Influence of Microstructure on Tribological Behaviors of Al6061 Metal Matrix Composite Reinforced with Silicon Nitride (Si3N4) and Silicon Carbide (SiC) Micro Particles

Dual step stir casting method was utilized to develop Al6061 composite reinforced with SiC (micron) and Si3N4 (submicron) particles in varying proportion and weight percentage. A homogeneous and uniform distribution of hybrid reinforcement without significant porosity was observed in the microstructure of the composites. X-ray diffraction (XRD) results manifest that Si3N4 and SiC particles are thermodynamically stable during the processing and no unwanted phases were detected. Elemental mapping was also performed for phase identification. With reference to the base alloy significant improvement was noticed in physical and tribological properties of hybrid composites. Maximum rise in hardness was 54.64%. Abrasive wear test results from pin on disc reveals that wear resistance get enhanced for all composition and load is found to be most dominating factor affecting wear behavior.


Introduction
Material scientists across the world have focused to develop new and better metal matrix composites. Over the past researches it has revealed that, addition of more than one reinforcement particles could improve the overall performance of the composite including mechanical and microstructure properties [1][2][3][4][5][6]. Hybrid composite is second generation of composite materials in which two or more reinforcements are used to reinforce the matrix material. Multiple reinforcement incorporation is intended to revamp the performance of the composite in terms of property enhancement, cost reduction, and mutual compensation for the negative impact of reinforcements. Al6061 as matrix material is becoming increasingly popular because of its excellent corrosion resistance, appreciable formability, and moderate strength in addition to low cost and ease to process characteristics. Moreover, Al6061 alloy can be heat treated, so further property improvement is absolutely possible. Three types of reinforcements are primarily used for composite manufacturing viz. particulates, continuous fibers and discontinuous fibers/whiskers. Among them, particulate matrix composites are easy to process and give isotropic characteristics that is why frequently developed in most of research work. Ceramic materials (oxides, carbides and nitrides) are frequently used as reinforcements since they have excellent combination of specific strengths and stiffness at both ambient and elevated temperatures. Silicon nitride (Si 3 N 4 ) & silicon carbide (SiC) are leading ceramic materials for high thermal applications. Both have excellent combination of properties e.g. low density, high hardness, and adequate corrosion and wear resistance [7][8][9]. Silicon nitride possess high temperature strength and is chemically stable at elevated temperature. It is manifested from the past research article that elongated silicon nitride particles improve impact strength/toughness [10,11]. Solid and liquid processing techniques are some most common ways to develop particulate matrix composites. 'Stir casting' is a kind of liquid processing technique that is frequently used due to its simplicity, flexibility, high material yield, incurring low cost, suitability for mass production and satisfactory results [12,13]. On the contrary, conventional stir casting process has significant faults such as poor wettability between constituent phases, non-homogeneous particle distribution, and formation of weak intermetallic compounds [14]. These faults/drawbacks can be limited by proper selection of stirring process parameters and using some aids [15]. From different literatures, preheating of reinforcing particles, addition of alloying element (Mg), metallic coating on particles, performing stirring in steps and use of ultrasonic assisted stirring are concluded as some of the useful techniques that can be adopted for getting better results [16][17][18]. In present research work, three techniques including dual step stirring, preheating and use of alloying element have been adopted to get anticipated results.

