Materials Characterization Volume 152, June 2019, Pages 169-179

Effect of SiC nanoparticles on the microstructure and texture of friction stir welded AA2024/AA6061

Mohammad Mahdi Moradia, Hamed Jamshidi Avala, Roohollah Jamaatia, Sajjad Amirkhanloub, Shouxun Jic

a-Department of Materials Engineering, Babol Noshirvani University of Technology, Shariati Ave., Babol 47148–71167, Iran

b-Department of Materials, University of Oxford, Parks Rd, Oxford Ox1 3PH, United Kingdom

c-Institute of Materials and Manufacturing, Brunel University London, Uxbridge UB8 3PH, United Kingdom

https://doi.org/10.1016/j.matchar.2019.04.020

Highlights

•The crystallographic relationship between aluminum and silicon carbide was .

•The addition of SiC nanoparticles has reduced the grain size of the SZ to 7.4 μm.

•The average size of particle in the stirred zone was much smaller than TMAZ.

•The nanoparticles prevented the rotation of the {001}〈100〉 oriented new grains.

Abstract

In the present study, the effect of SiC nanoparticles on the microstructure and texture evolution of friction stir welded AA2024/AA6061 dissimilar joint investigated. The results showed that the SiC nanoparticles homogeneously dispersed in the aluminum matrix. The Al/SiC interfaces had a good quality of bonding between particles and matrix. It was found that the new recrystallized aluminum grains tend to nucleate and grow on the crystal plane of SiC nanoparticle ( facet on surface). The average grain size in the stirred zone (SZ) decreased to 7.4 μm owing to the grain refinement through the occurrence of both dynamic and static recrystallization. Most grains in the center of the stirred zone grown due to the occurrence of continuous grain growth (CGG) mechanism after the stirring. Moreover, the incorporation of SiC nanoparticles led to a reduction of 26% in the mean grain size of the SZ. The average size of the precipitate particles in the stirred zone was much smaller than the thermo-mechanically affected zone (TMAZ) owing to the dissolution of precipitates during stirring and reprecipitation during cooling. In addition, the grain size of the stirred zone on the advancing side was lower than the retreating side (2.6 μm vs. 9.7 μm). The results revealed a weaker texture after FSW compared to the initial samples. The SiC nanoparticles largely suppressed the rotation of the {001}〈100〉 oriented new grains through particle pinning effect. Unlike the advancing side, there was no recrystallization texture component on the retreating side.

Keywords:

Aluminum alloysSiC nanoparticlesFriction stir welding (FSW)Transmission electron microscopy (TEM)Electron backscattered diffraction (EBSD)Crystallographic texture

1. Introduction

Aluminum alloys are one of the most important lightweight materials. They are widely used in structural, automotive, and aerospace industries due to their high strength-to-weight, good plasticity, high fracture toughness, good wear resistance, and excellent corrosion resistance [[1], [2], [3]]. However, the weldability of these alloys, especially the heat-treatable series, is relatively poor [[4], [5], [6]]. Hot tearing, oxidation, dendritic and eutectic structures, and development of hydrogen gas porosity during solidification are severe metallurgical problems associated with fusion welding techniques of aluminum alloys [[6], [7], [8]]. Thus, a solid-state method is appropriate for welding Al alloys.

Friction stir welding (FSW) is a desirable welding technique for aluminum alloys [9,10]. This process performed at a temperature below the melting temperature of materials [11,12]. Compared with the fusion techniques, there is no melt during friction stir welding. Consequently, many of the disadvantages associated with fusion welds are reduced or eliminated [7,10,11]. This technique has been successfully used for joining similar and dissimilar Al alloys [13,14].

FSW joint is divided into four regions based on the microstructural characterization that exhibit different mechanical properties. These regions are the stirred zone (SZ), thermo-mechanically affected zone (TMAZ), heat affected zone (HAZ) and base metal (BM) [6,7,10,11]. Frictional heat and severe plastic deformation results in a fine dynamically recrystallized microstructure in the SZ [10,13,14]. The TMAZ typically shows the elongated grains while the HAZ shows over-aging (in the heat-treatable alloys) and grain growth [10,[13], [14], [15]]. It is essential to investigate the microstructures of the regions and the interfaces between them to determine mechanical properties.

Many efforts have been performed to document the influence of FSW on the microstructural and mechanical properties of the similar and dissimilar joints [[4], [5], [6],8,10,[12], [13], [14], [15], [16]]. Recently, several studies conducted to produce the metal matrix composites via FSW. Owing to different thermo-mechanical response, friction stir welding in metal matrix composites can create complex structures. Karthikeyan and Mahdevan [17] investigated the influence of silicon carbide particle incorporation in the AA6351 alloy. They found that the incorporation of silicon carbide hinders the boundary growth by pinning and led to an improvement in the tensile properties. Abbasi et al. [18] used silicon carbide nanoparticles during the friction stir welding of Mg-Al-Zn alloy and studied the influence of nanoparticles on the microstructural and mechanical properties. Their results showed that the incorporation of silicon carbide nanoparticles improved the tensile strength and formability. Tabasi et al. [19] studied the influence of SiC nanoparticles during friction stir welding on the microstructure and mechanical response of dissimilar AA7075–AZ31 joint. They found that the creation of brittle intermetallic compounds decreased the tensile strength and elongation.

