2010) L1eL4.  L. Wang, F. Qiu, Q. Zou, D.L. Yang, J. tang, Y.Y. Gao, Q. Li, X. Han, S.L. Shu, F. Chang, Q.C. Jiang, Microstructures and tensile properties of nano-sized SiCp/ Al-Cu composites fabricated by semisolid stirring assisted with hot extrusion, Mater. Char. 131 (2017) 195e200.  S. Tahamtan, A. Halvaee, M. Emamy, M.S. Zabihi, Fabrication of Al/A206eAl2O3 nano/micro composite by combining ball milling and stir casting technology, Mater. Des. 49 (2013) 347e359.  J.Y. Li, S.L. Lü, S.S. Wu, Q. Gao, Effects of ultrasonic vibration on microstructure and mechanical properties of nano-sized SiC particles reinforced Al-5Cu composites, Ultrason. Sonochem. 42 (2018) 814e822.  Z.W. Xu, L. Ma, J.C. Yan, S.Q. Yang, S.Y. Du, Wetting and oxidation during ultrasonic soldering of an alumina reinforced aluminum-copper-magnesium (2024 Al) matrix composite, Composites Part A 43 (2012) 407e414.  Q. Gao, S.S. Wu, S.L. Lü, X.C. Xiong, Effects of ultrasonic vibration treatment on particles distribution of TiB2 particles reinforced aluminum composites, Mater. Sci. Eng. A 680 (2017) 437e443.  X. Yao, Z. Zhang, Y.F. Zheng, C. Kong, M.Z. Quadir, J.M. Liang, Y.H. Chen, P. Munroe, D.L. Zhang, Effects of SiC nanoparticle content on the microstructure and tensile mechanical properties of ultraﬁne grained AA6063-SiCnp composites fabricated by powder metallurgy, J. Mater. Process. Technol. 33 (2017) 1023e1030.  D. Yuan, K. Hu, S.L. Lü, S.S. Wu, Q. Gao, Preparation and properties of nanoSiCp/A356 composites synthesised with a new process, Mater. Sci. Technol. https://doi.org/10.1080/02670836.2018.1458479.  H. Zhang, Y.C. Zhao, Y. Yan, J.F. Fan, L.F. Wang, H.B. Dong, B.S. Xu, Microstructure evolution and mechanical properties of Mg matrix composites reinforced with Al and nano SiC particles using spark plasma sintering followed by hot extrusion, J. Alloys Compd. 725 (2017) 652e664.  S.S. Wu, H. Nakae, Nucleation effect of alumina in Al-Si/Al2O3 composites, J. Mater. Sci. Lett. 18 (1999) 321e323.  S.S. Wu, H. Nakae, T. Kanno, Y. You, Zone-melted unidirectional solidiﬁcation of particulate dispersed composites, J. Mater. Sci. 36 (2001) 225e229.  Q. Gao, S.S. Wu, S.L. Lü, X.C. Xiong, R. Du, P. An, Improvement of particles distribution of in-situ 5 vol% TiB2 particulates reinforced Al-4.5Cu alloy matrix composites with ultrasonic vibration treatment, J. Alloys Compd. 692 (2017) 1e9.  Z.P. Luo, Crystallography of SiC/MgAl2O4/Al interfaces in a pre-oxidized SiC reinforced SiC/Al composite, Acta Mater. 54 (2006) 47e58.  B.L. Bramﬁtt, The effect of carbide and nitride additions on the heterogeneous nucleation behavior of liquid iron, Metal Trans. 1 (1970) 1987e1991.  C. Lin, S.S. Wu, S.L. LÜ, P. An, Effects of ultrasonic vibration and manganese on microstructure and mechanical properties of hypereutectic Al-Si alloys with 2% Fe, Intermetallics 32 (2013) 176e183.  J.C. Yan, Z.W. Xu, L. Shi, X. Ma, S.Q. Yang, Ultrasonic assisted fabrication of particle reinforced bonds joining aluminum metal matrix composites, Mater. Des. 32 (2011) 343e347.  S.L. Lü, P. Xiao, S.S. Wu, X.G. Fang, Preparation and squeeze casting of nanoSiCP/A356 composite assisted with ultrasonic vibration process, Mater. Sci. Forum 879 (2017) 1188e1193.  R.R. Chen, D.S. Zheng, T.F. Ma, H.S. Ding, Y.Q. Su, J.J. Guo, H.Z. Fu, Effects and mechanism of ultrasonic irradiation on solidiﬁcation microstructure and mechanical properties of binary TiAl alloys, Ultrason. Sonochem. 38 (2017) 120e133.  S. Merouani, O. Hamdaoui, Y. Rezgui, M. Guemini, Theoretical estimation of the temperature and pressure within collapsing acoustical bubbles, Ultrason. Sonochem. 21 (2014) 53e59.  S. Amirkhanlou, B. Niroumand, Development of Al356/SiCp cast composites by injection of SiCp containing composite powders, Mater. Des. 32 (2011) 1895e1902.  