Materials and Composite Fabrication
Chemical composition of matrix metal (Al6061) has been enlisted in Table 1. Si 3 N 4 and SiC particles (99.9% purity) as reinforcements were sieved using standard sieving practice to get particles within the size range of 0.5-1 µm for Si 3 N 4 and 5-15 µm for SiC particles. FESEM images (Field emission scanning electron microscope) in Fig. 1 is showing the morphology of the particles. Electric resistance furnace of temperature capacity 1500℃ coupled with temperature controller and graphite crucible was employed to melt the alloy. After achieving a uniform temperature of the melt at 750℃, coverall 11 (1wt%) was used to cover the melt in order to avoid the gas absorption and impurities. Then the furnace is set to a lower temperature at 625℃ to cool the melt and finally reach to its semi solid state. Afterwards, small quantity of magnesium (approx. 0.7wt% or 10 g) in the form of chips and hexachloroethane tablets were incorporated into the melt before the addition of ceramic particles to improve the wettability and for degassing purpose respectively. Subsequently, floated impurities and slag were scooped out and manual stirring was performed to form the proper vortex and then preheated (at 550℃ for 2 h) reinforcement particles mixture was inducted into semi solid melt. Preheating of particles is done to eliminate moisture content and contaminants in addition to improve wettability by forming oxide layer on particle surfaces. Manual mixing of the slurry was done about 10 min, once it was completed reheating of composite slurry was started until it reaches to 750℃ ± 30℃. A mechanical stirring with three bladed stirrer (blade angle 45 o) was then performed at a speed of 550 rpm for 10 min. Thereafter molten melt was poured into preheated die (approx. 550 °C, to avoid porosity due to solidification shrinkage). Die was in the form of fingers and made of cast iron and then allowed to solidify. Finally the casts was permitted to cool and then machined to prepare different testing samples as per the specifications of ASTM standards. Figure 2a-e depicts the experimental setup, various casting stages and machined testing samples. The details of employed compositions has been tabulated in Table 2.

Heat Treatment
Heat treatment is a common practice adopted for betterment of the different properties. Programmable muffle furnace was used to perform heat treatment operation. Heat treatment of the investigated materials was involved three steps, those were:(i) Solutionizing: solutionizing at 543℃ for two hours in furnace with T6 temper for

Microstructural Analysis
X-ray diffraction (XRD) was employed to analyze the crystallinity and elemental composition of the material. X-ray source of Cu Kα radiation (λ = 1.5406 Ȧ) was used for the analysis of SiC, Si 3 N 4 , AA6061 aluminum alloy and its hybrid composite (powder form). The standard database (JCPDS database) for XRD pattern was used to recognize the crystalline phases. Microstructure was examined using Carl-Zeiss Axio Scope 5 Optical Microscope and Carl Zeiss make Ultra Plus FESEM to know particle distribution, grain boundaries and interfacial bonding [19][20][21][22]. Samples of 20 mm diameter and 15 mm thick were cut, cold mounted and polished as per standard metallographic technique and etched by Keller's reagent (3 ml HCl, 2 ml HF, 5 ml HNO 3 , 190 ml H 2 O). The required surface finish of the specimen was achieved by grinding it with 240, 400, 600 and 1000 grit papers. The subtle scratches caused by the final grinding operations were removed by polishing with flaky powders (as polishing abrasives) of different grades were used. A combination of powder and water was distributed over a rotating disc wrapped with polishing velvet cloth. Alumina powders (magnesium or diamond may also be used) of different fineness was used for successive polishing phases. The specimen carefully washed after each phase to prevent contamination. Elemental mapping to confirm the presence of elements was performed using the FESEM microscope.

Hardness
For hardness measurement Vickers microhardness test was performed following ASTM: E384-08 standard. Pyramidal shaped with square base diamond indenter, 1000gf test load and 15 s compression time was taken to produce the indent. An average of five readings (at different locations) is taken to get final hardness value of each sample.

Wear Tests
Using a pin-on-disc wear testing machine (Model: TE-165 Magnum) as shown in Fig. 3, two body abrasive wear test was performed on a cylindrical shaped specimen called a pin of 25 mm length and 10 mm in diameter. ASTM G132-96 standard was followed to perform the test. Normal load of 5-15 N (with a 5 N increment), sliding distances of 125 m, 250 m, abrasive wear paper of 180 and 320 grit (self-adhesive SiC abrasive paper),  Al6061/(1.0%Si 3 N 4 + 1.0%SiC) 1600 16 16 and constant sliding speed of 1.25 m/s at ambient temperature are the input parameters for wear tests. Specimen surfaces were ground with 80 SiC emery paper to remove rough surfaces before performing abrasive wear tests. The self-adhesive abrasive wear paper was firmly placed on the testing machine disc (EN31 hardened steel disc). Specimen was held by specimen holder against the

