Microstructure and Texture Evolution of Aluminum in the Al-Nb/ Ti/Ni Composite Fabricated by the ARB Process

*e Al-Nb/Ti/Ni composite was fabricated from pure Al, Ni, Ti, and Nb sheets by the ARB technology. *e microstructure evolution was observed by scanning electron microscopy, x-ray diffraction, and transmission electron microscopy. *e evolution was evaluated by the electron backscattered diffraction (EBSD) technique. A couple of results we obtained showed that the microstructure of Al changed from equiaxed grains to a lamellar structure, and the grain size in the ND decreased gradually. Finally, the average grain size in the ND was 0.31 μm. Additionally, the fraction of HAGBs increased after the third pass, resulting from the dynamic recovery and the shear bands. *e texture evolution was tested by electron backscattered diffraction. After the fourth pass, the Al exhibited a combination texture of rolling texture and shear texture. *e rolling texture components were composed of Copper{112}<111>, Dillamore{4 4 11}<11 11 8>, S{123}<634>, and Brass{011}<211>, and the shear texture components were Rotated Cube {001}<110> and {111}//ND. *e microhardness of Ni, Ti, Nb, and Al was improved in the ARB process and finally reached 226.4, 246.3, 187.2, and 44.2 HV, respectively.


Introduction
Metal matrix composites (MMC) combine the advantages of each component and makes up for the shortcomings, which has excellent comprehensive properties that cannot be matched by a single metal or alloy so that it becomes a focus in materials science today. Metal matrix composites are prepared by the bonding technology to achieve metallurgical bonding between two or more metals with different characteristics. [1][2][3][4][5].
At present, the major methods of producing metal composites are casting, powder metallurgy, spray deposition, etc. However, most of these technologies require expensive equipment and complex processes, limiting their industrial application [6][7][8][9][10]. e accumulative roll bonding (ARB) technology is a kind of severe plastic deformation (SPD) process which can fabricate bulk composites continuously and has a good potential for industrialization. Compared with other SPD processes, ARB technology, it can produce not only ultrafine grained metals but also metal matrix composites.
Aluminum is generally used as a matrix for metal matrix composites due to its lightweight and good formability. Recently, a number of researchers have reported the Almatrix composite produced by the ARB process, such as Al-Ti [11], Al-Cu [12], Al-Ni [13,14], Al-Mg [15][16][17], Al-Mg-Ti [18], Al-Ti-Nb [19], Al-Cu-Ni [20], Al-Cu-Sn [21], and Al-Cu-Zn [22]. However, the research about multimetal composites fabricated by the ARB process is insufficient. Titanium aluminum alloys are widely used in the automobile and aviation industries due to its low density-high strength, excellent corrosion resistance, and high-temperature performance. Adding nickel to the titanium aluminum alloy can improve its corrosion resistance, and adding niobium can enhance its high-temperature performance [19,23]. erefore, the Al-Nb/Ti/Ni composite will have excellent comprehensive properties. Metal sheets are arranged in a symmetrical form in order to ensure the stability of the composites. Aluminum sheets located on the surface and in the middle can improve the feasibility of the ARB process. erefore, the metal sheets were stacked in the order of Al/ Ni/Ti/Al/Nb/Al/Ti/Ni/Al. Texture analysis is a very critical means to explore the grain oriental of metals during plastic deformation so that we can more clearly understand the changes in the microstructure of the material. Existing researches have focused on the texture evolution of single metals or alloys [24][25][26][27][28][29][30][31] prepared by the ARB technology. However, the investigation on the texture of multimetal composites is insufficient.
In previous studies, we have prepared Ni-Ti-Al-Cu composites and investigated their microstructure evolution and mechanical properties. We will pay more attention to the texture evolution of the Al-Nb/Ti/Ni composite in this work.
e EBSD technology was used to investigate the texture evolution of the Al-Nb/Ti/Ni composite.