They observed that there was no direct relation between the grain size of the stirred zone grain size and its tensile strength due to the destructive effects of the intermetallic particles. Finally, Bodaghi et al. [20] used silicon carbide nanoparticles during the friction stir welding of AA5052 alloy. Their results indicated that the incorporation of SiC led to the remarkable effect on the grain refinement of the stirred zone; because the reinforcements are preferred nucleation sites and pin the boundaries so that the average grain size decreased from 36 μm in the BM to 5 μm in the stirred zone.

Texture evolution gives important information about recrystallization mechanisms and plastic deformation. Several studies were found in the literature describing the development of crystallographic texture through the application of FSW on the aluminum alloys. Fonda et al. [21,22] observed the shear texture orientations created in FSW and reported their evolution from simple shear components to complicated 〈111〉 //ND orientations. Wang et al. [23] investigated the texture evolutions during the FSW of AA5052-AA6061. They demonstrated that the {001}〈100〉 Cube and {123}〈634〉 S orientations in the BM transformed into shear component. Suhuddin et al. [24] studied texture evolution during FSW of AA6016 sheets. They reported that the {112}〈110〉 component as a shear texture produced via the

rotation of a recrystallized grain with {001}〈100〉 Cube component. However, very limited works have been done on the texture development of the friction stir welded aluminum-based composites. Feng et al. [25] investigated the misorientation of grain boundaries and texture evolutions in FSWed Al-Cu-Mg/silicon carbide composite. They found that the texture in the stirred zone remarkably weakened compared to that in the initial sample.

To the best of our knowledge, no information is available on the microstructure evolution and texture development in hybrid aluminum (bi-metallic) matrix nanocomposites prepared by FSW. The main objective of this study is to fabricate the AA2024/SiC/AA6061 nanocomposite via FSW. The effect of silicon carbide nanoparticles on the microstructural and textural evolutions of joints investigated.

2. Experimental procedure

The materials used in this work were AA2024 and AA6061 with a dimension of 100 mm × 50 mm × 6 mm. The chemical composition of the alloys is reported in our previous work [7]. The SiC nanoparticles with the average particle size of 42 ± 8.4 nm used as reinforcement materials. Fig. 1 shows the SiC nanoparticles obtained with field emission scanning electron microscopy (FESEM). In order to add the SiC nanoparticles, a groove of 0.3 mm in width and 5 mm in depth machined using a milling machine. The advancing and retreating sides were AA2024 and AA6061, respectively, on the machine bed without root gap. Then, the SiC nanoparticles compacted into the groove. An AISI H13 hot work steel tool used, having a conical geometry with 18 mm shoulder diameter, a 2° conical cavity, a square frustum probe measuring 3.5–7 mm in diameter, and 5.8 mm in length. The tool tilt kept constant at 2° for all experiments. Using a trial and error method, proper welding conditions obtained. Based on the results obtained in our previous paper [7] a tool rotation speed of 800 rpm and a traverse speed of 31.5 mm/min used. The welding direction (WD) was parallel

to the rolling direction (RD) of both alloys. The transverse direction (TD) of the welded joint denoted as cross-welding direction (CWD).

Fig. 1. FESEM micrograph of silicon carbide nanoparticles.

In order to investigate the microstructures of samples, optical microscope (OM), scanning electron microscope (SEM), Electron backscattered diffraction (EBSD), and transmission electron microscope (TEM) used. EBSD analyzed regions of the dissimilar joint are shown in Fig. 2. EBSD images obtained using a FESEM operated at a voltage of 20 kV, the working distance of 12 mm, a tilt angle of 70° and a step size of 0.5 μm. Specimens for EBSD characterization mechanically ground and polished with standard metallurgical methods. The specimens subsequently electropolished with a solution of 30 pct. nitride acid and 70 pct. methanol at 12 V and −30 °C. Grain orientation map, grain boundary map, kernel average misorientation (KAM), pole figure (PF), and orientation distribution function (ODF) were obtained using EDAX's OIM™ software.

Fig. 2. EBSD analyzed regions of the dissimilar joint.

3. Results and discussion

Microstructures of as-received AA2024 and AA6061 alloys presented in Fig. 3. The aluminum matrix for both AA2024 and AA6061 sheets shows elongated grains due to the pre-rolling process with the mean grain size of 49.8 μm and 56.5 μm, respectively. The AA2024 and AA6061 alloys contain intermetallic compounds distributed randomly in the alpha aluminum. Besides iron-rich intermetallics and metastable elements of the precipitation sequence, the basic particles of the AA2024 are Al2Cu and Al2CuMg, and that for the AA6061 is Mg2Si. From Fig. 3, the average particle sizes of AA2024 and AA6061 are about 1.3 and 4.2 μm, respectively.

Fig. 3. Microstructure of as-received (a) AA2024 and (b) AA6061.

