Enhanced Mechanical and Anti-Wear Properties of Magnesium by Alloying and Subsequent Extrusion–Aging Treatment

Abstract

In this work, the microstructure, mechanical, and anti-wear properties of the alloyed-extruded-aged Mg-8.3Gd-4.5Y-1.4Zn-0.3Zr (wt%) alloys were investigated by X-ray diffraction (XRD), scanning electron microscope (SEM), nanoindentation, and wear tests. Results showed that the alloying—extrusion processing induced a significant grain refinement of magnesium resulting in the formation of bulk Mg₂₄(GdYZn)₅ at the grain boundaries. The grain size decreased from 116 μm in pure magnesium to 17 μm in alloyed-extruded magnesium, while the grain refinement, solid solution and second phase strengthening led to a hardness enhancement from 0.67 GPa in pure magnesium to 1.64 GPa in alloyed-extruded magnesium. Aging treatment further drove the structural homogenization of the alloyed-extruded magnesium resulting in an enhanced hardness of 1.83 GPa. During the sliding wear tests, a large-area plastic deformation layer formed on the wear track surface of pure magnesium, leading to an unstable friction coefficient and a high wear rate of 2.64 × 10⁻³ mm³·N⁻¹·m⁻¹. The alloying—extrusion—aging treatments effectively inhibited the formation of the plastic deformation layer. The wear rate of the alloyed-extruded material decreased to 1.60 × 10⁻³ mm³·N⁻¹·m⁻¹. In contrast, the alloyed-extruded-aged material showed a lower wear rate of 1.16 × 10⁻³ mm³·N⁻¹·m⁻¹. The wear failure mechanisms of all fabricated materials were further discussed according to the characterization results.

1. Introduction

Magnesium alloys are widely used in lightweight components in the automotive and aerospace fields because of their low density, high specific strength, excellent machinability, and recyclability [1,2,3]. Since the 1920s, rare earth elements have been used as alloying elements to improve the mechanical properties of magnesium. Among them, heavy rare earth elements, such as Gd and Y, show great potential in improving the comprehensive properties of magnesium [4,5]. The Gd and Y atoms with larger atomic radii are usually enriched at grain boundaries to form a continuous network of eutectic phases, which is the main way to strengthen the Mg-RE alloy system [6]. Moreover, the Zn and Zr elements are usually added to the Mg-Gd-Y system alloys. The addition of Zn to Mg-RE (RE = Gd or Y)-(Zr) alloys produces the formation of a novel long-period stacking ordered (LPSO) structure, which plays a key role in enhancing the mechanical property of magnesium alloy [7,8,9,10]. The Zr element doping can significantly refine the grain size of magnesium alloys due to the uniform distribution of Zr, and further promotes the heterogeneous nucleation process of the magnesium matrix [11]. Therefore, Mg-Gd-Y-Zn-Zr alloys have become a research hot spot to expand the application range of magnesium alloys.

In past years, many studies have confirmed that Mg-Gd-Y-Zn-Zr alloys have better mechanical, anti-wear and anti-corrosion properties in comparison to traditional Mg alloys. For example, Chang et al. [12] prepared the thin-walled Mg-11Gd-3Y-1Zn-0.2Zr (wt%) alloy by rheological die casting. They found that the ultimate tensile strength and elongation of this alloy could reach 250 ± 10 MPa and 6.1 ± 0.8%, respectively. Zhao et al. [13] tested the fracture toughness of the Mg-5.5Gd-4.4Y-1.1Zn-0.5Zr (wt%) alloy. They found that the fracture toughness of the hot-compressed and the annealed sample was as high as 17.82 MPa·m^1/2. Chen et al. [1] found that the decisive factors determining good mechanical properties of AZ91D magnesium alloy were grain refinement and uniform distribution of intergranular compounds. Hu et al. [14] concluded that the Mg₁₂(GdY)Zn phase in Mg-10Gd-4Y-1.5Zn-0.4Zr (wt%) alloy played a key role in reducing the wear rate. Liu et al. [15] found that the LPSO phase had low relative potential and low volume fraction, which significantly reduced the microgalvanic corrosion acceleration rate of Mg-15Gd-2Zn-0.39Zr (wt%) alloy.

