© Wuhan University 2024

0 Introduction

As space exploration advances, the environmental conditions in space are becoming harsher. The demands placed on exploration machines and their components are increasing. Aerospace materials have become critical to the success of space exploration. Bearings are the heart of space machines. Bearings are a crucial component of machines due to their excellent resistance to pressure, corrosion, wear, and self-lubrication. The materials used to make bearings are of utmost importance. Ni₃X intermetallic compounds with geometrically close packed (GCP) structures exhibit high stability and strength, as well as excellent oxidation and corrosion resistance at high temperatures. However, it has been observed that these compounds have poor ductility at room temperature and low creep strength at high temperatures. To address these issues, microstructures based on Ni₃X (where X is Al, Ti, and Nb, or Si, Ti, and Nb, or Al, Ti, and V, or Al, Nb, and V) have been extensively studied. Among these alloy systems, the dual-phase, two-phase intermetallic alloy Ni₃Al-Ni₃V is particularly well-suited for high-temperature mechanical components. This two-phase metal alloy exhibits a unique microstructure, with high interconnectivity, and is significantly more highly alloyed than many conventional metal alloys^[1].

Ni₃Al-Ni₃V two-phase metal compounds are considered to be some of the best candidates for bearings in important machines in the land, marine, and aerospace gas turbine industries^[2]. To enhance the mechanical properties of bearings, scholars have incorporated solid solution strengtheners into two-phase alloy materials. For instance, Ta with Re^[3], Nb with Ti^[4], NbC^[5], Fe^[6], and C^[7] have been used. Research has shown that the inclusion of Zr elements has a plasticizing effect on Ni₃Al-rich alloys^[8]. Additionally, the addition of Zr improves the ductility and oxidation resistance of Ni₃V alloys^[8-11]. Furthermore, Zr in the Ni phase region is beneficial for enhancing the stability of O within the Ni-Ni₃Al system^[12]. Intermetallic compounds that contain aluminium are highly susceptible to hydrogen embrittlement, which occurs when atmospheric water vapour reacts with the aluminium element in the alloy to form hydrogen, leading to embrittlement of the material. However, the addition of zirconium elements to nickel-based alloys suppresses environmental embrittlement of the alloy and provides better wear properties than Ni₃Al-Ni₃V alloys^[13]. However, research on the friction and wear characteristics of nickel-based alloys that contain zirconium elements is limited. There are few studies on complex gaseous environments similar to those in outer space. Therefore, this paper focuses on studying the friction and wear properties of Ni₃Al-Ni₃V-Zr-Ni₅Zr alloys in different atmospheres, such as air, oxygen (O₂), and carbon dioxide(CO₂). It compares these properties with those of Ni₃Al-Ni₃V alloys. The study aims to provide a better understanding of the friction and wear properties of these alloys in different environments.

1 Experimental Materials and Methods

1.1 Alloy Preparation

High purity Ni, Al, V and Zr were selected and melted into ϕ100 mm×15 mm ingots using a DHL-1250 high vacuum induction furnace. The actual chemical composition (mole fraction) of the melted alloys was 76.18Ni-8.75Al-14.92V-0.15Zr (referred to as Ni₃Al-Ni₃V-Zr-Ni₅Zr alloy) and 76.21Ni-8.92Al-14.87V (referred to as the Ni₃Al-Ni₃V alloy).

1.2 Alloy Heat Treatment Processes

The alloy underwent heat treatment using a ZM-45-15 vacuum molybdenum wire furnace. The heat treatment process for this alloy is shown in Table 1. A wear specimen measuring 35 mm×35 mm×7 mm was wire cut and its surface was ground using 800 grit sandpaper.