H.B. Xu, Q.Y. Han, Effects of ultrasonic vibration on degassing of aluminum alloys, Mater. Sci. Eng. A 473 (2008) 96e104.  B.Y. Jiang, S.R. Zhang, The effects of strain rate and grain size on nanocrystalline materials: a theoretical prediction, Mater. Des. 87 (2015) 49e52.  Q. Zhang, B. Liu, Z. Niu, Z. Zhang, Z. Leng, Grain reﬁnement and mechanical properties of Mg-5Li-3Al alloy inoculated by Al-5Ti-1B master alloy, Mater. Sci. Eng. A 619 (2014) 152e157.  K. Tian, Y. Zhao, L. Jiao, S. Zhang, Z. Zhang, X. Wu, Effects of in situ generated ZrB2 nano-particles on microstructure and tensile properties of 2024Al matrix composites, J. Alloys Compd. 594 (2014) 1e6.  X.G. Fang, S.S. Wu, S.L. Lü, J. Wang, X. Yang, Microstructure evolution and mechanical properties of quasicrystal-reinforced Mg-Zn-Y alloy subjected to ultrasonic vibration, Mater. Sci. Eng. A 679 (2017) 372e378.  W. Zhou, S. Bang, H. Kurita, T. Miyazaki, Y. Fan, A. Kawasaki, Interface and interfacial reactions in multi-walled carbon nanotube-reinforced aluminum matrix composites, Carbon 96 (2016) 919e928.  X.H. Liu, J.J. Li, E.Z. Liu, Q.Y. Li, C.N. He, C.S. Shi, N.Q. Zhao, Effectively reinforced load transfer and fracture elongation by forming Al4C3 for in-situ synthesizing carbon nanotube reinforced Al matrix composites, Mater. Sci. Eng. A 718 (2018) 182e189.  Z. Zhang, T. Topping, Y. Li, R. Vogt, Y. Zhou, C. Haines, J. Paras, D. Kapoor, J.M. Schoenung, E.J. Lavernia, Mechanical behavior of ultraﬁne-grained Al composites reinforced with B4C nanoparticles, Scripta Mater. 65 (2011) 652e655.  C. Fang, G. Liu, H. Hao, X. Zhang, Effects of particle distribution on microstructural evolution and mechanical properties of TiB2/AZ31 composite sheets, Mater. Sci. Eng. A 684 (2017) 592e597.  W. Dai, S.S. Wu, S.L. Lü, C. Lin, Effects of rheo-squeeze casting parameters on microstructure and mechanical properties of AlCuMnTi alloy, Mater. Sci. Eng. A 538 (2012) 320e326.  J. Nampoothiri, R.S. Harini, S.K. Nayak, B. Raj, K.R. Ravi, Post in-situ ultrasonic treatment for generation of Al-4.4Cu/TiB2 nanocomposite: a route to enhance the strength of metal matrix composites, J. Alloys Compd. 683 (2016) 370e378.  C.S. Kim, I. Sohn, M. Nezafati, J.B. Ferguson, B.F. Schultz, Z. Bajestani-Gohari, P.K. Rohatgi, K. Cho, Prediction models for the yield strength of particlereinforced unimodal pure magnesium (Mg) metal matrix composites (MMNCs), J. Mater. Sci. 48 (2013) 4191e4204.  S.J. Hong, H.M. Kim, D. Huh, C. Suryanarayana, B.S. Chun, Effect of clustering on the mechanical properties of SiC particulate-reinforced aluminum alloy 2024 metal matrix composites, Mater. Sci. Eng. A 347 (2003) 198e204.  Z. Zhang, D.L. Chen, Consideration of Orowan strengthening effect in particulate-reinforced metal matrix composites: a model for predicting their yield strength, Scripta Mater. 54 (2006) 1321e1326.  J.B. Correia, H.A. Davies, C.M. Sellars, Strengthening in rapidly solidiﬁed age hardened Cu-Cr and Cu-Cr-Zr alloys, Acta Mater. 45 (1997) 177e190.  A. Matin, F.F. Saniee, H.R. Abedi, Microstructure and mechanical properties of Mg/SiC and AZ80/SiC nano-composites fabricated through stir casting method, Mater. Sci. Eng., A 625 (2015) 81e88.  N.J. Petch, The cleavage strength of polycrystals, J. Iron Steel Inst. 174 (1953) J. Li et al. / Journal of Alloys and Compounds 769 (2018) 848e857 25e28.  A. Sanaty-Zadeh, Comparison between current models for the strength of particulate-reinforced metal matrix composites with emphasis on consideration of Hall-Petch effect, Mater. Sci. Eng. A 531 (2012) 112e118.  M.L. Wang, D. Chen, Z. Chen, Y. Wu, F.F. Wang, N.H. Ma, H.W. Wang, 857 Mechanical properties of in-situ TiB2/A356 composites, Mater. Sci. Eng. A 590 (2014) 246e254.  D.J. Lloyd, Particle reinforced aluminium and magnesium matrix composites, Int. Mater. Rev. 39 (1994) 1e23.