Recently, the hot deformation process and aging treatment are widely developed to refine the microstructure of cast magnesium alloys due to their poor formability [16,17,18]. It has been reported that the growth textures of Mg-RE alloys were tuned by dynamic recrystallization, intermetallic compound dissolution/fragmentation, and secondary phase redistribution through the hot extrusion technique [19,20,21,22]. Subsequently, the ageing treatment proved to be effective in tuning the microstructure and optimizing the mechanical properties of the as-extruded Mg-RE alloys [8]. Up to now, the structural evolution and mechanical behaviors of the extruded-aged Mg-RE alloys are the research focuses in the forming field of magnesium alloy. In contrast, the anti-wear property is also a key performance index to determine the industrial application of these alloys. Unfortunately, relatively little work about the tribological behaviors of these extruded—aged Mg-RE alloys are carried out. Here, this study employed detailed sliding wear tests to characterize the tribological behaviors of pure magnesium, extruded, and extruded-aged magnesium alloys. The purpose of this work is to clarify the strengthening and anti-wear mechanism of these extruded-aged Mg alloys.

2. Materials and Methods

Smelting casting, hot extrusion molding, and ageing treatment technology were employed to prepare the alloyed-extruded-aged Mg-8.3Gd-4.5Y-1.4Zn-0.3Zr (wt%) alloy. Alloying was achieved by adding gadolinium (China Magnesium Technology Co., Ltd., Shanghai, China), yttrium (China Magnesium Technology Co., Ltd., Shanghai, China), zinc (China Magnesium Technology Co., Ltd., Shanghai, China), and zirconium (China Magnesium Technology Co., Ltd., Shanghai, China) elements to the pure magnesium matrix (China Magnesium Technology Co., Ltd., Shanghai, China). The specific process details are as follows: high-purity magnesium and zinc were first added to the intermediate frequency induction furnace, and then the furnace temperature was raised to 800 °C. Subsequently, Mg-30Gd and Mg-30Y alloys (China Magnesium Technology Co., Ltd., Shanghai, China) were added to the intermediate frequency induction furnace, and Mg-30Zr alloys were added after these above alloys were completely dissolved. Before the casting, the completely melted and evenly mixed molten alloy was allowed to stand for 60 min. After cooling and descaling, a bar with a diameter of 150 mm was finally obtained. The bar was kept in a protective atmosphere (Ar) of 510 °C for 16 h, and then was extruded into a bar with a diameter of 45 mm. The extrusion temperature and extrusion ratio were 420 °C and 1:1, respectively. The obtained bar was aged at 200 °C for 28 h, and was processed into a Φ 45 mm × 5 mm cylinder by a wire cutting. Finally, these samples were degreased, ground, and polished for testing. For comparison, both pure magnesium and extruded Mg-RE alloy (Φ 45 mm × 5 mm) were also prepared for this investigation. The material name and number of all samples are shown in

Table 1. Materials name and number.

Sample	Name
M0	Mg
M1	Extruded Mg-RE
M2	Extruded-aged Mg-RE

.