Fig. 4
Vickers microhardness test result for as cast and heat treated samples rotating abrasive media and load was applied to the specimen using a cantilever mechanism and dead weights. The traversal distance, track radius, and rotational speed of the disc were all adjusted to meet our requirements. Before and after testing, acetone-clean specimens were weighed using a digital weighing machine (with an accuracy of 0.01 mg). A load cell/force transducer attached to digital display was used to measure frictional force. The friction force measured is then converted into a friction coefficient. The weight loss method was used to calculate the volume loss and volumetric wear rate, as shown in Eqs. (1) and (2), while specific wear rate and coefficient of friction were calculated using the Eqs. (3) and (4) SiC (110) SiC (116)  Si3N4(321) Al (222) Al (311) Al (220) Al (200) SiC (102) Si3N4 (210) Intensity (a.u.) Si3N4 (101) SiC (101) Al (111)   where ρ is the density of sample (g/cm 3 ), S = sliding distance (m), Δm = mass loss (g) where F t is the tangential friction force and F n is the applied normal load in 'N'.

Density, Porosity and Hardness
Experimental density (actual density) based on Archimedes principle and theoretical density based on rule of mixture are measured and tabulated in Table 3. Difference of these densities is utilized to measure porosity percentage. A Precise and high accuracy (0.01 mg) electronic balance was used to weigh the samples. Slight increment in density can be observed in composite samples that may be due to presence of relatively dense ceramic particles. Low porosity level indicates to the success of processing technique. Microhardness test reveals a negligible difference in readings at different locations. It indicates uniform distribution of ceramic particles. Hardness improvement with hybrid composition can be easily observed in Fig. 4, it may also be noticed that hardness improvement is directly proportional to wt% of hybrid composition and it is more favorable to high proportion of Si 3 N 4 particles. Strong particle bonding, grain refinement and dislocations induced by virtue of the presence of hard particles and mismatch in elastic modulus and coefficient of thermal expansion (CTE) are some of the reasons of improved hardness [23,24]. Maximum rise has gained for sample S4 which is 55%. Precipitation of intermetallic phases and hardening region due to segregation of solute elements of matrix alloy have improved the hardness upon heat treatment and the gain is approximately 45-60%. Highest hardness for both as cast and heat treated samples was observed for sample S4 i.e. Al6061/1.0%Si 3 N 4 /1.0%SiC composite. Entangled morphology of Silicon nitride and its distribution with Silicon carbide particles in same proportion are the governing cause behind higher hardness of sample S4 compared to S3.

X-Ray Diffraction (XRD Analysis)
The X-ray diffraction (XRD) patterns as in Fig. 5 indicate peaks, corresponding 2θ angle (Bragg angle) available phases and miller indices for reinforcing particles, In XRD patterns for prepared Al6061 alloy (S0) and Al6061/wt%Si 3 N 4 /wt%SiC hybrid composites, the peaks correspond to Al, Al 12 Mg 17 , Mg, Mg 2 Si, Si, Si 3 N 4 , and SiC and no unwanted phases are identified. XRD pattern of fabricated Al6061 is correlated with the Aluminum (JCPDS Card No. 01-085-1327). The low intensity for the peaks of reinforcement particles in composite represents its low weight percentage. The XRD results manifest that Si 3 N 4 and SiC particles are thermodynamically stable during the processing.

Elemental Mapping
The micrographic image of Al6061 and its respective elemental mapped image has shown in Fig. 6, it exhibits the presence of Al, Mg, Si and Fe elements which authenticates  Fig. 7, it shows the presence of Si, N and C with the Al and Mg and subsequently verifies that Al6061 with SiC and Si 3 N 4 particle were utilized to form the composite for current investigation. Elemental mapped image of Al6061/(Si 3 N 4 + SiC) hybrid composite shows the presence of 'C' by red dots, 'Al' by green dots, 'Fe' by blue dots, 'N' by yellow dots, 'Mg' by dark turquoise color dots and 'Si' by purple dots.