Fig. 4 shows FESEM micrographs from the center of the stirred zone. Fig. 4(a) indicates the SiC nanoparticles homogeneously dispersed in the aluminum matrix. There are no clusters or

agglomeration of nanoparticles. With higher magnification, the particle distribution can be seen more clearly as in Fig. 4(b). It should be noted that these nanometer-size particles are silicon carbide, iron-rich intermetallics, and surface aluminum oxide fractured during FSW. The distribution of particles is a function of process parameters, such as traverse and rotational speeds [[17], [18], [19], [20]]. The uniform dispersion in Fig. 4 confirms that the chosen set of FSW parameters (rotational speed of 800 rpm and traverse speed of 31.5 mm/min) is suitable to produce the desired distribution of particles in the aluminum matrix. It is well-accepted that the severe plastic deformation together with the rotating action of the tool during friction stirring induces a change in the size and morphology of particles [17,26,27]. In fact, the FSW results in a considerable break up of ceramic particles. However, fragmentation of ceramic particles depends on their initial size and shape. The coarse and non-spherical particles have the tendency to break off during intense plastic deformation [17,[26], [27], [28]]. In the present work, with regard to Fig. 1, the fine and spherical SiC particles are used. Consequently, the silicon carbide particles did not undergo fragmentation and the shape and size of SiC particles remained almost unchanged after FSW (see Fig. 4).

Fig. 4. FESEM micrographs of the stirred zone at two different magnifications.

Fig. 5 shows the TEM micrographs of the stirred zone. The interfaces between the aluminum and the SiC nanoparticles are clear. It can be observed that there are no defects (pores or reaction products) at the Al/SiC interfaces. The interfaces have a good quality of bonding between particles and matrix. The interface plays a vital role in transferring mechanical load effectively to the particles. Good interfacial adhesion between matrix and reinforcement enhances the properties of the composite. The processing temperature is an important variable that affects the interfacial strength. FSW at high temperature leads to undesired reactions and/or porosity generation in the interfaces between matrix and particles that weakens the interfacial strength. As seen in Fig. 5, there is no porosity close to the silicon carbide

nanoparticles owing to sufficient plasticization of the aluminum matrix under the chosen experimental condition. However, some researchers reported porosity in the area adjacent to the reinforcement particles in the aluminum matrix composite due to the lack of adequate material flow during friction stirring [[29], [30], [31]].

Fig. 5. TEM micrographs of the stirred zone at two different magnifications.

Fig. 6 presents the HRTEM images of the stirred zone. There is a lack of dislocation within the aluminum grain interior as seen in Fig. 6(a). This can be attributed to the occurrence of recrystallization and formation of dislocation-free α grains in the stirred zone during FSW. Fig. 6(b) shows the interface between aluminum and silicon carbide nanoparticle. It is apparent that there are two groups of diffraction points on the selected area diffraction (SAD) pattern. The bigger and brighter diffraction points are corresponding to that of aluminum. The smaller and darker ones are corresponding to that of SiC particle. According to the HRTEM micrograph and the SAD pattern of the interfacial area, the crystallographic relationship between aluminum and silicon carbide is also identified as . This relationship indicates that the new recrystallized aluminum grain tend to nucleate and grow on the crystal plane of SiC nanoparticle ( facet on surface). It should be noted that there is a very small angle between

aluminum and SiC particle. In fact, is not totally parallel to during the recrystallization. However, this angle is very small, which means that there is a minor lattice misfit of them. As also suggested by Fig. 6(b) the Al/SiC interface is smooth and straight, and there are no reaction layers on it.

Fig. 6. HRTEM of (a) aluminum and (b) Al/SiC interface.

Fig. 7 shows the EBSD maps (orientation, boundary, and kernel), SEM micrograph, grain size distribution, and grain boundary misorientation distribution from the center of the stirred zone. In the grain boundary map, blue lines denote the high angle grain boundaries (HAGBs), green lines denote the low angle grain boundaries (LAGBs). The center of the SZ exhibited fine and relatively equiaxed grains. It should be noted that some low angle grain boundaries can be detected. The grain size distribution is narrow as shown in Fig. 7(e). The mean grain size of the aluminum is 7.4 μm compared to the initial samples grain sizes of 49.8 μm and 56.5 μm. The formation of fine grain structure is mainly attributed to the occurrence of dynamic recrystallization (DRX). However, it is possible to take place static recrystallization (SRX) once the stirring tool moves away from the stirred zone.

Fig. 7. (a) Orientation, (b) boundary, (c) kernel maps, (d) SEM micrograph, (e) distribution of grain size, and (f) distribution of grain boundary misorientation from the center of the stirred zone.