1.3 Experimental Methods

The friction and wear test is conducted using the WTM-2E controlled atmosphere miniature friction and wear tester. A vacuum hood is used to seal the specimen and the friction and wear apparatus. The chamber is evacuated before the test to remove the air from it. Then, the chamber is cleaned by passing in the gas required for the test. This process is repeated four times while the chamber continues to be evacuated. Friction and wear tests were conducted in three different atmospheres: air, O₂, and CO₂. A controlled flow of 0.05 MPa of O₂ and CO₂ was maintained, along with an ambient humidity of 10%-14% (measured by an Anymetre JR900 electronic hygrometer). The sample is to be rubbed with a 5 mm Si₃N₄ ball. The wear test involved applying pressure loads of 3, 10, and 20 N to the alloy wear specimen using a rotating ceramic ball. The sliding speed was set at 0.1 m/s, and the test lasted for 30 min. The friction coefficient was automatically recorded by the instrument. The wear depth and width of the specimens were measured using a NanoMap-500LS 3D contact surface profiler. The calculation of the wear rate is determined by the formula W=V /(F·L). Here, W represents the wear rate, V represents the wear volume, F represents the load, and L represents the wear distance. The FM-700e Vickers hardness tester was used to test the material hardness. The wear surface morphology and alloy microstructure were observed using a QUANTA 250 scanning electron microscope. The composition was analyzed with an energy spectrometer, and the volume fraction of the Ni₃Al phase was analyzed using Image J software.

2 Results and Analysis

2.1 Microstructure and Hardness Analysis of the Alloys

Figure 1 displays the histomorphology and X-Ray diffraction (XRD) patterns of the Ni₃Al-Ni₃V and Ni₃Al-Ni₃V-Zr-Ni₅Zr alloys. The black phase of the alloys is the Ni₃V (L12) precipitated phase, while the grey phase is the two-phase composite microstructure of Ni₃Al and Ni₃V^[13]. These findings are supported by the energy dispersive spectrometer (EDS) analysis results presented in Table 2.

Fig. 1 Microstructure and XRD patterns of Ni3Al-Ni3V and Ni3Al-Ni3V-Zr-Ni5Zr alloys (a) Ni3Al-Ni3V; (b) Ni3Al-Ni3V-Zr-Ni5Zr-Sample 1; (c) Ni3Al-Ni3V-Zr-Ni5Zr-Sample 2; (d) Ni3Al-Ni3V-Zr-Ni5Zr-Sample 3; (e) Ni5Zr-phase; (f) XRD of alloys

The high structural stability of the two-phase structure^[14,15] results in superior high-temperature mechanical properties, such as tensile and creep strength, compared with many conventional high-temperature alloys. The Ni₃Al-Ni₃V-Zr-Ni₅Zr alloy contains a high Zr content in the white phase. The transmission electron microscope (TEM) analysis reveals the presence of the brittle Ni₅Zr phase in this region, which results in poor plasticity throughout the workpiece^[16]. Sample 1 has a Ni₃Al phase size of approximately 250 nm, sample 3 has a Ni₃Al phase size of approximately 500 nm, and sample 2 has a Ni₃Al phase size of approximately 1-4 μm. The nanoscale structure is the primary cause of the "bun peaks" in samples 1 and 3 XRD patterns.

Table 3 presents the hardness data for various materials relative to the hardness of Ni₃Al-Ni₃V, which ranges from 498.3 to 521.4 HV. Additionally, the volume fractions of Ni₃Al in samples 1, 2, and 3 were 41.601%, 62.179%, and 55.216%, respectively. The Ni₃Al+Ni₃V phase is the hard phase, while the Ni₃Al phase is the soft phase. Therefore, as the volume fraction of Ni₃V increases, the alloy's hardness decreases.

2.2 Frictional Wear Properties of the Alloy

Figure 2(a) displays the friction coefficients for the Ni₃Al-Ni₃V and Ni₃Al-Ni₃V-Zr-Ni₅Zr alloys. The coefficient of friction for all alloys decreases with increasing load. In an air environment, the coefficient of friction of the Ni₃Al-Ni₃V alloy is significantly larger than that of the Ni₃Al-Ni₃V-Zr-Ni₅Zr alloy. The coefficient of friction of the Ni₃Al-Ni₃V-Zr-Ni₅Zr alloy is highest in an O₂ environment, followed by CO₂ and air environments. Figure 2(b) displays the wear rates of the Ni₃Al-Ni₃V and Ni₃Al-Ni₃V-Zr-Ni₅Zr alloys. It is evident that the wear rate of the Ni₃Al-Ni₃V alloy is significantly higher in the air environment compared to that of the Ni₃Al-Ni₃V-Zr-Ni₅Zr alloy. Additionally, the wear rate of the Ni₃Al-Ni₃V-Zr-Ni₅Zr alloy is higher in the CO₂ environment than in the O₂ environment and air environment. At low load, the wear rate of the Ni₃Al-Ni₃V-Zr-Ni₅Zr alloy is the same in both air and O₂ environments. As the load increases, the wear rate in the air environment remains relatively stable, while in the O₂ environment it increases rapidly.