Experimental Procedures
e chemical compositions and dimensions of raw sheets (pure aluminum, nickel, titanium, and niobium) are depicted in Table 1.
Al, Ni, Ti, and Nb sheets with a thickness of 0.2 mm were annealed, respectively, at 623 K, 1073 K, 1073 K, and 1473 K, for 30 minutes. e surface of the metals needed to be polished with a steel brush due to the oxides formed during the annealing process and then degreased in alcohol and acetone. Four aluminum sheets, two nickel sheets, two titanium sheets, and one niobium sheet were stacked (as depicted in Figure 1). e stacked sheets were fastened by steel wires at the four corners to keep stability. e rolling process was conducted on a two-high mill with a 50% reduction at room temperature. e diameter of the roller was 350 mm, and the rolling speed was 10 rpm. e bonded samples were cut into two, degreased, and brushed. e samples were stacked and continued to roll with a target of 50% reduction. e experiment procedure was repeated for four times without lubrication. e Al-Nb/Ti/Ni composite was fabricated successfully after four passes accumulative roll bonding process. e Von Mises equivalent strains for each pass of the ARB process are 0.8, 1.6, 2.4, and 3.2, respectively. e complete accumulative roll bonding process is illustrated in Figure 1.
Microstructure evolution of the Al-Nb/Ti/Ni composite was observed by scanning electron microscopy (SEM, JSM-7001F) and transmission electron microscopy (TEM, JEM2010, 200 kV). e texture evolution of the Al layer in the ARB process was analyzed using an electron backscattered diffraction (EBSD) technique, and the EBSD test was conducted at a scanning electron microscopy (SEM, JSM-7001F) equipped with a TSL-OIM EBSD analysis system. e EBSD measurements were performed at a voltage of 20 kV, a tilt angle of 70°, a magnification of 2500, and a scan step size of 0.06 microns. e EBSD samples were electropolished in an ethanol and perchloric acid (5 vol.%) solution at 20 V for 20∼30 s. e microhardnesses (HV) of the composites were measured by a hardness tester at a load of 50 g for 10 s. Figure 2 is the SEM image of the Al-Nb/Ti/Ni composite on the RD-ND plane. e microstructural evolution of various layers of the Al-Nb/Ti/Ni composite after the different passes (1, 2, 3, and 4) is demonstrated in Figure 2. A good bonding between individual layers could be observed in Figure 2(a). e size of nickel, titanium, and niobium decreased with the further deformation, and the thickness of the niobium fragments was visibly smaller than that of Ni and Ti due to the smaller deformation resistance. In Figure 2(d), Nb fragments with extremely small size could be observed.

