Features of the Formation of Surface Structures under Dry Friction of Al-30Sn Composite against Steel

Abstract

Features of the frictional interaction of a sintered Al-30Sn composite, which is used as a coating of bearing inserts, with steel counterbody in the absence of liquid lubricant were studied. The tribological tests were carried out according to the pin-on-disk scheme at room temperature. The friction coefficient $\mu$ of the friction pair is increased up to approximately 0.6 during the running-in process. Its growth stops when the stage of steady state friction begins, and then $\mu$ fluctuates around some relatively high and constant value. The study of the friction surfaces of the friction pair showed that the aforementioned increase in $\mu$ is due to the formation of a discrete transferred layer on the friction track surface and strain hardening of the subsurface layer of the Al-30Sn sample under the action of hard particles of the transferred layer. It was established that tin is transferred on the friction track surface mainly in a composition of wear particles, despite the high content of the solid lubricant in the investigated samples. The wear intensity of the samples subjected to processing by equal channel angular pressing with route A is much lower than that of the sintered (unprocessed) ones. The main wear mechanism of Al-30Sn composites under dry friction against steel is a delamination of the highly deformed subsurface aluminum grains along their interphase boundaries.

1. Introduction

Increasing the service life of friction units in machines and mechanisms is one of the most important problems of modern materials engineering [1,2]. Wide application of aluminum alloys as bearing materials is limited by their property of seizure with a hard counterbody under dry friction, even under small loads [2,3,4,5,6,7]. However, their pressure of seizure can be increased by an addition in aluminum of at least one of the soft metals (Sn, Cd, Pb, Bi, et al.), which are not dissolved in solid Al [8,9,10,11]. Additions of tin are most often used in this case. It is supposed that the high resistance of such alloys to adhesive wear is caused by the ability of inclusions of the soft metal to perform the function of additional lubricant under dry and boundary friction [12,13]. That is, it is implicitly assumed that, as in the case of porous oil-filled sliding bearings, a thermal expansion of the subsurface layer of Al-Sn alloys leads to extrusion of tin on the friction surface from the harder Al matrix. It can be expected that such an extrusion will be especially effective when tin is melted, and its volume is additionally increased by 2.6%.

However, this is a misconception about the lubricating ability of Sn. Unlike an oil lubricant, the wetting angle of Al by liquid tin is more than 100°, and it is sharply decreased only at temperatures above 600 °C [14]. Therefore, the spreading of liquid tin over the aluminum surface, which is also covered with oxide films, is impossible at the lower temperatures. In the case of an absence of strong adhesive bonds, the extruded liquid Sn can neither spread nor stay on the aluminum surface. The uniform spreading of thin Sn layers over the friction surface using a mechanical method is also unlikely. Some amount of tin can be saved on the friction surface only in deep cavities.

However, it was found that during the dry friction of a pure tin sample against steel, some part of it is transferred from the sample surface on the friction track and covers it with a thick layer. The phenomenon of transfer of tin atoms on the surface of a counterbody at the beginning of the frictional interaction has been also confirmed by the labeled atoms method in the case of dry friction contact between the Al-Sn alloy and steel. It was also established [15] that if such a transferred layer is not renewed periodically, it quickly wears out.

Tin is always present on the friction track surface during the stage of steady state friction of sintered Al-Sn alloys against steel without lubrication [16]. However, it was found in the composition of particles transferred on the friction track of the steel counterbody only when they reached a certain thickness [17]. The authors of this work suppose that an alloying of the transferred layer with elements of the steel counterbody and atoms of the surrounding atmosphere provides its high antifriction properties with an effect of solid lubricant.

The last statement is debatable, although high oxygen content is found on both coupled friction surfaces during the frictional interaction between the Al-Sn composite and steel [16,17,18]. However, it is known that oxygen is not dissolved in solid aluminum and tin, and, therefore, it can be only in the form of oxide films on the friction surface. Such hard films are not a solid lubricant for softer Al-Sn composites, but they can cover the friction surface and effectively prevent an adhesion between the coupled bodies [19].

A material is characterized as an antifriction one in the tribology if its friction coefficient does not exceed 0.3 in the absence of liquid lubricant. Hence, antifriction materials should have an effect of self-lubrication, which consists of supply of the soft phase on the friction surface. A protective layer of the formed secondary structures acts as a solid lubricant in critical conditions of friction (dry friction, shock loads) and prevents a fracture of the shaft [20]. According to this view, the sintered Al-30Sn composite can be considered as a prospective antifriction material because it contains a lot of soft phase. For example, it is known that the addition of tin in aluminum alloy leads to an increase in its pressure of seizure, and the wear intensity of the Al-Sn alloys is decreased with increasing tin content up to 40 wt% under the dry friction against steel [21].