Fig. 2 Comparison of alloy friction coefficients (a) and wear rates (b)

2.3 Wear Surface Morphology

Figure 3 displays the wear surface morphology of the Ni₃Al-Ni₃V alloy under air conditions. At a load of 3 N, the surface of the alloy exhibits significant cracks, characterized by ploughing, spalling, and slight adhesion, with blocky spalling around the cracks. This is where the friction coefficient and wear rate of the alloy reach their maximum. As the load increases to 10 N, ploughing and adhesion become more pronounced, while cracking weakens and the alloy's coefficient of friction and wear rate decrease significantly. When the load increases to 20 N, adhesion increases, cracking weakens further, and the alloy's wear rate and coefficient of friction decrease even more. Furthermore, observations of the parallel wear surface indicate the presence of a distinct sharp angle feature and cracking characteristics. Cracks were observed sprouting on the wear surface and expanding in a direction approximately 18 degrees from the wear surface (Fig. 3(d)), indicating hydrogen embrittlement cracking characteristics. Therefore, both load and hydrogen embrittlement significantly influence the frictional wear properties of the Ni₃Al-Ni₃V alloy.

Fig. 3 Surface abrasion profile of Ni3Al-Ni3V alloy (a) 3 N; (b) 10 N; (c) 20 N; (d) Crackle

The EDS analysis and the wear surface morphology of the three Ni₃Al-Ni₃V-Zr-Ni₅Zr alloys are displayed in Table 4 and Figs 4 to 6, respectively. The wear surfaces of the alloys under air conditions (Figs. 4(a)-(c), 5(a)-(c) and 6(a)-(c)) exhibit mainly ploughing and adhesion (Fig. 4(b)-I). The EDS analysis (Table 4) shows a high content of Ni and O elements in this adhesion, along with a small amount of cracking and spalling features on the alloy surfaces. As the load increases, the adhesion features weaken, and the grooved features slightly increase. The weakened adhesion features reduce the friction coefficient of the alloy.

Fig. 4 Surface wear of alloys of Sample 1 (a)-(c): Air of 3, 10 and 20 N; (d)-(f): O2 of 3, 10 and 20 N; (g)-(i): CO2 of 3, 10 and 20 N

The wear surfaces of the alloy in O₂ (Figs. 4(d)-(f), 5(d)-(f), and 6(d)-(f)) exhibit more pronounced grooving, adhesion, and spalling features than in the air. There is also a small amount of surface cracking (Fig. 4(e)-II) and an increased loading of black material (Fig. 4(f)-III). The oxygen content (Table 4) and surface black material are also increased. According to EDS analysis, the black material found at the wear site is Si compounds that were transferred from Si₃N₄. The high O content observed at the cracked features of the alloy wear surface indicates that an oxide film covered the wear surface. Energy spectral analysis revealed a large number of spalling bands on the wear surface (Fig. 5(d)-IV), which were identified as the alloy matrix. It can be inferred that these bands were caused by Si₃N₄ cutting the alloy matrix. The wear surface of the alloy in the CO₂ environment exhibited slight ploughing, adhesion, and spalling features at 3 N load, with a small amount of flake spalling also observed. The adhesion features appeared powder-like. At 10 N load, the ploughing on the wear surface increased, and slight plastic deformation occurred, along with cracks perpendicular to the direction of wear.

Fig. 5 Surface wear of alloys of Sample 2 in different atmospheres (a)-(c): Air of 3, 10 and 20 N; (d)-(f): O2 of 3, 10 and 20 N; (g)-(i): CO2 of 3, 10 and 20 N

3 Discussion

3.1 Analysis of the Frictional Properties

In Fig. 2(a), it can be observed that Sample 2 has a slightly larger coefficient of friction compared to Samples 1 and 3. This is due to the fact that the hardness of the material has a significant impact on the coefficient of friction. As the hardness increases, the contact area between the workpiece and the two objects that rub against each other decreases^[17], and the frictional wearability of the material is directly proportional to the hardness^[18]. Thus, the coefficient of friction of sample 2 is higher than that of samples 1 and 3.