The well-accepted DRX mechanisms of the metals and alloys during FSW are particle stimulated nucleation (PSN), geometric dynamic recrystallization (GDRX), continuous dynamic recrystallization (CDRX), and discontinuous dynamic recrystallization (DDRX) [10,13,14]. Aluminum and its alloys undergo restoration mainly by continuous dynamic recrystallization when subjected to severe plastic deformation at high temperature because they have high stacking fault energy (SFE). The CDRX occurs by the gradual accumulation of dislocations, the annihilation of some dislocations, rearrangement of other dislocations as

subgrain boundaries (i.e. LAGBs), increasing misorientation (by rotation of existing subgrains or by absorption of dislocations into LAGBs) and transforming subgrain boundaries into HAGBs [10,14,32]. Regarding Fig. 7(b), (e), and (f), the CDRX seemed to be operating as the main grain refinement mechanism owing to the presence of some LAGBs and a relatively uniform distribution of grain size. On the other hand, the easier cross-slip in the AA2024 and AA6061 prevent stored strain energy necessary for discontinuous dynamic recrystallization to occur [32]. This mechanism is more likely to occur in low SFE alloys like copper alloys. From Fig. 7(b), the geometric dynamic recrystallization happened in the AA2024/SiC/AA6061. This DRX mechanism results from the impingement of serrated boundaries in the highly elongated grains during FSW. It can be concluded that CDRX and GDRX are the plausible mechanisms for DRX in the stirred zone of AA2024/SiC/AA6061.

In the grain boundary map (Fig. 7(b)) some blue areas were found which cannot be indexed. To better track the blue areas, it is suggestible to consider corresponding SEM image (Fig. 7(c)). The poor-indexed regions can be attributed to the presence of pores as indicated by SEM micrograph. The presence of pores leads to shadowing and remarkably decreases the indexing. Consequently, the blue areas on the grain boundary map represent the places of the pre-existing particles. A similar observation was made in the study by Ma et al. [33] and Hoziefa et al. [34] in their EBSD investigation of aluminum matrix nanocomposite.

The KAM map of the stirred zone presented in Fig. 7(d). This map indicates the average misorientation that each pixel has with all its immediate neighbors. The average local misorientation within an area is an indication of stored energy. The red- and blue-colored regions represent the highest and lowest deformed area and the KAM value, respectively. Most of α-Al grains in the center of stirred zone show low KAM values (Fig. 7(d)) which is a pointer to recrystallized grains. However, when the 1–3° boundaries (i.e. green and yellow colors) are considered, they overlie on the area with moderate KAM value (includes grains of medium dislocation density), indicating the occurrence of dynamic recovery (DRV). Fig. 7(d) clearly shows that the KAM value in some regions is the highest (orange and red colors). This may be due to the presence of particles observed in Fig. 7(b) and (c).

One can also be observed in Fig. 7, most grains in the center of the stirred zone have grown. This would appear to be due to the occurrence of continuous grain growth (CGG) mechanism after the stirring. The grain growth occurs through two major mechanisms: (i) reduction in curvature of grain boundaries, and (ii) geometrical coalescence of grains [32,35]. As can be seen in Fig. 7(a), some regions have similarly oriented grains and the boundaries between them are LAGBs. The geometrical coalescence of grains can easily take place in such favored situations and two or more grains can coalesce together and form a single larger grain. The grain growth continues until the temperature significantly decreases or a stable crystallographic orientation develops or particles pin the grain boundaries.

It was reported in our previous paper [36] that the mean grain size of the stirred zone without particle addition was about 10.0 μm. The presence of silicon carbide nanoparticles decreased the grain size from 10.0 μm to 7.4 μm. Therefore, the addition of SiC nanoparticles results in

lower grain size in the stirred zone. The ceramic particles have a tendency to pin the movement of the grain boundaries and inhibition of CGG caused by DRX. Such observations were found in the SZ of FSW sample in Fig. 7(b). It should be noted that the finer grain size can only be attributed to the addition of the silicon carbide nanoparticles because the welding conditions remained the same. As a result of the present work, the addition of SiC nanoparticles led to a reduction of 26% in the grain size. This finding is similar to the results obtained by other investigators [3,37]. For example, Heidarzadeh et al. [37] studied the effect of Al2O3 nanoparticles on the pinning of copper grain boundaries using EBSD technique. The mean grain size of the stirred zone without Al2O3 nanoparticles was 2.8 μm, while, in the case of the sample with alumina, the average grain size was about 0.7 μm. The same results have been reported by Khodabakhshi et al. [3] examining the effect of TiO2 nanoparticles on the grain structure of Al-Mg alloys. The mean grain size without the addition of TiO2 nanoparticles was about 10 μm. The addition of titanium oxide nanoparticles enhanced the grain refinement of the stirred zone. The average grain size was 4.6 and 2 μm for the composites consist of 2.3% and 6.1% TiO2 nanoparticles, respectively.

As also suggested by Fig. 7, no PSN has occurred in the stirred zone of FSW sample. In general, the PSN phenomenon can take place during friction stirring of aluminum matrix composite [36,[38], [39], [40]]. However, the PSN will only occur if dislocations accumulate at the matrix/particle interface during deformation [32]. In other words, if there is not enough net stored energy in the vicinity of the particles, no PSN will occur. The dislocation accumulation only occurs around large (>1 μm) non-shearable particles at low processing temperature [32]. In the present investigation, there is no PSN in the SZ because the particle size was very small and the temperature of the stirred zone was quite high. It can be concluded that the formation of finer grains in the presence of particles is attributed to the Zener pinning effect by the silicon carbide nanoparticles, which hindered the grain growth of the aluminum matrix.

