Search for articles:
Sahayam Joyson Abraham
, Isaac Dinaharan, Jebaraj David Raja Selvam, Esther Titilayo Akinlabi
Corresponding authors:
Received: 2018-05-23
Revised: 2018-07-28
Online: 2019-01-10
Copyright: 2019 Editorial board of Acta Metallurgica Sinica(English Letters) Copyright reserved, Editorial board of Acta Metallurgica Sinica(English Letters)
More
Abstract
Rutile (TiO2) particle-reinforced aluminum matrix composites were prepared by friction stir processing. The microstructure was studied using conventional and advanced characterization techniques. TiO2 particles were found to be dispersed uniformly in the composite. Clusters of TiO2 particles were observed at a higher particle content of 18 vol%. The interface between the TiO2 particle and the aluminum matrix was characterized by the absence of pores and reactive layer. Sub-grain boundaries, ultra-fine grains and dislocation density were observed in the composites. TiO2 particles improved the mechanical properties of the composites. However, a drop in tensile strength was observed at a higher particle content due to cluster formation. All the prepared composites exhibited ductile mode of fracture.
Keywords：
Every kind of industry at the present has been seeking to reduce the weight of the components and improve the specific strength. Aluminum matrix composites (AMCs) satisfy the requirement in the pursuit of new materials. AMCs are commonly reinforced with ceramic particles [1, 2]. SiC and Al2O3 were regularly used as reinforcement phases due to low cost and good wettability with molten aluminum [3]. The improvement in existing and the advent of new production techniques enable to reinforce several potential reinforcement particles to improve the performance of AMCs without deviating from the primary objective of light weight. Titanium dioxide (TiO2) is found in its natural mineral form of rutile which is easily available, cheaper and possesses desirable mechanical, wear and corrosion properties [4, 5].
Al/TiO2 AMCs were prepared using conventional production routes by some investigators. Ramesh et al. [6] manufactured AA6061/(2-10) wt% TiO2 AMCs using stir casting. They predicted wear coefficients using Archard and Yang relationships and compared with experimental wear coefficients. Sivasankaran et al. [7] produced AA6061/10 wt% TiO2 AMCs using mechanical alloying and investigated the influence of the milling time on the morphology, density and flow characteristics of the composite powders. Soltani and Hoseininejad [8] developed Al-Cu/(0.5-3.5) wt% TiO2 AMCs using stir casting and showed an improvement in wear resistance of the composites due to TiO2 particles. Shin et al. [9] prepared Al/(15vol%) TiO2 AMCs by consolidated milled powders using hot rolling and studied the interface characteristics in detail. Sivasankaran et al. [10] fabricated AA6061/10 wt% TiO2 AMCs using mechanical alloying. They used nanocrystalline and coarse grain aluminum powders to form a trimodaled composite powders and evaluated the microstructure and tensile properties. Kumar et al. [11] synthesized Al/(0-12) wt% TiO2 AMCs using mechanical alloying and presented the microstructure and dry sliding wear behavior.
Although Al/TiO2 AMCs were produced by several methods, stir casting was inexpensive to produce the composite but the wettability of TiO2 particles with molten aluminum is poor and needs an additional coating [12] of particles resulting in higher cost. The distribution of TiO2 particles in the aluminum matrix was uneven and clustering was unavoidable [6, 8]. Mechanical milling provides desirable distribution of the particles [7, 10, 11]. Nevertheless, it is time to consume high amount of energy for the production of the composites [13]. Friction stir processing (FSP) has been materialized as an energy-efficient method to synthesize AMCs [14]. The interaction between a non-consumable rotating tool and the substrate material produce severe plastic deformation. The packed ceramic particles are mechanically mixed with the deformed substrate by the tool. The temperature rise is insufficient to melt the substrate material. The solid-state nature assists to remove the defects of conventional casting routes. The physical and the chemical properties of the ceramic particles do not influence the process significantly. The density gradient and the wettability issues do not affect the resultant distribution of the reinforcement phase in the composite [15].