The coefficient of friction of the Ni₃Al-Ni₃V and Ni₃Al-Ni₃V-Zr-Ni₅Zr alloys tends to decrease with increasing load (Fig. 2(a)) due to the fact that the coefficient of friction is determined by the ratio of the shear strength to the yield strength of the softer of the two metals that rub against each other. The ratio of shear strength to yield strength of the same metal is determined by the fact that the shear strength and yield strength of the same metal are certain and that they vary simultaneously. During frictional wear, the heat generated by friction is confined to the surface layer. At higher loads, the change in shear strength is more pronounced for softer metals compared to yield strength, leading to a decrease in the coefficient of friction, i.e. the decrease in shear strength is confined to the surface layer of the metal and the metal matrix remains very large The change in yield strength follows the trend of Ref. [17].

3.2 Cracking and Environmental Embrittlement Mechanisms on Wear Surfaces

According to Hertz's classical equation (1)^[17], the Ni₃Al-Ni₃V alloy produced significant cracking features on the wear surface (Fig. 3), while the Ni₃Al-Ni₃V-Zr-Ni₅Zr alloy had no significant cracking features on the wear surface (Figs. 4 to 6). An analysis was carried out.

$a=\mathrm{1.1}{\{\frac{Fr}{\mathrm{2}}\left(\frac{\mathrm{1}}{E_{\mathrm{1}}}+\frac{\mathrm{1}}{E_{\mathrm{2}}}\right)\}}^{\frac{\mathrm{1}}{\mathrm{3}}}$(1)

where a (mm) is the radius of the contact zone, r (mm) represents the radius of the collimated sphere Si₃N₄, F(N) represents the load, and E₁ (GPa) and E₂ (GPa) represent the modulus of elasticity of Si₃N₄ and the sample alloy, respectively (refer to Tables 5 and 6). A= πa² will be proportional to W^2/3 and the average pressure in the contact zone, P_m=W/πa² will be proportional to W^1/3[17]. Equation (1) calculates A for the Ni₃Al-Ni₃V alloy and the Ni₃Al-Ni₃V-Zr-Ni₅Zr alloy at different loads (Tables 5 and 6). The results show that the true contact surface is in plastic contact, and this is deduced from the value of E₁, which is approximately 310 GPa.

In terms of strength factor, Gerberich et al^[19] concluded that the effective stress field strength factor at the crack tip is K_I, which increases with the number of dislocations emanating from the crack tip, and when it is greater than or equal to K_IC, the material should fracture, which is referred to as "plasticity-induced disintegration".

$K_{\mathrm{I}}=\mathrm{1.12}\tau$(2)

In Eq. (2), K_I is the crack tip stress intensity factor. According to the calculation, it can be seen that K_I of Ni₃Al-Ni₃V (1.63 (MPa·m)^1/2) is much smaller than K_IC of Ni₃Al-Ni₃V (>20 (MPa·m)^1/2), so the Ni₃Al-Ni₃V alloy wear surface cracking is not caused by contact stress.

The environmental sensitivity of nickel aluminium alloys increases with increasing aluminium content due to the higher activity of aluminium compared to nickel. This is reflected in the decreasing free enthalpy change of reaction with water vapour, with the order of Ni < Ni₉₄Al₆ < Ni₃V < NiAl < NiAl₃^[20]. Table 6 displays the Gibbs free energy for the reaction of each element with water vapour in the Ni₃Al-Ni₃V alloy. When the Gibbs free energy of each element in the alloy is less than -120 kJ/mol, the alloy becomes highly sensitive to the environment^[20]. The main cause of environmental hydrogen embrittlement in the alloy is the reaction of Al and V elements with water vapour, which produces H atoms.

In comparison to the air wear surface of the Ni₃Al-Ni₃V alloy shown in Figs. 3(a)-(c), the air wear surfaces of the Ni₃Al-Ni₃V-Zr-Ni₅Zr alloys in Fig. 4 (a)-(c), Fig. 5(a)-(c) and Fig. 6 (a)-(c) exhibit minimal crack generation. This is attributed to the presence of Zr, an oxygen-loving element that is distributed within the grain as an oxygen trap, thereby preventing oxygen from entering the grain boundaries and reducing oxygen levels. The diffusion of oxygen along the grain boundaries is reduced, which in turn reduces the generation of grain boundary cracks. The plasticity of the alloy is improved by the bias and diffusion of Zr elements to the grain boundaries^[22]. Additionally, the element Zr readily forms an oxide film of Zr with oxygen that covers the surface of the workpiece. Pre-oxidation of the sample produces a protective oxide film that reduces the material's susceptibility to hydrogen embrittlement^[23]. The presence of Zr results in the consumption of surface Al in the alloy. This reduces the proportion of H atoms produced by the combination of Al and atmospheric moisture, leading to an improvement in environmental embrittlement^[13].