Fig. 8 shows EBSD maps, SEM image, distribution of grain size, and distribution of grain boundary misorientation of SZ/TMAZ interface on the advancing side. The interface of the SZ with the TMAZ is sharp. The average grain sizes of stirred and thermos-mechanically affected zones on the advancing side are about 2.6 and 50.5 μm, respectively. The size of grains achieved in the TMAZ ranged from 0.5 μm to 86.7 μm, and there are many coarse and elongated grains in the distribution. However, the grain size distribution of SZ on the advancing side is narrow, and there are fine equiaxed grains as indicated in Fig. 8(a) and (b). Similar to the center of SZ, the dominant restoration mechanism in the SZ on the advancing side of the weld is CDRX. At the TMAZ, most grains are highly elongated and the TMAZ exhibits a flow pattern in the stirred zone due to severe shear deformation induced by the rotating tool. On one hand, many subgrains surrounded by low angle boundaries formed in the elongated grain interiors indicating that the DRV widely occurred. On the other hand, between the coarse grains, some fine grains (as shown by purple arrows in Fig. 8(b)) appear indicating that the CDRX took place.

Fig. 8. (a) Orientation, (b) boundary, (c) kernel maps, (d) SEM micrograph, (e) distribution of grain size, and (f) distribution of grain boundary misorientation of the advancing side.

As can be seen in Fig. 8(b) and (c), the particle sizes of SZ and TMAZ are different. The mean particle size in the stirred zone is much lower than TMAZ due to the dissolution of precipitates during stirring and reprecipitation during cooling. It should be noted that the transformation of precipitate in the thermos-mechanically affected zone is similar to that in the SZ, but the level of particle dissolution changes owing to the temperature difference. The level of particle dissolution in the SZ is higher than the TMAZ because of the higher

temperature and strain. Mijajlovic et al. [41] and Gholamnejad et al. [42] expressed that the temperature in the SZ of AA2024 aluminum alloy during FSW reached between 410 °C and 470 °C. This temperature is high enough to force precipitate particles to go into solution. On the other hand, due to the fast diffusion of copper in the aluminum matrix, the cooling rate after stirring is not high enough to avoid reprecipitation. At the TMAZ, owing to its lower temperature, only the relatively small metastable precipitates dissolve and the large precipitates overage to equilibrium phases. For this reason, the particle coarsening occurred in the TMAZ.

Unlike the center of SZ, the PSN occurred in the thermo-mechanically affected zone on the advancing side because the particles are coarse enough (i.e. larger than 1 μm) and the temperature is not very elevated. Thus, similar to Refs. [1,[43], [44], [45]], particle stimulated nucleation could have happened as shown by red arrows in Fig. 8(b). The presence of large particles results in the formation of additional dislocations at the interfaces during stirring. In addition, the thermal expansion coefficients of Al2Cu and aluminum are 17.2 × 10−6 K−1 [46] and 23.1 × 10−6 K−1 [1], respectively. The mismatch in the coefficients of thermal expansion could introduce large thermal stress and correspondingly increase the dislocation density in the interfaces during FSW. Therefore, the regions near to large particles have higher dislocation density. As the dislocations rearrange by DRV and generate subgrain and even grain boundaries, a higher fraction of both low and high angle boundaries created in the regions adjacent to particles compared to other regions in the TMAZ as indicated in Fig. 8(d).

Fig. 9 demonstrates EBSD maps, distribution of grain size, and distribution of grain boundary misorientation of SZ/TMAZ interface on the retreating side of the joint. Unlike the advancing side, the SZ/TMAZ interface is relatively diffused on the retreating side. In other words, the microstructure of the SZ/TMAZ interface is homogeneous and the changes are gradual. The mean grain size of SZ/TMAZ interface on the retreating side is about 10 μm. The microstructure consists of fine equiaxed grains and the grain size distribution is narrow as indicated in Fig. 9. The dominant restoration mechanism in the SZ/TMAZ interface on the retreating side of the weld is CDRX.

Fig. 9. (a) Orientation, (b) boundary, (c) kernel maps, (d) distribution of grain size, and (e) distribution of grain boundary misorientation of the retreating side.

Fig. 8, Fig. 9 confirmed that the grain size of the stirred zone on the advancing side (2.6 μm) is lower than the retreating side (9.7 μm). In fact, more equiaxed grains with smaller grain sizes observed for the SZ–AA2024 alloy, compared to the SZ–AA6061 alloy. This can be attributed to the faster aging kinetics observed in the AA2024 alloy [45,47] and the heterogeneous plastic flow in the stirred zone [1,48]. Since the cooling rates after the welding are moderate, the particles re-precipitated heterogeneously at the grain boundaries. These particles can suppress further grain growth during cooling. However, for the AA6061 alloy, aging kinetics is slower than that of AA2024, thus larger grain size observed. The second factor to be noticed is that there is a strain gradient and therefore different temperatures and degree of deformation within the stirred zone. This is owing to the asymmetric material flow within the stirred zone during FSW [1,48].