Several ceramic particles including SiC [16], Al2O3 [17], TiC [18], B4C [19], TiB2 [20], ZrO2 [21], Si3N4 [22] and TiN [23] were successfully reinforced to prepare AMCs using FSP. However, reports on TiO2-reinforced AMCs are very little. Mirjavadi et al. [24] incorporated TiO2 particles along the joint line to improve the tensile strength of friction stir-welded AA5083 aluminum alloy. Some investigators used small quantity of nanolevel TiO2 particles to synthesize in situ Al3Ti and Al2O3 particles using FSP by applying six passes [25, 26]. Madhu et al. [27] attempted to prepare Al/TiO2 AMCs through FSP by applying six passes to change the initial precursor titanium tetra-isopropoxide Ti{OCH(CH3)2}4 into TiO2 particles. Gudla et al. [28] introduced TiO2 particles to a depth of 2 mm using FSP on pure aluminum and observed the anodizing behavior. The present work is focused to use submicron-level TiO2 particles to produce AMCs by FSP and assess the role of TiO2 particles in microstructure and tensile behavior. A detailed microstructural characterization was reported.
10-mm-thick aluminum alloy AA6063 plates having a width of 50 mm and a length of 100 mm were procured to prepare the composites. Table 1 represents the composition of the aluminum alloy plates. Rectangular slots of cross section x mm (x = 0, 0.4, 0.8 and 1.2 mm) × 5.5 mm were machined throughout the length of the plates. The grooves were stuffed with TiO2 particles (~ 0.56 µm). Figure 1 displays SEM, EDAX and particle size analysis of TiO2 particles. The size of TiO2 particles was measured using a particle size analyzer (Malvern Mastersizer 3000). The variation in the geometry of the groove leads to designed variation in volume fraction which was computed according to the following expression as reported in literature [29] and observed to be 0, 6, 12, and 18 vol%.
Theoretical volume fraction=(Area of groove/Projected area of tool pin)×100 (1)
Area of groove=Groove width×Groove depth. (2)
Table 1 Chemical composition of AA6063 aluminum alloy (wt%)
| Mg | Si | Fe | Mn | Cu | Cr | Zn | Ti | Al |
|---|---|---|---|---|---|---|---|---|
| 0.45 | 0.35 | 0.15 | 0.02 | 0.01 | 0.09 | 0.04 | 0.02 | Bal. |
Fig. 1 a SEM, b EDAX analysis, c particle size distribution map of TiO2 particles
FSP was conducted in accordance with the procedure [30] portrayed in Fig. 2 using a robust computer numerical controlled vertical machining center. Table 2 documents the processing conditions. Specimens for microstructural characterization were machined from the friction stir-processed plates. They were mounted (STRUERS Citopress), polished (STRUERS Labopol) and etched with Keller’s reagent. The etched specimens were observed using an optical microscope (Olympus BX51M), field emission scanning electron microscope (Carl Zeiss SIGMAHV), electron backscatter diffraction (EBSD) and high-resolution transmission electron microscope (JEOL JEM 2100). EBSD was carried out in a FEI Quanta FEG SEM equipped with TSL-OIM software. Mini tensile specimens of gauge length of 30 mm, width of 4 mm and thickness of 4 mm were machined from the FSP zone by wire EDM. The ultimate tensile strength (UTS) was estimated using a computerized tensile tester (Instron 1195). The fracture surfaces were studied using FESEM.
Fig. 2 FSP procedure to fabricate the composite: a compacting the groove with particles, b closing the groove top using a pinless tool, c processing using a tool with pin (two passes were applied in opposite directions)
Table 2 Processing conditions
| Parameter | Values |
|---|---|
| Rotational speed (rpm) | 1600 |
| Traverse speed (mm/min) | 60 |
| Shoulder diameter (mm) | 18 |
| Pin diameter (mm) | 6 |
| Pin length (mm) | 5.7 |
| Pin shape | Straight cylindrical threaded |
| Tool material | HCHCr steel |
Typical photograph of a friction stir-processed AA6063 aluminum alloy plate is shown in Fig. 3. The crown surface looks flat and smooth. The surface is continuous throughout the length of FSP line. No abnormal surface flaws are noticed. The crown appearance ensures proper material flow during processing across the stir zone. Surface irregularities act as precursor for further macroscopic-level defects in the stir zone and damage the integrity of the composite structure. Arc like features matching to the periphery of a half circle are observed on the crown. This is formed due to the scrubbing of the tool shoulder over the plasticized material. The average distance between any two arcs corresponds to revolutions of the tool per unit distance of FSP length [31].
