Reinforced Powder Metallurgy Magnesium Alloy Composites under Oil Lubrication Condition

For the evaluation of wear behavior of Mg composites under oil lubrication conditions, powder metallurgy Mg97Y2Zn1 alloy reinforced with additive Mg2Si particles were fabricated by the repeated plastic working (RPW) and hot extrusion. The RPW process was effective in refining both Mg2Si reinforcements and α-Mg grains causing the matrix hardening. When increasing the repetition number of RPW process from 200 to 600 cycles, the particle size of Mg2Si additives changed from 8 μm to 1∼2 μm, and αMg grain size was 1 μm or less. With regard to the defensive and offensive properties of Mg alloys reinforced with Mg2Si dispersoids, the composite had superior adhesive wear resistance compared with the conventional Mg alloys because of its extremely high microhardness of 95∼180 Hv by RPW process. The uniform distribution of refined Mg2Si particles was useful for improving both defensive and offensive properties against AZ31B counter disk specimens. The Mg2Si prominent dispersoids in the matrix were also effective in forming the oil grooves around them, and caused the low and stable friction coefficient. On the other hand, in the case of the composite containing coarse Mg2Si particles, severely deep scratches were given on the counter face of the AZ31B disk, and resulted in an unstable and high friction coefficient.


Introduction
Magnesium alloys are applied to engineering components in the automotive industries due to their low density, less than 2 g/cm 3 , and effective for weight reduction, assisting the improvement of the fuel efficiency [1]. In addition, they also possess a good damping capacity and superior machinability to the other light metals such as aluminum and titanium alloys. Furthermore, Mg matrix composites show outstanding thermomechanical properties because of the hard dispersoids. On the other hand, magnesium silicide (Mg 2 Si) intermetallic compounds is one of the useful reinforcements of Mg alloys because of its low density of 1.91 g/cm 3 , high melting point of 1358 K, 120G Pa Young's modulus, high micro-Vicker's hardness of 600-700 Hv, and low coefficient of thermal expansion of 7.5 × 10 −6 K −1 [2,3]. Mg composites with in situ synthesized Mg 2 Si dispersoids were studied by a casting method and solid state reaction process [4][5][6]. In the casting process, the needle like Mg 2 Si compounds distributed in the matrix caused the reduction of mechanical properties. On the other hand, fine Mg 2 Si particles via a solid state reaction between added silicon particles and the Mg matrix were effective for improving the strength of the composites. By employing SiO 2 particles as additives, the microstructural and mechanical responses of Mg alloy composites reinforced with Mg 2 Si/MgO hybrid dispersoids were studied [7]. The reduction of SiO 2 particles by Mg was completely conducted by controlling the reaction temperature. It was set as an exothermic temperature of the DTA profile, when using the elemental mixture of Mg-SiO 2 powder. The Mg matrix composites revealed high mechanical properties and good wear resistance due to a good coherence of in situ synthesized Mg 2 Si and the matrix [8]. In the previous study, Mg alloy composites reinforced with some additives such as nano-sized alumina, SiC, and feldspar particles were used to evaluate their tribological properties and wear resistance under dry sliding conditions [9][10][11]. This was because the weight reduction of 2 Advances in Materials Science and Engineering Mg alloy composites was effective in reducing the friction loss and saving the consumed energy in operating. The conventional AZ91 alloys and pure Mg were employed as the composite materials, and then the adhesion phenomena mainly occurred on the sliding surface due to the poor heat resistance of the matrix materials. In particular, with regard to AZ91 alloy, the sliding wear map under dry conditions was investigated by changing the applied load and sliding speed [12]. It was concluded that high strength and hardness of the matrix were required to control the