To study the texture evolutions during FSW of AA2024/SiC/AA6061 joint, {100}, {110}, and {111} pole figures and ODFs of three different locations (center of stirred zone, advancing side, and retreating side) obtained by EBSD plotted in Fig. 10, Fig. 11, respectively. One should note that the PF and ODF of starting materials reported in our previous paper [36]. Fig. 10 indicates that the textures of the center of the stirred zone, advancing side, and retreating side is fully asymmetric due to the non-uniform plastic deformation of FSW. As shown in Fig. 11, the main orientations in the center of the stirred zone can be characterized as the {001} 〈 110 〉 , {223} 〈 112 〉 , {012} 〈 012 〉 , {552}

〈115〉, and {011}〈100〉 with the maximum intensity of 8.3 × R, 8.3 × R, 5.9 × R, 2.9 × R,

and 1.5 × R, respectively. In addition, the 〈 221 〉 //ND fiber and {114} 〈 041 〉 , {011}

〈011〉 , {001}〈100〉 , and {111}〈112〉 orientations generated on the advancing side with the maximum intensity of 8.2 × R, 3.5 × R, 2.3 × R, and 1.5 × R, respectively. Moreover,

the texture components on the retreating side are the {012}〈510〉, {110}〈223〉, {112}

〈 012 〉 , {012} 〈 012 〉 , {012} 〈 100 〉 , {111} 〈 110 〉 , and {112} 〈 111 〉 with the maximum intensity of 4.2 × R, 2.8 × R, 1.9 × R, 1.9 × R, 1.6 × R, and 1.6 × R, respectively.

Fig. 10. {100}, {110}, and {111} pole figures of (a) center of stirred zone, (b) advancing, (c) and retreating sides.

Fig. 11. ODFs of (a) center of the stirred zone, (b) advancing, (c) and retreating sides.

The results reveal a weaker texture for all the three locations compared to the initial samples. Other authors also reported similar results [49,50]. This is owing to the high-temperature plastic deformation during FSW and therefore the occurrence of various types of recrystallization mechanisms such as CDRX, GDRX, and PSN. Therefore, the final texture consists of deformation and recrystallization components. Since the deformation and recrystallization texture orientations are completely different and there is a large number of stable components, the intensity of them is low.

By comparing the ODFs of advancing side for the friction stir welded AA2024/SiC/AA6061 and AA2024/AA6061 (see Ref. [36]), it can be found that the intensity of {001}〈110〉 and

{011}〈211〉 components are considerably reduced in SiC added joint. On the other hand,

the texture intensity of {001} 〈 100 〉 orientation increased. In other words, the shear components weakened and the recrystallization orientation strengthened in the presence of SiC nanoparticles. This may be due to the pinning of grain boundaries by the nanoparticles. In the particle-free joint, the new nucleated grains with {001} 〈 100 〉 orientation as a

recrystallization texture rotate easily towards {001} 〈 110 〉 and {011} 〈 211 〉 shear components. However, the presence of SiC nanoparticles can largely prevent the rotation of the {001}〈100〉 oriented new grains through particle pinning phenomenon. Therefore, the

intensity of {001} 〈 100〉 component on the advancing side of the AA2024/SiC/AA6061 joint increased compared to the AA2024/AA6061 joint.

From Fig. 10(c), the initial texture of AA6061 sheet i.e., {001} 〈 100 〉 , {011} 〈 011 〉 ,

{112}〈011〉, and {011}〈211〉 disappeared after FSW on the retreating side. However,

some shear components including strong {012} 〈510〉 , moderate {110} 〈223〉 , {112}

〈012〉 , {012} 〈012〉 , {012} 〈100〉 , and weak {111} 〈110〉 and {112} 〈111〉 created on the retreating side. This is consistent with shear textures found in the literature [6,23,51]. The formation of multiples shear textures during friction stir welding on the retreating side is due to the severe shear deformation. It is interesting to note the there is no DDRX texture component (i.e. {001}〈100〉) on the retreating side while, as indicated in Fig. 9, the recrystallization occurred. Therefore, it can be concluded that the dominant recrystallization mechanisms on the retreating side are CDRX and GDRX.

For wider industrial applications, the use of lightweight materials such as aluminum alloys and aluminum matrix composites is a driving force for the development of joining methods. The addition of ceramic particles remarkably decreases the weldability of aluminum matrix composites and it is hard to fabricate the defect-free joints. A major concern is the quality of bonding between reinforcement particles and aluminum matrix. The present paper indicated a proper bonding between SiC and aluminum with a crystallographic relationship without porosity. Therefore, this aluminum matrix composite is potentially useful in industrial applications.

4. Conclusions

In the present work, the effect of SiC nanoparticles on the microstructure, texture, and tensile properties of friction stir welded AA2024/AA6061 dissimilar joint studied by EBSD and TEM. The main conclusions are:

1. The SiC nanoparticles homogeneously dispersed in the aluminum matrix and there were no clusters or agglomeration of nanoparticles.

2. After the FSW, the silicon carbide particles did not undergo fragmentation and the shape and size of SiC nanoparticles remained nearly unchanged due to its small size and spherical shape.

3. There was no porosity close to the SiC nanoparticles owing to sufficient plasticization of the aluminum matrix under the chosen experimental condition and good interfacial bonding observed.