Fig. 3 Typical photograph of friction stir-processed AA6063/TiO2 AMC plate showing crown appearance
Figure 4 presents the macrostructures of AA6063/TiO2 AMCs. The stir zone area which holds the AMC is apparently evident in all the figures. The boundary of the stir zone is delineated with black line. The identity of the groove on the aluminum plate prior to FSP is completely lost. It shows that the formation of the composite is complete and the plasticized material flow is continuous. The rubbing of the tool shoulder and the shearing of the pin creates adequate frictional heat which causes the aluminum matrix around and below the tool to plasticize. The rotating and the translation movement of the tool transfer the plasticized aluminum from advancing side to retreading side. This transportation of material flow forces the groove to cave in and combines the compacted TiO2 particles with the plasticized aluminum alloy. The speed at which the tool rotates and translates decides the intensity of mixing leading to the formation of the composite. It is evident from Fig. 4 that the stir zone is free from defects such as pin holes, tunnels, voids, cracks and kissing bonds. These defects are frequently encountered in FSP [14]. Defects shrink the area of the stir zone and cause the composite vulnerable to sliding wear and tensile loading. None of these defects appeared in the stir zone. Defects occur owing to multiple factors which are not limited to inadequate heat generation, material flow and consolidation. The process parameters govern these factors significantly. Nonexistence of defects can be related to chosen process parameters used in this work.
Fig. 4 Macrostructures of AA6063/TiO2 AMCs containing TiO2: a 6 vol%, b 12 vol%, c 18 vol%
Figure 5 displays example SEM micrographs of the composite specimens having different contents of TiO2 particles. The micrographs reveal that the TiO2 particles are dispersed over the stir zone area effectively. Every unit area of the micrograph is filled with TiO2 particles. There are no blank spaces (regions without any reinforcement of particles) of considerable size. The nature of the dispersion of TiO2 particles in the aluminum matrix can be equitably reckoned as uniform dispersion. Unlike stir casting method, the density variation between the aluminum matrix and the TiO2 particle do not play a major role in dictating the final dispersion in the composite. The dispersion is predominantly affected by the process parameters compared to the physical properties of the reinforcement. Rana and Badheka [32] reached a conclusion that tool rotational speed and number of passes alter the reinforcement distribution remarkably. Hence, a desirable dispersion in the micrographs can be directly related to the use of proper combination of process parameters which were optimized in an earlier work [33]. Few clusters of agglomerated TiO2 particles are noticed in Fig. 5c. It suggests that more number of passes are required to improve the dispersion at a higher content of TiO2 particles. But, increased number of passes accumulates more strain energy and may lead to the conversion of TiO2 particles into Al2O3 and Al3Ti [25]. This factor limited the number of passes to two in the present study to retain TiO2 particles without any kind of in situ reaction.
Fig. 5 FESEM micrographs of AA6063/TiO2 AMCs containing TiO2 particles: a 6 vol%, b 12 vol%, c, d 18 vol% (arrows point clusters)
Although TiO2 particles are dispersed over the entire aluminum matrix, they do not align themselves along the grain boundaries and show a continuous network of segregated particles. In spite of the observation that the grain boundaries are not revealed in Fig. 5, some TiO2 particles may be located directly on the grain boundaries. The above statements give a hint that the dispersion is mainly intragranular, i.e., most of the particles are retained inside the grain boundaries. The behavior of a ceramic particle inside a molten aluminum and a plasticized aluminum is entirely different. The density variation and the force of the solidification front allow a movement of particles during the former process. The ceramic particle cannot move by itself due to the above factors in the latter process. Subsequently, the unrestrained motion is completely arrested and the particle is forced to follow the trajectory of the rotating tool to locate a place where it is led. This ensures that the possibility of the occurrence of segregation in FSP is distant.