The phase compositions of all prepared materials were investigated by a X-ray diffraction (XRD, X’ Pert Powder, PANalytical B.V., Almelo, The Netherlands). The scanning range was in the range of 20–90°, and the scanning rate was 10°/min. The Cu Kα ray with a length of 0.154156 nm was selected during the XRD characterization. The average grain size of all prepared samples was calculated by the linear intercept method [16]. An optical microscope (VHX-500F, Keyence, Osaka, Japan) and a field emission scanning electron microscope (SEM, EVO MA 10, Carl Zeiss AG, Jena, Germany) were employed to observe the surface morphologies of all prepared materials. The local chemical compositions of the materials were analysed and quantified using energy dispersive spectrometer (EDS, XFlash6l100, Bruker, Karlsruhe, Germany). The operating parameters of SEM were as follows: images recorded using a backscatter detector, incident voltage of 15 KeV and spot diameter of 60 μm. The nanomechanical properties of all prepared samples were measured by a nanoindenter (Nano Indenter^® G200, Kla Corporation, Milpitas, CA, USA). The nanoindenter calculated all samples by continuous stiffness measurement using a Bose indenter. The nano-hardness and Young’s modulus of these materials were calculated according to the Oliver–Pharr method. Among them, the nano-hardness and Young’s modulus of all samples were calculated according to the increase of the test depth (2 μm), respectively. The tribological properties of all prepared materials were evaluated by using a ball—disk tribometer (MS-T3000, Lanzhou Huahui Instrument Technology Co., Ltd., Lanzhou, China). These sliding wear test parameters were 5 mm test radius, 300 rpm sliding speed, and sliding time of 30 min. A 0.5 N load was applied to the samples via Si₃N₄ ball (Φ = 6 mm). The wear test was repeated at least three times to ensure the repeatability of experimental results. A surface profiler (ALPHASTEP D-100, KLA Tencor, Milpitas, CA, USA) was employed to read the wear track profile. The optical microscope and SEM were used to characterize the wear track morphologies of all materials after the sliding wear tests. In order to better understand the friction damage mechanism, this cross-sectional specimen was first processed by a wire cutting, and only the subsurface layer of the wear track was retained. Subsequently, the specimen was rapidly broken apart, and the fracture morphology was further characterized by the SEM.

3. Results and Discussion

The XRD patterns of all fabricated materials are shown in Figure 1. It can be found that these main diffraction peaks of material M0 correspond to the typical α-Mg matrix phases [16]. After doping with Zn, Zr and rare earth elements, these α-Mg matrix phases are also characterized in material M1. In addition, some visible Mg₂₄(GdYZn)₅ and Mg₁₂(GdY)Zn phases also appear in this material. Previous researchers have reported that the precipitation sequence of Mg-RE-Zn-Zr alloy aged at 200 °C is (SSSS)→${\beta }^{″}$→${\beta }^{\prime }$ [8]. However, the content of the ${\beta }^{\prime }$ phase in material M2 is too small to be detected by the XRD. Therefore, material M2 is mainly composed of Mg₂₄(GdYZn)₅, Mg₁₂(GdY)Zn, and α-Mg matrix phases.

Figure 2 shows the metallographic images and grain sizes of all fabricated materials. Material M0 exhibits a coarsely equiaxed grain morphology with a grain size of about 116 μm, which is following the same grain type as the recent study [23]. In contrast, material M1 undergoes prominent recrystallization and shows an equiaxed grain morphology with a grain size of approximately 17 μm [24]. These results are under previous reports that thermo-mechanical plastic deformation and second-phase nucleation induced a significant grain refinement in magnesium alloys [25,26,27,28]. However, these grains in material M1 are disorderly piled together. After the aging treatment, the precipitation hardening rebalances the microstructure of material M2. Moreover, this process further enabled material M2 to obtain an uniform and rounded grain morphology with a smaller grain size of 15 μm.