4. The crystallographic relationship between aluminum and silicon carbide was indicating the new recrystallized grain tend to nucleate and grow on the crystal plane of SiC nanoparticle.

5. The grain size of the aluminum in the stirred zone decreased from 49.8 μm and 56.5 μm to 7.4 μm due to the occurrence of recrystallization. Most grains in the center of the stirred zone grown. This would appear to be due to the occurrence of CGG mechanism after the stirring.

6. The addition of SiC nanoparticles has reduced the grain size of the SZ, from 10.0 μm to 7.4 μm. The main associated mechanisms in grain refining of the stirred zone during FSW of AA2024/AA6061 with SiC nanoparticles were CDRX and GDRX. There was no PSN in the SZ because the particle size was very small and the temperature of the stirred zone was high.

7. The average size of the precipitate particles in the stirred zone was much smaller than TMAZ due to the dissolution of precipitates during stirring and reprecipitation during cooling. At the TMAZ, owing to its lower temperature, only the relatively small metastable precipitates dissolved and the large precipitates overaged to equilibrium phases and therefore, the particle coarsening occurred in the TMAZ.

8. The textures of the center of the stirred zone, advancing side, and the retreating side was fully asymmetric owing to the non-uniform deformation of FSW. The results exhibited a weaker texture for all the three locations after FSW compared to the as-received sheets.

9. The shear components weakened and the recrystallization orientation strengthened in the presence of SiC nanoparticles.

10. The presence of SiC nanoparticles could largely prevent the rotation of the {001}

〈 100 〉 oriented new grains through particle pinning phenomenon. Unlike the

advancing side, there was no recrystallization texture component (i.e. {001}〈100〉) on the retreating side.

Acknowledgments

The authors acknowledge the funding support of Babol Noshirvani University of Technology through Grant program No. BNUT/370167/98 and BNUT/393044/98.

References

[1] D. Yadav, R. Bauri, A. Kauffmann, J. Freudenberger Metall. Mater. Trans. A, 47 (2016), pp. 4226-4238

[2] Y. Zhao, X. Huang, Q. Li, J. Huang, K. Yan, Int. J. Adv. Manuf. Technol., 78 (2015), pp. 1437-1443

[3] F. Khodabakhshi, A. Simchi, A.H. Kokabi, M. Nosko, F. Simanĉik, P. Švec Mater. Sci. Eng. A, 605 (2014), pp. 108-118

[4] Z.L. Hu, X.S. Wang, S.J. Yuan, Mater. Charact., 73 (2012), pp. 114-123

[5] D. Texier, Y. Zedan, T. Amoros, E. Feulvarch, J.C. Stinville, P. Bocher, Mater. Des., 108 (2016), pp. 217-229

[6] M. Imam, V. Racherla, K. Biswas, H. Fujii, V. Chintapenta, Y. Sun, Y. Morisada

Int. J. Adv. Manuf. Technol., 91 (2017), pp. 1753-1769

[7] M.M. Moradi, H. Jamshidi Aval, R. Jamaati, J. Manuf. Process., 30 (2017), pp. 97-105

[8] P. Kah, R. Rajan, J. Martikainen, R. Suoranta, Int. J. Mech. Mater. Eng., 10 (2015), pp. 26-35

[9] W.M. Thomas, E.D. Nicholas, J.C. Needham, M.G. Murch, P. Templesmith, C.J. Dawes, GB Patent Application No. 9125978.8; 1991 December.

[10] R.S. Mishra, Z.Y. Ma, Mater. Sci. Eng. R, 50 (2005), pp. 1-78

[11] O.S. Salih, H. Ou, W. Sun, D.G. McCartney, Mater. Des., 86 (2015), pp. 61-71

[12] Y.H. Yau, A. Hussain, R.K. Lalwani, H.K. Chan, N. Hakimi Int. J. Miner. Metall. Mater., 20 (2013), pp. 779-787

[13] M.S. Węglowski, Arch. Civ. Mech. Eng., 18 (2018), pp. 114-129

[14] H. Sidhar, N. Kumar, W. Yuan, Friction Stir Welding of Dissimilar Alloys and Materials, Elsevier Science (2015)

[15] E.E. Patterson, Y. Hovanski, D.P. Field, Metall. Mater. Trans. A, 47 (2016), pp. 2815-2829

[16] H. Jamshidi Aval, Mater. Des., 87 (2015), pp. 405-413

[17] P. Karthikeyan, K. Mahadevan, Int. J. Adv. Manuf. Technol., 80 (2015), pp. 1919-1926

[18] M. Abbasi, A. Abdollahzadeh, B. Bagheri, H. Omidvar J. Mater. Eng. Perform., 24 (2015), pp. 5037-5045

[19] M. Tabasi, M. Farahani, M.K. Besharati Givi, M. Farzami, A. Moharami, Int. J. Adv. Manuf. Technol., 86 (2016), pp. 705-715

[20] M. Bodaghi, K. Dehghani, Int. J. Adv. Manuf. Technol., 88 (2017), pp. 2651-2660

[21] R.W. Fonda, K.E. Knipling, Sci. Technol. Weld. Join., 16 (2011), pp. 288-294