Figure 5d shows a SEM micrograph of AA6063/18 vol% TiO2 AMC at a higher magnification. TiO2 particles are merged with the matrix in an excellent manner. There are no undesirable features such as pockets of unfilled space (similar to pores in castings) or reaction layer between them. The mechanism of emergence of pores in FSP is contrary to pore formation in stir casting. Unlike absorption of atmospheric gases by the molten aluminum, the nonalignment of material flow over the particle surface form pore like structure. The smaller size and spherical shape of TiO2 particles did not obstruct or deflect the pathway of plasticized aluminum. The entire surface of the TiO2 particles was immersed by the plasticized aluminum which might have avoided the formation of pores. A reaction between a ceramic particle and the aluminum matrix commences if either the peak temperature or the applied strain rate under the processing conditions is excessive. A closer examination of the interface shows no signs of an emerging diffusion layer which may hint the initiation of such reaction. This suggests that the applied processing conditions limited the peak temperature and the strain level below the limit which is required for initiating an interfacial reaction. Interface contributes significantly to allow smooth transfer of the applied load into the reinforcement particle during tensile or wear test. AMCs will fail easily if the interface contains pores or reactive layer in spite of uniform dispersion of reinforcement particles in the aluminum matrix.
It is observed in Fig. 5d that the average size of TiO2 particles after FSP is lower compared to the initial size measured using particle size analyzer. Ceramic particles are subjected to a refinement in size and shape during FSP [15]. The mechanism is presented below. The aluminum matrix is a highly deformable material which has higher toughness. On the other hand, the ceramic particle TiO2 cannot be easily deformed and possess lower toughness. FSP is treated as one of the severe plastic deformation (SPD) processes. The strain rate produced in FSP is highest among all conventional SPD processes. The aluminum matrix would respond to this enormous strain by easily deforming. TiO2 particles would not withstand above a particular strain rate, and the particles simply fracture due to its innate brittleness. This process creates variation in size and shape after FSP. The fine debris generated during breaking of TiO2 particles also mix well with the plasticized aluminum and dispersed uniformly in the aluminum matrix. The tool has a tendency to abrade the particles and produce a rounding off effect. The breaking of TiO2 particles would lead to a reduction in average interparticle distance which will boost up mechanical properties by enhancing more interaction among the particles.
Figure 6 represents optical micrographs of AA6063/12 vol% TiO2 AMC which was taken from top, middle, bottom and advancing side of the stir zone. The variation in the dispersion of TiO2 particles across the stir zone is observed to be minimum. The small variation at the bottom is explained later. A constant dispersion of reinforcement particles is preferred to avoid anisotropic behavior of the composite. It is difficult to achieve constant dispersion in stir casting method due to the tendency of a ceramic particle to either float or sink in the molten aluminum according to the density variation. Poor selection of process parameters may lead to uneven dispersion of reinforcement particles across the stir zone [19, 22, 34]. The solid-state nature of the process and appropriate combination of process parameters and tool design produced a preferable dispersion across the stir zone. Successive layers of high and low dispersion are witnessed at the bottom of the stir zone (Fig. 6d). Those layers are part of the evolved onion ring structure during FSP [35]. The interaction between circular and vertical material flow at the chosen process parameters forms such a structure at the bottom of the stir zone.