Optical and FESEM Microstructural Examination
Microstructural studies are quite useful for property prediction of end material. Figures 8 and 9 displays the Optical Microphotographs (OM) at 100 × magnification of Al6061 alloy and Al6061/ (Si 3 N 4 + SiC) hybrid composite for as cast and heat treated samples respectively. Microstructural phases includes α-Al with small amount of primary Mg, Si and intermetallic compounds (Mg 2 Si) scattered in the form of strips (needle shaped) or patches, eutectic Mg 2 Si and reinforcement particle phases. Microstructure of base alloy reveals a fine distribution of small and coarse grains of alpha aluminum (α-Al) solid solution Fig. 9 Optical micrographs of heat treated samples a Al6061 b Al6061/(0.5%Si 3 N 4 + 0.5%SiC) c Al6061/ (0.25%Si 3 N 4 + 0.75%SiC) d Al6061/(0.5%Si 3 N 4 + 1.5%SiC) e Al6061/ (1.0%Si 3 N 4 + 1.0%SiC) hybrid composite 1 3 with dendritic patterns of grains. Segregation of solute elements (Si, Mg) in interdendritic regions can also be also noticed. Reinforcement particles behaves as a nucleation sites for grain formation, they act as an obstacle for grain growth which results in grain refinement. It can be clearly seen that the grains of the matrix alloy are getting refined with the addition of reinforcing particles which results in increasing the grain boundary concentration and finally turns into increase in strength and wear resistance. Grain boundaries obstruct the dislocation movement so high concentration of grain boundaries offers more obstacles. From optical micrograph, one can easily infer that there is no significant clustering and porosity has found around the particles. Coarse intermetallic compound (Mg 2 Si) and eutectic phase (α-Al)-(Mg 2 Si) also observed in the composite micrograph. Few literatures manifests, addition of SiC and Si 3 N 4 particles accelerate the aging and quench sensitivity of the Al6061 matrix alloy due to the high dislocation densities produced as a result of high thermal mismatch between reinforcement and matrix metal. Optical micrographs of heat treated samples as in Fig. 9 depicts the dendritic structure is breaking down and became more uniform with equiaxed grain structure, in which intermetallic precipitates distributed both at the grain boundary and within the grains. The precipitates near the grain boundaries are much finer. Figure 10, exhibits the FESEM micrographs of Al6061 and Al6061/ (Si 3 N 4 + SiC) composites. These images confirms the grain refinement and particle clustering of Si 3 N 4 and SiC particles. Al6061 micrograph consists of dendritic α-Al with segregation of primary Si and mg precipitates and stripped Mg 2 Si intermetallic phase [25,26]. Microstructure of composites displays distribution of two reinforcing particles, somewhere particle clustering and moderate interfacial reaction. However, no significant porosity has been noticed. Sometimes, moderate interfacial reaction is required to bind the reinforcing particle with the matrix which results in enhanced interfacial bonding. Consequently, mechanical properties get improved. Clustering region revealed as cementations, which bind the whole nearby regions. In clustered regions SiC particle are surrounded by Si 3 N 4 particles. Whereas in some regions Sic particle are surrounded by elongated rod-like Si 3 N 4 particles which jammed the SiC particles  [26]. Few literature revealed that there is an increase in the porosity with increase in the SiC particles which indicates that the interfacial adhesion is weak between the particles and matrix, it will promote to lower the mechanical properties of the composite with its higher proportion.