Figure 3a shows the SEM surface image of material M1. Five growth textures are characterized in the SEM image including a dark gray Mg matrix, white bulk, white particle, gray bulk, and gray lamellar phases. Large-size second phase is distributed in the form of a network at the boundaries of the grains, and the white particle phase is scattered inside the grains and the second phase. Previous literature [8,29] reported that white particles usually formed at grain boundaries or in the α-Mg matrix of the Mg-RE-Zn-Zr system, and these white particle phases are mainly attributed to the RE, Zn, and Zr-rich intermetallics. As shown in Figure 3A, the chemical compositions of the gray phase are 87.45 at.% Mg, 3.06 at.% Gd, 4.37 at.% Y, and 5.12 at.% Zn, indicating that the gray phase corresponds to the Mg₁₂(GdY)Zn as determined by the XRD characterization. According to the XRD analysis of Figure 1 and the EDS energy spectrum of Figure 3B, the bulky white phase is attributed to the Mg₂₄(GdYZn)₅. As shown in Figure 3a, numerous visible whisker growth morphology of Mg₁₂(GdY)Zn also appear in the grain [30,31]. In addition, plenty of micro-cracks appear at the interface of the eutectic phase and α-Mg matrix, which likely originated from the non-equilibrium solidification process of the Mg₂₄(GdYZn)₅ and Mg₁₂(GdY)Zn phases. The SEM surface image of material M2 is displayed in Figure 3b, after the aging treatment, these dark gray Mg matrix, white bulk, white particle and gray bulk also appear in material M2. According to the EDS spectra of Figure 3C,D, these white and gray bulks correspond to the Mg₂₄(GdYZn)₅ and Mg₁₂(GdY)Zn, respectively. However, these whisker-like Mg₁₂(GdY)Zn phases dissolve into the interior of the α-Mg matrix. Meanwhile, these micro-cracks hardly appear in material M2. Apparently, this aging treatment drives the composition homogenization of the growth texture and further promotes the fusion of micro-cracks between the matrix and the eutectic phase. According to the previous literature [9,32], the precipitation process of the aged M2 material likely obeys to this evolution of (SSSS) →${\beta }^{″}$→${\beta }^{\prime }$.

Figure 4 presents the nanoindentation characterization results of all materials. As shown in Figure 4a, the hardness and modulus of material M0 are 0.67 GPa and 43.9 GPa, respectively. Material M1 exhibits a hardness of 1.64 GPa and a modulus of 55.2 GPa. The hardness and modulus of material M2 are 1.83 GPa and 55.8 GPa, respectively. The H/E ratio of all materials is displayed in Figure 4b. It can be found that material M0 possesses a low H/E ratio value of about 0.015, while the H/E ratio increases to 0.03 for material M1, and reaches a higher value of 0.033 for material M2. Apparently, material M1 shows better mechanical properties than material M0, which is mainly attributed to the synergistic strengthening effect of grain refinement, solid solution and second phase. As identified by the metallographic images of Figure 2 and SEM results in Figure 3, the grain size of material M1 decreases significantly, while these eutectic compounds also form at its grain boundaries, both of which effectively inhibit the basal slip [33,34] resulting in an enhanced mechanical property. In addition, the solid solution of doping atoms into the α-Mg matrix is also a key factor in enhancing the mechanical property of material M1. Compared with material M1, this ageing technique drives the structural homogenization of the α-Mg matrix resulting in a prominent solid solution strengthening. Therefore, this strong synergistic strengthening effect of grain refinement, solid solution, and second phase eventually leads to a slight enhancement in the hardness and modulus of material M2 [35]. According to the previous literature [9,36], the formation of ${\beta }^{\prime }$ phase is also expected to benefit the mechanical property improvement of material M2.

Figure 5 shows the friction curves of all fabricated materials. Material M0 exhibits an unstable tribological performance during the whole sliding wear test, and its friction coefficient fluctuates between 0.35 and 0.60. In comparison, materials M1 and M2 show relatively stable tribological performances during the whole sliding wear tests. The friction coefficient of material M1 slowly climbs from 0.20 to 0.35 with the increase of the sliding distance, while material M2 has a similar evolution law in friction coefficient to material M1, and its friction coefficient gradually increases from 0.20 to 0.3.