adhesive wear phenomena in addition to the intermetallic compounds of the alloy. Yttrium and Zinc additives to Mg alloys were also useful for improving the mechanical properties, in particular high temperature strength due to their long-periodic stacking ordered (LPSO) structures [13,14]. LPSO phases were formed in not only rapid solidification but also conventional ingot metallurgy process, and then Mg-Y-Zn cast ingots also showed much higher tensile strengths than the conventional AZ series alloys. In this study, the tribological property of powder metallurgy (P/M) Mg 97 Y 2 Zn 1 (at%) alloy reinforced with additive Mg 2 Si particles was investigated under oil lubrication conditions. In particular, the microstructures of the composites were refined by repeating the severe plastic working on the elemental mixture of Mg alloy and Mg 2 Si powders. The effects of the matrix hardness, Mg 2 Si particle size, and content of the composite on their wear behavior were experimentally evaluated. 2 Si Particles. Pure Mg powders, having a mean particle size of 112 μm and purity of 99.9%, were used as starting raw materials. Pure silicon powders with 21 μm diameter and 99.9% purity were prepared as another raw materials. The elemental mixture of Mg-33.33 at% Si powders was consolidated by a pressure of 15 MPa in vacuum (<4 MPa) by using the spark plasma sintering (SPS) equipment (SPS SYNTEX Inc., SPS-1030). The sintering conditions at 893 K for 600 seconds were sufficient for the solid state reaction between Mg and Si powders to synthesize Mg 2 Si intermetallics [15]. XRD analysis on the sintered material indicated the only Mg 2 Si peaks, and no peak of Mg and Si of raw powders was detected in the XRD profile. Fine Mg 2 Si powder with a mean particle size of 8.2 μm was obtained by ball milling (SEIWA GIKEN Co., RM-30) for 7.2 kiloseconds in air.

Powder Metallurgy Mg Alloy
Composite with Mg 2 Si via RPW Process. Machined chips of Mg 97 Y 2 Zn 1 (at%) cast ingot, containing 30 ppm Fe and 17 ppm Cu as impurities, were used as input materials. The elemental mixture of the above Mg-Y-Zn alloy powders and Mg 2 Si additives (Mg 2 Si content; 0, 5 and 10 wt%) was prepared by ball-milling process for 7.2 kiloseconds. The repeated plastic working (RPW) process, schematically illustrated in Figure 1 [16], was applied to the elemental mixture to refine Mg 2 Si particles and uniformly distribute them over the matrix of the green compact. In RPW process, the alternate operation between cold compaction and backward extrusion (an extrusion ratio; 5.2) was carried out on the powder mixture filled in the die, which was installed in a 1000 kN screw-driven press machine. The fragmentation of the additive Mg 2 Si particles occurred by the above severe plastic deformation during the RPW process. At the same time, the refined particles were embedded in the Mg alloy matrix. A lot of strains, which were effective for the dynamic recrystallization during hot extrusion, were also induced into the matrix of Mg-Y-Zn alloy powder.
The maximum repetitive number was 600 cycles in this study during RPW process. After RPW process with a suitable repetitive number, the columnar green compact was fabricated, and served to hot extrusion process. The green compact via RCP process was heated at 623 K for 300 seconds in nitrogen gas atmosphere, and immediately consolidated by hot extrusion. The extrusion ratio was 30, and temperature of the extrusion container was 673 K. The microstructure coarsening with large grains over 10 μm occurred via dynamic recrystallization by heating the green compact at 673 K. Therefore, the heating temperature of the powder compacts was 623 K in this study. Micro-Vicker's harness measurement, optical microstructure, and scanning electron microscope (SEM, JOEL JSM-6500F) observation were carried out on the extruded Mg-Y-Zn alloy composites reinforced with Mg 2 Si particles. α-Mg grain and Mg 2 Si particle size distribution of the Mg alloy composite was measured by the image analysis software (Image Pro-plus) on optical microstructure photos.

Tribological Evaluation by Pin-on-Disk Type Wear Test.