Fig. 6 Surface wear of the alloy of Sample 3 in different atmospheres (a)-(c): Air of 3, 10 and 20 N; (d)-(f): O2 of 3, 10 and 20 N; (g)-(i): CO2 of 3, 10 and 20 N

3.3 Mechanisms for the Influence of Oxide Films on the Frictional Wear Properties of Alloy Surfaces

When elements such as Al in the Ni₃Al-Ni₃V-Zr-Ni₅Zr alloy are exposed to air, oxides such as aluminium oxide are formed with water vapour and O₂ in the air, and the friction surfaces are rapidly oxidized as soon as they come into contact with air. The oxidation does not take long, and an oxide film 10 to 100 angstroms thick can be obtained in a short time^[17], and this oxide provides good protection for the alloy substrate. Therefore, the main wear mechanisms of the Ni₃Al-Ni₃V-Zr-Ni₅Zr alloy in the air environment are oxidative and abrasive wear, with low friction coefficients (Fig. 2(a)) and wear rates (Fig. 2(b)).

In an O₂ environment, the oxygen concentration is approximately 2.5 times that in air due to the O₂ pressure of 0.5 MPa. According to Table 7, Zr and Al will preferentially react with O₂ to form ZrO₂ and Al₂O₃ (Zr+O₂ = ZrO₂ with 4/3Al+O₂ = 2/3Al₂O₃ ), V will react with O₂ to form V₂O₅ (4/5V+O₂ = 2/5V₂O₅ ), V₂O₅ and Al to form Al₂O₃, and in the chemical formula 4/3Al+O₂= 2/3Al₂O₃ with low Gibbs free energy synergy, the alloy surface will continue to produce Al₂O₃ oxide film; accompanied by friction-induced surface temperature increase, the wear surface rapidly reacts to generate oxide film, which on the one hand protects the workpiece surface, but when the oxide film superimposed reaches a thickness of 200 Å, the oxide film may break, increasing the friction coefficient of the alloy^[17]. When the load is 3 N, the temperature rise of the wear surface and the ploughing of the micro convex body on the Si₃N₄ surface of the wear part are weak, and the oxide film is generated and peeled off at a slower rate, resulting in a lower wear rate of the alloy. When the load is 10 N (Fig. 4(e), Fig. 5(e), Fig. 6(e)), the temperature of the alloy surface rises rapidly in frictional wear, the oxygen atoms become more active, accelerating the formation of oxides, and when the oxide film reaches a certain thickness, it becomes fragile and easily cracks or flakes off. The wear rate of the alloy then increases. At a load of 20 N (Fig. 4(f), Fig. 5(f) and Fig. 6(f)), the oxide film on the wear surface rubs violently against its Si₃N₄ counterpart, resulting in violent flaking of the oxide film against the Si₃N₄, and the flaked Si₃N₄ fragments are transferred to the wear surface of the alloy. These hard Si₃N₄ shear forces combine with the high load to increase the alloy's wear, and the alloy's wear rate (Fig. 2(b)) continues to increase. In the O₂ environment, the main wear mechanisms are oxidative and adhesive wear at low loads and abrasive and oxidative wear at high loads.

3.4 Mechanisms for the Formation of Oxide Films on Alloy Wear Surfaces

A simplified model is proposed based on the analysis of the frictional wear process of the Ni₃Al-Ni₃V-Zr-Ni₅Zr alloy, as shown in Fig. 7. In the early stages of wear, friction tends to be unstable due to insufficient frictional heat and the presence of elements such as Zr, Al, and Ni distributed in the alloy (Fig. 7(a)). As wear progresses, the material surface undergoes plastic deformation accompanied by extrusion deformation, generating a large amount of frictional heat. During extrusion and deformation (as shown in Fig. 7(b)), elements like Zr, Al, and Ni tend to accumulate towards the surface. These elements then react with oxygen to form their respective oxides (such as ZrO₂, Al₂O₃, and NiO) which are compressed to create oxide films that adhere to the surface (as seen in Fig. 7(c)). As wear continues, cyclic loading leads to a gradual reduction in wear flakes, a gradual build-up of frictional heat, an increase in the adhesive properties of the oxide film, and gradual compaction of the oxide film (see Fig. 7(d)). The thickness of the oxide film increases gradually under the load compaction. However, when the oxide film reaches a thickness of 200 Å, it becomes brittle and tends to crack or flake off (see 1 in Fig. 4(e)). Continued friction results in the production of a new oxide film, leading to flaking and an increase in wear rate. This cycle repeats, further increasing the wear rate. Localized oxide films can break down under excessive contact stress, causing surface oxide particles to enter the coating and slightly increasing the coefficient of friction and wear rate.