The wear track profiles of all fabricated materials are displayed in Figure 6. A large amplitude fluctuation is observed in the wear profile of material M0, indicating an unstable tribological performance. This phenomenon is in accordance with the evolution of the friction curve. The wear track depth and width of material M0 are 31.17 μm and 895.61 μm, respectively, and its wear rate is about 2.64 × 10⁻³ mm³·N⁻¹·m⁻¹. In contrast, the fluctuation in the wear profile of material M1 obviously decreases. Meanwhile, material M1 exhibits a shallower wear track depth of 20.50 μm and a narrower wear track width of 742.35 μm. The wear rate of material M1 decreases to 1.60 × 10⁻³ mm³·N⁻¹·m⁻¹. Material M2 possesses the best wear resistance accompanied by a smaller profile fluctuation, but also shows the shallowest wear track depth of 16.79 μm and the narrowest wear track width of 662.38 μm. The wear rate of material M2 is about 1.16 × 10⁻³ mm³·N⁻¹·m⁻¹. Apparently, the wear resistance of magnesium is greatly enhanced by the alloying and subsequent extrusion-annealing treatment.

Figure 7 shows the wear track topographies of all fabricated materials after the wear tests. As displayed in Figure 7a, the wear track surface of material M0 is characterized by plenty of large and deep grooves along the sliding direction. These grooves also appear on the wear track of material M1, as shown in Figure 7b, but their size and depth obviously decrease, indicating an alleviative plastic deformation during the sliding wear test. In contrast, material M2 exhibits a relatively smooth wear topography, as shown in Figure 7c, and some very small and shallow grooves distribute on the surface of the wear track. Apparently, friction force induces a significant extrusion effect of magnesium resulting in a severe plastic deformation on the surface of the wear track. However, the alloying and extrusion-annealing treatment greatly enhance the mechanical properties of magnesium, triggering a strong resistance to deformation and fracture, and consequently resulting in alleviative wear damage.

The SEM wear track images of all fabricated materials after the sliding wear tests are shown in Figure 8. As displayed in Figure 8a, the wear track surface of material M0 is characterized by numerous large size cracks, indicating severe structural damage and fracture during the sliding wear test. As shown in Figure 8b, these cracks also appear in the wear track of material M1, but their sizes obviously decrease, indicating an enhanced fracture resistance. In contrast, material M2 shows a smooth wear track morphology (see Figure 8c), without coating fracture or crack initiation, indicating a considerable fracture resistance during the sliding wear test.

To further investigate the wear mechanism of material M0, SEM cross-sectional characterizations are conducted to clarify the structural evolution of this sample after the sliding wear test. As shown in Figure 9a, the cross-sectional wear track of material M0 exhibits three distinct regions, including a flat fracture morphology, a honeycomb-like fracture morphology, and a fiber-like morphology. According to the previous literature [23,37,38], this honeycomb-like morphology away from the wear track represents a typical ductile fracture characteristic [37], while this flat morphology in the subsurface layer of the wear track represents a typical brittle fracture characteristic [23,38]. Lu et al. [39] reported that frictional contact induces plastic deformation underneath the wear track during the sliding wear test, inducing the formation of delaminating tribolayers, and eventually resulting in a high friction coefficient and wear rate of metals. Apparently, these fiber-like morphologies and visible fatigue cracks in Figure 9a confirm the formation of a plastic deformation layer in the wear track during the wear test. According to the partially enlarged view of the wear track in Figure 9b, it can clearly explain the fracture failure mechanism of material M0 during the sliding wear test. The continuous contact load and repeated sliding wear induce a significant plastic strain on the surface of material M0, as shown in Figure 9b, these plastic strains propagate and accumulate in the subsurface layer of the wear track resulting in the formation of a plastic hardening layer [39]. When the friction-induced shear stress exceeds the yield limit of magnesium, the plastic deformation layer shows poor stability, and shear instability usually occurs, which induces the initiation of cracks in the deformation layer. Subsequently, these cracks propagate inside the deformed layer and finally cause local material spalling [39]. Therefore, a large-scale block shedding of the material can be observed on the wear track surface of material M0 (Figure 8a). Moreover, the severe material fracture and spalling eventually lead to a fluctuated friction coefficient during the sliding wear test (Figure 5).