Pin-on-disk type wear test equipment (Rhesca Co. Ltd., FPR-2100) was used to evaluate the tribological properties of the above Mg alloy composites under oil lubrication conditions. Pin specimens, having 5 mm diameter and 10 mm length, were machined from the extruded Mg alloy composites. The conventional AZ31B and ADC12 (die-cast Al-12 wt% Si) alloys were used as the reference materials, and also machined to the pin specimen. AZ31B disks were employed as the counterpart materials. The surface roughness of the counter face after polishing was measured by the surface texture measuring instrument (Surfcom 1400D, TOKYO SEIMITSU CO). The R max value was 0.81∼0.93 μm. Wear tests took place in 10W30 motor oil controlled at 308∼311 K. The applied load on the pin specimen, sliding speed and test time were 100 N, 1 m/s and 10 kiloseconds, respectively. The changes in friction coefficient during wear test were evaluated. The volumetric wear loss of the pin specimens was investigated by measuring the change of the pin length. The sliding surfaces of the pin and disk specimens were observed by optical microscope to investigate wear behavior of each Mg alloy composite.  composites with Mg 2 Si particles via the RPW process with N = 200 ∼ 600 cycles.

Microstructural and Mechanical Responses of Mg Alloy Composites via RPW Process.
In both Mg 2 Si contents of 5 wt% and 10 wt%, the refinement and uniform distribution of the additives obviously occurs with increase in the repetition number of the RPW process. For example, the initial mean particle size of Mg 2 Si additives is 8.2 μm, and drastically decreases to less than 2 μm after the RPW process with N = 600 cycles as shown in Figures 2(c) and 2(f). Macroscopic photos indicate that a large plastic deformation of the matrix is observed, and some primary particle boundaries also exist in the case of 200 cycles (seeing Figures 2(a) and 2(d)). Each material reveals a layer microstructure in the matrix, which is typically caused by hot extrusion. The distance between layers gradually decreases with increasing the repetition number. Figure 3 shows optical microscope and SEM observation photos on hot extruded Mg 97 Y 2 Zn 1 alloy composites with additive Mg 2 Si particles via the RPW process with 400 cycles. Their content is (a) 5 wt% and (b) 10 wt%, respectively. The grain size of both composites is 0.7∼1 μm or less via dynamic recrystallization during hot extrusion. The optical microstructures obviously reveal Mg 2 Si spherical particles with a diameter of 1∼3 μm. In the SEM photos, the white particles correspond to refined Mg 2 Si additives. Very fine particles with 300 nm or less, which are impossible to be detected by optical microscope observation, are uniformly  distributed inside α-Mg grains of the matrix. Figure 4 shows the hardening dependence of hot extruded Mg alloy composites on the repetition number in the RPW process.
In each material, the hardness is proportion to the repetitive number because α-Mg grain refinement by dynamic recrystallization occurs with increase in the repetition number. The composite including 10 wt% Mg 2 Si is harder than that with no additive when applying RPW process with 600 cycles. In this study, the hardness was measured on the matrix of the composite, except for the Mg 2 Si particles which could be detected by optical microscope observation. The reason for this increase of the hardness is considered as below. Very fine Mg 2 Si dispersoids with 100∼300 nm in diameter, shown in SEM photos of Figure 3, effectively cause the increase of microhardness of the matrix. On the other hand, many strains are induced into α-Mg grains around Mg 2 Si hard fine particles prior to the others far from the particles by the repetition of severe plastic deformation during RPW process. They accelerate the formation of fine grains via dynamic recrystallization during hot extrusion. Therefore, the hardness of the Mg alloy via RCP process increases with an increase in the Mg 2 Si content of the composite as shown in Figure 4. Figure 5 shows the comparison of the total volumetric wear loss of the pin specimens made of Mg 97 Y 2 Zn 1 alloy composite with various contents of Mg 2 Si additives. The loss was calculated by measuring the changed height of the pin specimen after the wet sliding test. When increasing the additive content  of Mg 2 Si particles and the repetition number of the RPW process, the wear loss of pin specimens is remarkably reduced. This means that their wear resistance is obviously improved by the hardening of the matrix and the uniform distribution of fine Mg 2 Si hard particles. For example, the total wear loss of the composite with 10 wt% Mg 2 Si via 600 cycles RPW process is about 35% of that including no Advances in Materials Science and Engineering additives (0%). The discussion on damages of the sliding surface of each specimen is given in below. First of all, Energy Dispersive X-ray Spectroscopy (SEM-EDS, JEOL, JED-2300) analysis was carried out on the sliding surface of the Mg alloy composite reinforced with Mg 2 Si additives. As shown in Figure 6, Mg 2 Si dispersoids with 1∼3 μm diameter reveal round shape, not angular by the severe plastic deformation via RPW process. They are uniformly distributed in the matrix with no segregation. It also indicates no trace and no hollow of Mg 2 Si particles detached from the matrix. Figure 7 shows the comparison of sliding surfaces of the pin specimens of Mg-Y-Zn alloy with no Mg 2 Si (a) and 10 wt% Mg 2 Si additives. The repetition number of RPW process was 600 cycles. In general, the conventional Mg alloy shows the adhesive wear at the sliding surface when it directly contacts the counter face in wear test because of its soft matrix compared to the aluminum alloys and steels [9,10]. As shown in Figure 7(a), however, the sliding surface of the pin specimen reveals very slight adhesion wear, and a small change of the surface roughness is observed. This is quite different from the previous results on tribological properties of the conventional AZ91D alloy and their matrix composites [17] as mentioned above. This means that high hardness of about 140 Hv of the extruded Mg 97 Y 2 Zn 1 alloy via RPW process is enough to improve the defensive to the adhesion wear due to the plastic deformation in sliding and contacting with the counter disk specimen.