Abrasive Wear Test
It is well understood that a material's wear behavior is a system dependent property rather than a material property, so in this way material's abrasion wear resistance is determined by its metallurgical properties, operating and environmental conditions, and the nature of the counter surface. Among these factors, the hardness of the pin material, abrasive grit size on paper, and load play an important role in determining material wear response in terms of wear mechanism [27]. Calculation results of volume wear loss (∆V), volumetric wear rate (W r ) and specific wear rate (W s ) are tabulated in Table 4 whereas Figs. 11 and 12 depicts the responses of Al6061 alloy and its hybrid composite Al6061/Si 3 N 4 /SiC in terms of specific wear rate (mm 3 /Nm) for four different wear conditions (180 and 320 grit size, 125 m and 250 m sliding distance, and normal load (5-15 N) in a step of 5N at constant sliding speed, 1.25 m/s). Specific wear rate is found to be decreased with increasing grit size and sliding distance (especially in the case of hybrid composition). The results also reveal that hybrid composites have a lower specific wear rate than base alloys, which attributed to the improved hardness due to presence of hard ceramic particles. Particle strengthening and increased hardness provide greater resistance to material removal [28]; additionally, it has been discovered that as hardness increases, the actual (real) area of contact decreases (actual area of contact is the ratio of applied load to pin material hardness) which results in lowering wear rate for all compositions of hybrid composites according to hardness value [29]. While effect of load is considered, specific wear rate decreases as load increased for all four operating conditions (grit size and sliding distance). Coefficient of friction (COF) is one of the important parameters, which is also taken into consideration for the present investigation. Figure 13 depicts the variation of COF for different operating conditions. It is found from result that the coefficient of friction gets decreased with the addition of reinforcement particle and increases with increase in load for almost all operating conditions and all compositions. In view of abrasive wear behavior performance, it is concluded that S4 hybrid composition outperforms than all remaining compositions and load is found to be most dominating factor affecting wear behavior.

Conclusions
Influence of Microstructure on Tribological Behaviors of Al6061 Metal Matrix composite reinforced with Silicon Nitride (Si 3 N 4 ) and Silicon Carbide (SiC) micro particles as the above experimental work following results have been obtained: Fig. 11 A field emission scanning electron micrograph of as cast Al6061/(Si 3 N 4 + SiC) hybrid composite showing distribution of elongated rod-like Si 3 N 4 in Al6061 alloy matrix • Density increases with the induction of reinforcing material, obviously due to presence of dense reinforcement particles. Effectiveness of processing method is got to know as porosity level has found within the acceptable range. • Hardness of hybrid composite for all composition rises significantly as rise in reinforcement content. While comparing the hardness with base alloy, it has been increased by 35.6%, 32.1%, 41% and 55% for composite sample S1, S2, S3 and S4 respectively. It means that the highest hardness gain is achieved with the S4 composition. Grain refinement, high dislocation densities as a result of the existence of hard ceramic particles and CTE mismatch along with good interface bonding, are few cause of improved hardness. • Highest hardness (for both as cast and heat treated) corresponds to sample S4 i.e. Al6061/1.0%Si 3 N 4 /1.0%SiC composite, while in this composition harder reinforcement (SiC) is present in fewer quantity compared to sample S1 i.e. Al6061/0.5%Si 3 N 4 /1.5%SiC. Morphology of Silicon nitride and its distribution with Silicon carbide particles in same proportion is the ruling cause behind higher hardness of S2 compared to S1. An equal propor- Fig. 12 Variation of specific wear rate tion of reinforcements leads to better hardness because, in this case, Si 3 N 4 properly surrounds the SiC particle and provides a better barrier for dislocation movement as well as produces more lattice disturbance around the particles. • The XRD pattern manifest that Si 3 N 4 and SiC particles are thermodynamically stable during the processing. • After study the results of specific wear rate it is found that specific wear rate generally decreases with the increase of the grit size and sliding distance. The result also indicates that the hybrid composite of any composition gives low specific wear rate compared to base alloy might be due to presence of hard ceramic particles. Particle strengthening and improved hardness offer higher resistance to material removal. For all four combinations of grit size and sliding distance, specific wear rate decreases as load increases.