Compared with material M0, material M1 shows better anti-wear performance during the sliding wear test, as shown in Figure 10a. The cross-sectional morphology in the subsurface layer of the wear track shows a typical composite fracture characteristic, such as local brittle fracture and ductile fracture. Some visible local plastic straining accumulation can be observed on the cross-sectional image of the subsurface layer. Meanwhile, a very thin plastic deformation layer also forms on the surface of the wear track. According to XRD and SEM results in Figure 1 and Figure 3a, the enhanced wear resistance of material M1 is likely attributed to a series of synergistic strengthening effects induced by the alloying and subsequent extrusion treatment. According to the Hall-Petch strengthening theory [33], this significant grain refinement introduces numerous grain boundaries in α-Mg matrix, which is expected to benefit the improvement of mechanical and tribological properties [28]. Moreover, these doped atoms are dissolved in α-Mg matrix and induce the formation of the intermetallic compounds at the grain boundaries. Both of which greatly strengthen α-Mg matrix and effectively inhibit rapid initiation and propagation of fatigue cracks, suppressing the plastic strain accumulation and the formation of delaminating frictional layer [39], and consequently resulting in a relatively stable friction coefficient (Figure 5), slight friction induced fracture damage (Figure 8b), and enhanced wear resistance of material M1.

In contrast, the anti-wear performance of material M1 is further improved by the ageing treatment, as shown in Figure 10b, the cross-sectional SEM morphology in the subsurface layer of the wear track of material M2 exhibits a typical ductile fracture characteristic, without plastic strain accumulation or fatigue cracks, indicating a stronger resistance to wear and fracture damage. According to the previous literature [8], this aged magnesium alloy showed improved mechanical properties due to the formation of ${\beta }^{\prime }$ phase and homogenized texture. The SEM image in Figure 3b clearly reveals that aging treatment drives the homogenization of the growth texture and heals these micro-cracks between the matrix and eutectic phase. Therefore, the enhanced synergistic strengthening effect of grain refinement, solid solution, and second phase provides a stronger inhibitory effect in α-Mg matrix, as shown in Figure 8c, which effectively suppresses the generation of the delaminating frictional layer in the aged texture, and a lower wear rate and a smoother wear track are characterized in material M2.

4. Conclusions

This study investigated the effect of alloying—extrusion—aging treatments on the structure, mechanical and tribological properties of magnesium. The main conclusions are described as follows:

• The alloying—extrusion process refined the structure and induced the formation of Mg₂₄(GdYZn)₅ phases at the grain boundaries of magnesium. The ageing treatment drove the structural homogenization of alloyed-extruded magnesium. The grain size of magnesium decreased from 117 μm to 17 μm after the alloying—extrusion, and reached to 15 μm after the alloying—extrusion—aging treatments.
• The alloying—extrusion—aging treatments greatly enhanced the mechanical properties of magnesium due to a strong synergism of grain refinement, solid solution, and second phase strengthening. The hardness of magnesium increased from 0.67 GPa to 1.64 GPa after the alloying—extrusion and reached to 1.83 GPa after the alloying—extrusion—aging treatments.
• Magnesium experienced severe fracture during the sliding wear test, triggering the formation of a plastic deformation layer, and consequently resulting in a high wear rate of 2.64 × 10⁻³ mm³·N⁻¹·m⁻¹. The alloyed-extruded magnesium showed an improved wear resistance and a decreased wear rate of 1.60 × 10⁻³ mm³·N⁻¹·m⁻¹ because the alloying—extrusion process greatly inhibited the formation of the plastic deformation layer. In contrast, the alloying—extrusion—aging treatments showed stronger capability in suppressing the formation of a plastic deformation layer. The alloyed-extruded-aged magnesium exhibited better wear resistance and a lower wear rate of 1.16 × 10⁻³ mm³·N⁻¹·m⁻¹.