Fig. 7 Mechanisms for the formation of alloy oxide films

4 Conclusion

1) The addition of Zr elements to the Ni₃Al-Ni₃V alloy resulted in a significant decrease in its hardness. Furthermore, an increase in the volume fraction of Ni₃V in the alloy led to a decreasing trend in its hardness.

2) The magnitude of the friction coefficient of Ni₃Al-Ni₃V-Zr-Ni₅Zr alloy follows the order of O₂ environment > CO₂ environment > air environment, while the magnitude of the wear rate follows the order of CO₂ environment > O₂ environment > air environment. The addition of Zr is important to the frictional wear performance of Ni₃Al-Ni₃V-Zr-Ni₅Zr alloy, which exhibits better friction coefficient and wear rate than Ni₃Al-Ni₃V alloy.

3) The wear surface cracks on Ni₃Al-Ni₃V alloys are not caused by contact stress, but rather by H atoms generated through the reaction of Al elements with water vapour, leading to environmental hydrogen embrittlement of the alloys. The incorporation of Zr elements enhances the ductility of Ni₃Al-Ni₃V alloys, reduces the surface concentration of Al elements, and decreases the amount of H atoms produced by the reaction between Al and atmospheric moisture, thereby mitigating the environmental embrittlement phenomenon.

4) In the air environment, the surface of the alloy is oxidized under the action of air, accompanied by frictional wear, and the surface generates an Al₂O₃ oxide film to protect the workpiece; at this time, the friction coefficient and wear rate of the Ni₃Al-Ni₃V-Zr-Ni₅Zr alloy are extremely low, and the main wear mechanisms in this environment are abrasive wear and oxidative wear; the oxide film generated on the surface in the O₂ environment is very good at low loads When the oxide film reaches a certain thickness it becomes fragile and is prone to cracking or flaking. The flaking surface oxide particles plough through the alloy surface, increasing wear rates. The low O₂ and water vapour content of the CO₂ environment prevents the corresponding oxides from covering the surface and reducing wear; the main wear mechanism in the CO₂ environment is abrasive wear.

References

All Tables

All Figures

	Fig. 1 Microstructure and XRD patterns of Ni3Al-Ni3V and Ni3Al-Ni3V-Zr-Ni5Zr alloys (a) Ni3Al-Ni3V; (b) Ni3Al-Ni3V-Zr-Ni5Zr-Sample 1; (c) Ni3Al-Ni3V-Zr-Ni5Zr-Sample 2; (d) Ni3Al-Ni3V-Zr-Ni5Zr-Sample 3; (e) Ni5Zr-phase; (f) XRD of alloys
In the text

	Fig. 2 Comparison of alloy friction coefficients (a) and wear rates (b)
In the text

	Fig. 3 Surface abrasion profile of Ni3Al-Ni3V alloy (a) 3 N; (b) 10 N; (c) 20 N; (d) Crackle
In the text

	Fig. 4 Surface wear of alloys of Sample 1 (a)-(c): Air of 3, 10 and 20 N; (d)-(f): O2 of 3, 10 and 20 N; (g)-(i): CO2 of 3, 10 and 20 N
In the text

	Fig. 5 Surface wear of alloys of Sample 2 in different atmospheres (a)-(c): Air of 3, 10 and 20 N; (d)-(f): O2 of 3, 10 and 20 N; (g)-(i): CO2 of 3, 10 and 20 N
In the text

	Fig. 6 Surface wear of the alloy of Sample 3 in different atmospheres (a)-(c): Air of 3, 10 and 20 N; (d)-(f): O2 of 3, 10 and 20 N; (g)-(i): CO2 of 3, 10 and 20 N
In the text

	Fig. 7 Mechanisms for the formation of alloy oxide films
In the text