[22] R.W. Fonda, K.E. Knipling, D.J. Rowenhorst, JOM, 66 (2014), pp. 149-155

[23] B. Wang, B. Lei, J. Zhu, Q. Feng, L. Wang, D. Wu, Mater. Des., 87 (2015), pp. 593-599

[24] U.F.H.R. Suhuddin, S. Mironov, Y.S. Sato, H. Kokawa, Mater. Sci. Eng. A, 527 (2010), pp. 1962-1969

[25] A.H. Feng, B.L. Xiao, Z.Y. Ma, Mater. Sci. Eng. A, 497 (2008), pp. 515-518

[26] M. Barmouz, P. Asadi, M.K. Besharati Givi, M. Taherishargh, Mater. Sci. Eng. A, 528 (2011), pp. 1740-1749

[27] D. Storjohann, O.M. Barabash, S.A. David, P.S. Sklad, E.E. Bloom, S.S. Babu, Metall. Mater. Trans. A, 36 (2005), pp. 3237-3247

[28] I. Dinaharan, R. Sathiskumar, N. Murugan, J. Mater. Res. Technol., 5 (2016), pp. 302-316

[29] V. Sharma, U. Prakash, B.V. Manoj Kumar J. Mater. Process. Technol., 224 (2015), pp. 117-134

[30] H.R. Akramifard, M. Shamanian, M. Sabbaghian, M. Esmailzadeh Mater. Des., 54 (2014), pp. 838-844

[31] H.S. Arora, H. Singh, B.K. Dhindaw Int. J. Adv. Manuf. Technol., 61 (2012), pp. 1043-1055

[32] J. Humphreys, G.S. Rohrer, A. Rollett Recrystallization and Related Annealing Phenomena (Third edition), Elsevier Science Ltd., Oxford, United Kingdom (2017)

[33] S.M. Ma, P. Zhang, G. Ji, Z. Chen, G.A. Sun, S.Y. Zhong, V. Ji, H.W. Wang J. Alloys Compd., 616 (2014), pp. 128-136

[34] W. Hoziefa, S. Toschi, M.M.Z. Ahmed, Al. Morri, A.A. Mahdy, M.M. El-Sayed Seleman, I. El-Mahallawi, L. Ceschini, A. Atlam Mater. Des., 106 (2016), pp. 273-284

[35] F.C. Campbell Elements of Metallurgy and Engineering Alloys (First edition), ASM International, Materials Park, Ohio (2008)

[36] M.M. Moradi, H. Jamshidi Aval, R. Jamaati, S. Amirkhanlou, S. Ji J. Manuf. Process., 32 (2018), pp. 1-10

[37] A. Heidarzadeh, H. Pouraliakbar, S. Mahdavi, M.R. Jandaghi Ceram. Int., 15 (2018), pp. 3125-3133

[38] M. Sarkari Khorrami, M. Kazeminezhad, A.H. Kokabi Metall. Mater. Trans. A, 46 (2015), pp. 2021-2034

[39] J.F. Guo, J. Liu, C.N. Sun, S. Maleksaeedi, G. Bi, M.J. Tan, J. Wei Mater. Sci. Eng. A, 602 (2014), pp. 143-149

[40] Z. Zhang, R. Yang, Y. Guo, G. Chen, Y. Lei, Y. Cheng, Y. Yue Mater. Sci. Eng. A, 689 (2017), pp. 411-418

[41] M. Mijajlovic, D. Milcic, V. Nikolic-Stanojevic, M. Milcic Informat. Mech., 4 (2012), pp. 65-70

[42] S. Gholamnejad, M. Yektapour, A. Akbarifard J. Mech. Sci. Technol., 31 (2017), pp. 5435-5445

[43] N. Nadammal, S.V. Kailas, J. Szpunar, S. Suwas Metall. Mater. Trans. A, 48 (2017), pp. 4247-4261

[44] L.B. Johannes, I. Charit, R.S. Mishra, R. Verma Mater. Sci. Eng. A, 464 (2007), pp. 351-357

[45] N. Nadammal, S.V. Kailas, J. Szpunar, S. Suwas Metall. Mater. Trans. A, 46 (2015), pp. 2823-2828

[46] A.A. Aksenov, D.G. Eskin, N.A. Belov Iron in Aluminium Alloys: Impurity and Alloying Element CRC Press (2002)

[47] Z. Hu, S. Yuan, X. Wang, G. Liu, Y. Huang Mater. Des., 32 (2011), pp. 5055-5060

[48] J.Q. Su, T.W. Nelson, C.J. Sterling Mater. Sci. Eng. A, 405 (2005), pp. 277-286

[49] M.M.Z. Ahmed, S. Ataya, M.M.E.S. Seleman, H.R. Ammar, E. Ahmed

J. Mater. Process. Technol., 242 (2017), pp. 77-91

[50] J.H. Cho, W.J. Kim, C.G. Lee Mater. Sci. Eng. A, 597 (2014), pp. 314-323

[51] X. Xu, Y. Lu, F. Zheng, B. Chen J. Mater. Eng. Perform., 24 (2015), pp. 4297-4306