Microstructures.
In the first pass, the composite material also exhibited a layered structure with the locally uneven deformation. Obvious neckings and fractures appeared in Ti, Ni, and Nb layers when the ARB pass increased. A couple of investigations [18,19] reported similar results. e plastic instability occurred because of the different mechanical properties of individual layers, and then nenckings eventually breaking were visible in the hard phase. Since the metals Ni, Ti, and Nb are hard phases relative to the matrix Al, then the Ti, Ni, and Nb layers necked and fractured with the further deformation. In the fourth pass, the metal fragments (nickel, titanium, and niobium) were homogeneously distributed in the multimetal composite.
Another reason for the formation of metal fragments was that the reinforcements (nickel, titanium, and niobium) were cut by the shear band during the plastic deformation process, and finally, the metal fragments are  formed. In Figure 2(d), we could observe that there were shear bands in the matrix. e shear bands were at an angle to the RD and the angle was generally about 30°-45°. Because of the different mechanical properties of the four metals, a shear band was generated with the effect of shear forces [18]. At last, the reinforcements (Ni, Ti, and Nb) were further broken and distributed homogeneously due to the effect of shear bands. Interfacial bonding is a key issue of composites. e improvement of interface strength is beneficial to mechanical properties. Intermetallic compounds will be generated in different metal interfaces. In composites, brittle intermetallic compounds formed at the interface seriously affect the overall performance. A large amount of heat is released during the severe plastic deformation, which will lead to the formation of brittle intermetallic compounds. e XRD tests were conducted to check if there was a new phase. e results showed that there were only peaks of the raw metals in the XRD pattern, and no new intermetallics were observed (as shown in Figure 3.) e orientation maps on the RD-ND plane after the 1st, 2nd, 3rd, and 4th pass are illustrated in Figure 4. Since the severe plastic deformation of the ARB process, an electropolishing treatment should be conducted on the surface of samples before the EBSD test. e EBSD data is collected by an OIM software. In Figure 4, e correspondence between the colors and the orientations is shown in the triangle. Al grains were initially compressed, gradually elongated, and finally formed a lamellar structure. According to Alizadeh [32,33], typical dislocation cells were generated in the initial stage. With the increasing ARB pass, a lamellar structure forms. e grain size in the ND was decreased gradually. e grain size in the ND of Al can be calculated, and results are demonstrated in Figure 5. e range of size distribution was smaller and nearer to Y-axis, indicating that the grain size in the ND decreased. Based on the data in Figure 5, the average grain size is calculated, and the results are shown in Figure 6. e average grain size in the ND decreased significantly. After the fourth pass, the average grain size in the ND was 0.31 μm. Such a result suggests that rolling deformation plays a critical role in the grain size change in the ND.  Interestingly, we can find that the rate of grain size reduction in the ND is slower gradually in Figure 6. e reasons for this result may be more complicated and related to various factors, such as properties and status of metals. One of the most important reasons is the increase in metal strength and deformation resistance in severe plastic deformation. Due to the effects of work hardening and fine grain strengthening, it is becoming more and more difficult for metals to deform. Consequently, the rate of grain size reduction decreases. e bright-field TEM micrographs and corresponding selected area diffraction (SAD) patterns on the rolling direction (RD) normal direction (ND) plane are shown in Figure 7. e microstructure of Al observed in Figure 7 was consistent with that in Figure 4. Moreover, the average grain size calculated in Figure 6 was also confirmed in Figure 7. Obviously, the microstructure of the annealed Al sheets was changed from equiaxed grains to a lamellar structure during the ARB process. Grains had not been wholly elongated along the rolling direction after the first pass (as shown in Figure 7(a)). Dislocation tangles and dislocation cells are formed in the metal. By increasing ARB passes, Al grains were gradually elongated along the RD, and then the elongated grains were subdivided into smaller grains by boundaries [32][33][34][35].
In Figure 7(a), a great number of dislocations were visible. After the second pass, the dislocation density was increased and the grains became finer. e dislocation density was significantly reduced in Figure 7(c), and there were a large number of clear grains with a low density of dislocations (indicated by arrows) among other grains with high dislocation density [36,37]. Deformation heat during the ARB process provided energy for the dynamic recovery of aluminum. e temperature of aluminum rose, and then the dynamic recovery occurred in the ARB process. e dynamic recovery easily occurred in metals with high stack fault energy (SFE), such as aluminum and aluminum alloys. Figure 8 depicts the misorientation angle distributions after different ARB passes. e fraction of HAGBs after the 1st, 2nd, 3rd, and 4th pass was 0.407, 0.34, 0.515, and 0.646, respectively. In the initial stage, the structure of Al changed from a recrystallized structure to a lamellar structure, and high-angle grain boundaries were reduced due to the rolling deformation. In the cold-rolling process, the dislocation density accumulates and the internal distortion energy increases, which causes the initially clear grains to be divided into substructures by a large number of entangled dislocations. ese substructures are small-angle grain boundaries resulting in the decrease of HAGBs. After the third pass, the occurrence of dynamic recovery led to an increase in HAGB fraction. At the same time, the elongated grains could be divided by the local shear bands, forming the equiaxed grains, and more HAGBs are generated [19,24]. Several investigations [38][39][40] reported that the grain refinement could form high-angle grain boundaries. Figure 9 shows the orientation distribution function (ODF) of Al in the Al-Nb/Ti/Ni composite. e φ2 � 45°, 65°, 90°sections are illustrated in the figure. For face-centered cubic (FCC) metals, these sections can contain all the relevant texture components [19,26,41]. Figures 9(a)-9(d) show the ODF diagrams of the aluminum after the first, second, third, and fourth pass, respectively, and Figure 9(e) shows the ideal texture composition of Al. e texture distribution on the RD-ND plane was complex because of the repetition of the rolling process. e texture components obtained in the 1st pass can be characterized as the strong cube {001}<100> and weak S{123} <634>. e Al layer had a strong recrystallization texture due to a full annealing treatment before rolling, and then a small amount of recrystallization texture changed to the rolling texture after the first pass. e recrystallization texture disappeared and completely transformed into the   is result was different from the texture components in the aluminum ARB process. Pirgazi and Kim [26,27] reported that the main textural components were the Copper {112}<111> and Dillamore{4 4 11}<11 11 8> components in aluminum alloys ARB process, and no obvious shear texture component was observed. e formation of shear texture component was related to two main factors [19,26]. One was the generating of shear bands in the composite. e other was the shear deformation caused by great friction between the sample and the roller. Due to the repetition of the rolling, cutting, and stacking of the ARB, the surface of samples moved to the center in the next pass. erefore, the shear texture also appeared in the center.