Wear Phenomenon under Wet Sliding.
Therefore, the large volumetric wear loss of the Mg-Y-Zn alloy without Mg 2 Si additives shown in Figure 5 seems to be mainly caused by the abrasive wear. On the other hand, in the case of the composite reinforced with 10 wt% Mg 2 Si particles, very slightly abrasive wear is observed on the sliding surface as shown in Figure 7(b). In particular, the magnified observation indicates that fine Mg 2 Si particles are distributed in the matrix, and no particle is detached from the surface after wear test. It is similar to the polished surface observed in the microstructure. This means that mild wear with no adhesion is formed at the sliding surface because in addition  to the matrix hardening as mentioned above, fine Mg 2 Si hard dispersoids in the matrix are also effective for the defensive to both abrasive and adhesive phenomena in sliding.
With regard to the effect of Mg 2 Si particle size on the wear behavior, Figure 8 shows the sliding surfaces of the pin specimens and AZ31B disks when employing Mg-Y-Zn alloy composite reinforced with 5 wt% Mg 2 Si particles via RPW process with (a) 200, (b) 400 and (c) 600 cycles. In the case of N = 200 cycles, coarse Mg 2 Si additives with about 20 μm are observed in the matrix, and their particle size gradually decreases with increase in the repetition number of the RPW process. This corresponds well to the microstructure changes shown in Figures 2(a)-2(c). Every pin specimen reveals slight scratches, not adhesive wear on the sliding surface. In the case of N = 200 cycles, however, severely deep scratches are observed at the counter face of AZ31B disk specimen. This means coarse Mg 2 Si hard dispersoids attacked the counter disk material in sliding. With increase in the repetition number, the scratches become slight due to the refinement of their hard particles of the pin specimens. When employing N = 600 cycles, very narrow abrasion are detected on the counter face shown in Figure 8(c). Figure 9 indicates changes in friction coefficient during the wet sliding wear test, which correspond to the damage of the pin and disk specimens as mentioned above. When using Mg 97 Y 2 Zn 1 alloy with no Mg 2 Si particle (N = 600 cycles), friction coefficient was unstable. In particular, some sudden increment of the coefficient is detected. This is caused by the stick-slip phenomena due to the repetition of adhesive and abrasive wear at the contacting surface between the pin and disk specimen (shown in Figure 7  both materials was constant and stable. As mentioned above, the sliding surfaces of both materials shown in Figures 7(b) and 8(c) revealed very slight damages with extremely narrow abrasion, and seemed as the polished surfaces. In particular, Mg 2 Si dispersoids are clearly observed as the prominent dispersoids at the sliding surface because the matrix is softer than Mg 2 Si particles and its wear occurs much early. This means that a mixed lubrication condition was stably formed, and the sliding behavior was very mild due to the formation of the oil grooves around Mg 2 Si prominent dispersoids. This wear behavior is similar to that in the case of Al-Si eutectic alloys under wet sliding conditions [18]. As a result, friction coefficient became stably lower when Mg 2 Si additives were contained in the Mg alloy. In the case of the Mg 2 Si content of 5 wt% via RPW process with 200 cycles, coarse Mg 2 Si hard particles have a severely offensive property on the counter disk surface (shown in Figure 8(a)), and the friction coefficient gradually increases due to the abrasive wear of the disk specimen. 