Fig. 6 Optical photomicrographs of AA6063/TiO2 AMC at various locations within the stir zone: a top portion, b interface at the advancing side, c middle portion, d bottom portion
Figure 7 depicts the EBSD images of as-received AA6063 and friction stir-processed AA6063/12 vol% TiO2 AMC. The base material shows a bimodular grain structure consisting of elongated and fine grains. This structure is due to the rolled condition of the base plates. The average grain size was roughly 30 µm. In contrast, the prepared composite shows equiaxed fine grains. The average grain size was roughly 5.5 µm. There are two factors which can contribute to the evolution of such a fine structure. The primary factor is related to the inherent nature of the FSP process which is widely acknowledged as dynamic recrystallization. FSP is categorized as a hot working process in which the temperature due to friction and deformation soar above the recrystallization temperature of the substrate material. FSP is also recognized for its extreme deformation of material during processing. Both the aforementioned factors invoke the occurrence of dynamic recrystallization. It is established in literature that dynamic recrystallization occurs either in a continuous or discontinuous process [36]. The deformed material recovers rapidly by means of complete destruction of dislocations in the former process. Fresh and comparatively moderate or zero dislocation grains are formed in the latter process. The factors which switch on the particular dynamic recrystallization process are strain rate history and stacking fault energy. Higher strain rate and higher stacking fault energy, respectively, encourage and discourage the occurrence of discontinuous dynamic recrystallization. Aluminum alloys are characterized by high stacking fault energy, and hence, the dynamic recrystallization process can be counted as continuous. Secondly, the reinforcement of TiO2 particles may promote pinning effect known as Zener pinning into operation as illustrated in Fig. 8. The free displacement of grains during recrystallization and grain growth is pinned by the incorporation of reinforcement particles [15, 37]. Therefore, the prepared composite exhibits fine-grained structure.
Fig. 7 EBSD (IPF + grain boundary) maps of AA6063/TiO2 containing TiO2 particles: a 0 vol%, b 12 vol%
Fig. 8 Schematic illustration of a particle distribution and formation of grains, b pinning of grain boundaries by particles (irregular polygonal shape represents grains, black circular shape represents reinforcement TiO2 particles)
Figure 9 displays the TEM micrographs of AA6063/12 vol% TiO2 AMC. It reveals several features present in the composite. Extremely fine grains are seen in Fig. 9a. The grains are at different stages of the recovery process. A closer look shows the existence of sub-grain boundary structure at the boundary of the grains. Few TiO2 particles of various sizes are observed in Fig. 9b, c. The discrepancy in size might be related to the breaking up of the particles during processing. Interface between the particle and the aluminum matrix appears discrete without any evidence for either diffusion or reaction. Large amount of dislocation is present in the composite (Fig. 9b-d). Some are found in the matrix, and some are found close to the particle. The creation of dislocation density is related to deformation of the FSP process and the strain misfit. The extreme deformation automatically gives birth to dislocations in the aluminum matrix. Since TiO2 particles cannot deform at the same rate of aluminum matrix, the dislocations are generated to accommodate the differential strain rate. Those dislocations are located closer to the particle (Fig. 9b). All the grains do not have the same degree of dislocations. This observation suggests a variation in the level of energy storage which may dictate the orientation of grains.
Fig. 9 HRTEM micrographs of AA6063/12 vol% TiO2 AMC showing: a sub- and ultra-grain boundaries b, c particle interface and dislocations, d dislocations in the aluminum matrix
Figure 10 presents the mechanical properties of the prepared AA6063/TiO2 AMCs as a function of TiO2 content. The microhardness (Fig. 10a) improved due to the inclusion of submicron-level TiO2 particles. The microhardness was tested to be 62 Hv at 0 vol% and 142 Hv at 18 vol%. The improvement in microhardness at 18 vol% TiO2 particles is 129%. The microhardness distribution across friction stir-processed aluminum alloy AA6063 with 12 vol% TiO2 particulates taken at 2.5 mm from the shoulder region is depicted in Fig. 10b. The variation in hardness across the stir zone is minimum due to the consistent dispersion of TiC particulates. The UTS was found (Fig. 10c, d) to improve up to 12 vol% and dropped with further increase in the content of TiO2 particles. The UTS was estimated to be 222 MPa at 0 vol% and 325 MPa at 12 vol%. The improvement in UTS at 12 vol% TiO2 particles is 46.4%. The UTS of AA6063/18 vol% TiO2 AMC was 288 MPa. Overall, there was a strengthening of the composite due to the incorporation of TiO2 particles in the aluminum matrix. Various mechanisms might have operated to improve the mechanical properties which are discussed one by one as follows. There was a uniform dispersion of fine-size TiO2 particles in the aluminum matrix. The dislocation is forced to bow around the fine TiO2 particles which is known as Orowan strengthening [38, 39]. The uniform dispersion changes the path of the dislocations multiple times and the motion is retarded. Secondly, the dislocations generated due to plastic deformation and strain misfit establishes strain fields in the composite. The motion of dislocation faces opposition from those strain fields. Thirdly, the excellent interfacial bonding between the aluminum matrix and the TiO2 particle allows a smooth transfer of applied load to the particles. TiO2 particle may eventually fail due to fracture or pull out. Finally, the generation of extremely fine equiaxed grains strengthens the composite according to well-known Hall-Petch equation. The increase in the content of TiO2 particle further multiplies the contribution from each of the above factors which boost up the mechanical properties additionally. However, the drop in UTS of AA6063/18 vol% TiO2 AMC can be related to clusters found in the microstructure. The elongation was measured to be 25.5% at 0 vol% and 12% at 18 vol% as plotted in Fig. 10d. The percentage elongation reduced with increased content of TiO2 particles. The reduction can be related to the deformation behavior of aluminum and TiO2 particles in response to tensile load. Aluminum is highly deformable while TiO2 particles do not deform easily. TiO2 particles hold back the deformation of aluminum during tensile test resulting in a reduction of ductility.