Texture Evolution.
For the cold-rolling process of aluminum or aluminum alloys [24][25][26][27][28][29][30][31], the main rolling texture components were β-fiber and τ-fiber. Generally, the intensity of β-fiber and τ-fiber increased with the increasing amount of cold-rolling deformation. Pirgazi et al. [27] reported that the intensity of β-fiber and τ-fiber increased significantly in the deformation. Kim et al. [26] also observed similar results of the intensity of τ-fiber after four passes of the ARB process.
However, no obvious increase in rolling texture intensity could be seen in the Al-Nb/Ti/Ni composite (Figure 9). e  reason why the rolling texture intensity is not significantly improved is the effect of the shear bands. Qu et al. [19] found that shear bands led to not only a 15°deviation of rolling texture from the RD but also a decreasing intensity of rolling texture in aluminum. Chang et al. [15] reported that the shear bands reduced the intensity of rolling texture and t changed the texture orientation in the Mg/Al composite. Li et al. [42] observed a similar phenomenon in the Mg/Al composite. Furthermore, Toroghinejad et al. [24] showed that the intensity decrease of major texture components could be explained by the continuous recrystallization. erefore, the dynamic recovery in this ARB process might influence the intensity of rolling texture components. Figure 10 shows the microhardness variations of the Al-Nb/Ti/Ni composite. e microhardness of nickel, titanium, niobium, and aluminum was enhanced significantly from 157. 9,161.3,85.6,and 23.8 HV to 199.5,213.9,169.4,and 35.6 HV,respectively. e rapid increase of microhardness after the first pass was due to the hardening. In the subsequent deformation, the microhardness change was mainly determined by the fine grain strengthening. Due to the dynamic recovery of Al, its microhardness growth rate was less than the other three. Finally, the microhardness of Ni, Ti, Nb, and Al reached 226.4, 246.3, 187.2, and 44.2 HV.
One of the reasons for the increasing hardness in the composite may be the different thermal conductivities between the four metals [18,43]. Generally, the temperature of the sample increases due to plastic deformation and the friction between the roller and the sample. Finally, due to the different cooling rates of the four metals, regions with severe thermal stress are generated. According to the thermal expansion coefficients (CTE) of each layer (Ni, Ti, Nb, and Al are 13, 8.6, 7.1, and 23 μm/mK, respectively), dislocations occur at the interface during deformation, thereby improving the microhardness of the composite.
In addition, we can observe that the microhardness growth rate of Al is the slowest in Figure 10. It is majorly due to the dynamic recovery in the deformation process, which has a softening effect. At the same time, pure aluminum as a relatively soft metal is difficult to reach a high level of microhardness. Furthermore, niobium has the fastest hardness growth among the other three metals. It is not difficult to find out the reasons through the analysis of the microstructure of composites (Figure 2). e size of Nb fragments is significantly smaller than that of nickel and titanium, indicating that niobium obtains greater deformation during the ARB process. erefore, the microhardness of niobium gets a larger increment due to the effect of strain accumulation. In general, the changes in the mechanical properties of the individual metals in the composites are still determined by their own characteristics.

Conclusions
In summary, the multimetal composite was successfully fabricated by the ARB technology. e following conclusions were obtained.
(1) Nickel, titanium, and niobium fragments were homogeneously distributed in the multimetal composite, and the average size of Nb fragments was smaller than the other two. In the fourth pass, Nb fragments with extremely small size were distributed in the Al matrix. Additionally, Neckings and fractures appeared in the reinforcements (Ni, Ti, and Nb). ere was no new intermetallic compound generated in the rolling at room temperature.
(2) In the ARB process, the microstructure of Al changed from equiaxed grains to a lamellar structure, and the grain size in the ND decreased gradually. Finally, the average grain size in the ND was 0.31 μm. (3) After the third pass, the dynamic recovery occurred in the Al matrix, resulting in an increase of the HGAB fraction and the decrease of the dislocation density. Additionally, the increasing fraction of HGABs was also contributed by shear bands.

Data Availability
e experimental data used to support the findings of this study are included in the article. e other data are available from the corresponding author upon request.  Journal of Chemistry