10 wt% fine Mg 2 Si additives of the Mg composite (N = 600 cycles) cause the remarkably offensive effect on the counter material (shown in Figure 7(b)), and result in the stable but slightly high friction coefficient. The distribution of refined Mg 2 Si particles and matrix hardening of Mg 97 Y 2 Zn 1 alloy composites are suitable to keep the balanced wear resistance. That is, both of the defensive and offensive properties are given to the Mg composites when adding 5 wt% Mg 2 Si particles and applying the RPW process with 600 cycles on Mg 97 Y 2 Zn 1 alloy powder. The reference materials of AZ31B and ADC12 alloys were used as a pin type specimen to evaluate their tribological properties under the same sliding conditions in this study. In the case of AZ31B alloy, the total volumetric wear loss was about 37 mm 3 , and extremely large compared to the Mg-Y-Zn alloy composite with no Mg 2 Si particles shown in Figure 5. The friction coefficient gradually increased from 0.03 to 0.08, and was unstable. The microhardness (Hv) of the matrix of AZ31B was about 46, and quite smaller than that of Mg-Y-Zn alloys via RCP process (microhardness; 97). It was concluded that the poor wear resistance of AZ31B alloys was mainly caused by the extremely low hardness. The large wear loss was also due to no additive of hard particles, such as Mg 2 Si and Si, contained in the conventional AZ31B alloy. On the other hand, when ADC12 alloy was used as a pin specimen, the wear loss was about 14 mm 3 , and its friction coefficient was 0.03∼0.04, and very stable. This means coarse Si particles are effective to improve wear resistance, and also have an important role to make a stable sliding condition by oil grooves around Si dispersoids as mentioned in the previous reports [18]. Figure 10 shows the dependence of volumetric wear loss on microhardness of pin specimens. This indicates that wear resistance strongly depends on the matrix hardness. That is, the abrasion is a major wear phenomenon in the case of Mg 97 Y 2 Zn 1 alloys via RPW process. Furthermore, Mg 2 Si fine hard dispersoids are also very effective to improve the wear resistance and prevent the abrasion. The above results are quite different from wear behavior in the case of the conventional Mg alloys resulting in the adhesive wear [10,19,20]. This is because Mg 97 Y 2 Zn 1 alloys have superior mechanical properties to the conventional ones, and prevent the plastic deformation, causing at the sliding surface. It is concluded that the addition of Mg 2 Si particles is also effective for improving the abrasive wear resistance. It is however, important to handle their suitable particle size and content of the composite by microstructure control.

Conclusion
The wear behavior of powder metallurgy Mg 97 Y 2 Zn 1 alloy composites reinforced with additive Mg 2 Si particles via RPW process was investigated under oil lubrication conditions. When increasing the repetition number of RPW process, 8 Advances in Materials Science and Engineering the severe plastic deformation by RPW process was effective for the uniform distribution of refined Mg 2 Si prominent reinforcements with 1∼3 μm. It also accelerated the matrix hardening (microhardness, Hv = 95∼180) by α-Mg grain refinement of the composite. The former was useful for improving the wear resistance and offensive property against AZ31B counter disk specimens. It also caused a low and stable friction coefficient less than 0.03 by the oil groove formation around the Mg 2 Si particles, making mixed lubrication conditions. The hardening of the matrix was useful to prevent the adhesion wear phenomenon when contacting the AZ31B counter face. When no Mg 2 Si hard particle was dispersed in the matrix of the Mg alloys, the adhesive was the main wear mechanism. On the other hand, when coarse Mg 2 Si additives were contained in the composite, they severely attacked the counter face with deep scratches of AZ31B disk, and the abrasive was the main wear mechanism.