Fig. 10 Effect of TiO2 particles on: a microhardness, b hardness distribution in 12 vol% AMC, c stress strain graph, d extracted values
Figure 11 reveals the morphology of the fractured specimens of AA6063/TiO2 AMCs under the tensile load. The fracture morphology of the unreinforced aluminum alloy (Fig. 11a) shows a network of well-developed dimples. The composite specimens (Fig. 11b-d) also characterized in a similar manner by uniform distribution of dimples. This observation suggests that the aluminum matrix yielded significantly before failure which is the feature of ductile mode of fracture. Numerous fractured TiO2 particles are present in Fig. 11c as seen as white dots. It indicates proper interfacial bonding of TiO2 particles with the aluminum matrix. Large empty pockets are observed in the fracture morphology of AA6063/18 vol% TiO2 AMC. These pockets probably represent the location of TiO2 particle clusters. The cluster of particles was pulled out all together beyond a particular tensile load and reduced the area for supporting the tensile stress. Hence, the UTS reduced at this higher percentage.
Fig. 11 FESEM micrographs of fracture surfaces of AA6063/TiO2 AMCs containing TiO2 particles: a 0 vol%, b 6 vol%, c 12 vol%, d 18 vol%
(1) Submicron level TiO2 particles (0, 6, 12 and 18 vol%) were effectively incorporated into the aluminum alloy AA6063 to prepare the composites.
(2) Dispersion of TiO2 particles in the aluminum matrix was reasonably uniform. Few clusters of the particles were observed at a higher volume fraction of 18%. More number of passes may be required to improve the dispersion at higher content of the particles.
(3) TiO2 particles were characterized by excellent interfacial bonding with the aluminum matrix. There was no pore or reactive layer around the particles. TiO2 particles encountered breakage due to high strain rate of the FSP process.
(4) The variation in the dispersion across the stir zone was minimum. Onion ring structure showing successive layers of high and low dispersion of particles was found at the bottom of the stir zone.
(5) There was a significant reduction in the grain size due to dynamic recrystallization and pinning effect of TiO2 particles. Sub-grain boundaries, ultra-fine grains and dense dislocation were observed by TEM micrographs. The dense dislocation was due to severe deformation and strain misfit.
(6) There was an improvement in mechanical properties due to the reinforcement of TiO2 particles in the aluminum matrix. The microhardness was tested to be 62 Hv at 0 vol% and 142 Hv at 18 vol%. The UTS was estimated to be 222 MPa at 0 vol% and 325 MPa at 12 vol%. Various mechanisms were identified based on the microstructural characterization. The tensile strength reduced at a higher particle content of 18 vol%. The elongation reduced with an increase in TiO2 particles. The fracture surfaces revealed a network of well-developed dimples which suggested ductile mode of failure.
The authors are grateful to Vigshan Tools at Coimbatore, Microscopy Lab at University of Johannesburg, FESEM lab at Coimbatore Institute of Technology, OIM and Texture Lab at Indian Institute of Technology Bombay and PSG Institute of Advanced Studies for providing the facilities to carry out this investigation.
The authors have declared that no competing interests exist.
WeChat
/
| 〈 |
|
〉 |