Thus, it can be concluded that the positive effect of tin additives on the wear resistance of aluminum alloys is obvious. However, the mechanism of this effect has not been studied in detail. For example, it is possible that the high wear resistance of Al-Sn alloys with high Sn content can be caused by the action of tin as a solid lubricant on the friction surface and also between the Al grains [22]. For better understanding of the aforementioned problem, this work was carried out. The aim of this work was to study the features of formation of secondary structures on the friction surfaces of a steel disk and an Al-Sn composite sample depending on the duration of the dry friction test.

2. Materials and Methods

2.1. Materials Preparation and Structural Studies

The investigated samples were prepared by liquid phase sintering of briquettes consisting of the mixture of industrial aluminum and tin powders according to the method described in [23]. The content of Sn in the mixture was 30% by weight. The sintered Al-30Sn alloy is a composite material consisting of two mutually insoluble solid phases. The method of sintering avoids problems with uniform distribution of the phases, which differ greatly in density.

In order to strain harden the aluminum matrix of the composites and eliminate their residual porosity, the sintered samples were subjected to equal channel angular pressing (ECAP) with route A at a temperature of 200 °C in the press mold with channels intersecting at a right angle. ECAP with route A is carried out without rotation of the sample between the passes [24]. The value of deformation ($ϵ$) of the sample is approximately 1 during each pass.

This processing leads to a significant decrease in the porosity of the sintered samples and an increase in their strength [23]. Moreover, its structure has a layered form in the flow plane that is perpendicular to the shear plane of the sample. It was established that the thickness of Al grains and tin interlayers is decreased in the flow plane of the sample because the deformation is carried out according to the scheme of simple shear during each pass of ECAP with route A [25].

Cross sections for metallographic analysis of the structure of the Al-30Sn composites were prepared by mechanical grinding of the sample surfaces with emery paper with different abrasive grain sizes and subsequent polishing with a cloth with applied paste containing small (<1 $\mathsf{\mu }$m) abrasive diamond particles. The samples were investigated in a cross-section to avoid the presence of surface defects. Next, the short-term chemical etching of the polished surfaces was made in a 4% solution of nitric acid in ethanol for 15 s.

The structure of the Al-30Sn composites and friction surfaces of the coupled bodies was studied using an AXIOVERT-200MAT optical microscope (Carl Zeiss AG, Jena, Germany) and a LEO EVO 50 scanning electron microscope (Carl Zeiss AG, Jena, Germany) with energy-dispersive analysis (EDX). The devices were given by Shared Use Center Nanotech (Institute of Strength Physics and Materials Science, Siberian Branch, Russian Academy of Sciences).

2.2. Tribological Tests

The tribological tests of the friction pair “Al-30Sn composite sample–steel” were carried out on a tribotester (Tribotechnic, Clichy, France) according to the pin-on-disk scheme (Figure 1). The disk made of a hardened structural steel 40H (AISI 5140 steel) with a hardness of 47 ± 2 HRc was used as a counterbody. Rectangular samples with an end friction surface of 3 × 3 mm were cut from the sintered and ECAP-ed briquettes. The flow plane of the ECAP-ed composites was selected as a friction surface, and the sliding direction was perpendicular to the elongated interlayers of the phases.

The friction surfaces of the samples and counterbody were prepared by the same method as the metallographic sections. These surfaces were cleaned with acetone before the tribological tests. The pressure on the friction surface (P) was 1–5 MPa. The sliding velocity (${\mathrm{V}}_{sl}$) was 0.07 m/s, and the radius of the friction track was 15 mm for each test. The relief of the friction surfaces was studied with a Zygo New View 6200 scanning whitelight interference microscope (Zygo Corporation, Middlefield, OH, USA).

The tribotester was equipped with a built-in computer that allows automatic measurement of the value of the friction coefficient ($\mu$) according to the law of Amontons–Coulomb: $\mu$ = F/N, where F is the friction force, and N is the normal load. The value of $\mu$ is assumed to be constant for the selected friction pair because the aforementioned law assumes that properties and composition of the friction surfaces are not changed during the measuring process. If the sliding distance is large enough for steady state friction, the relief, chemical composition, and physical–mechanical properties of the friction surfaces will be significantly changed as compared to the initial ones. The real area of contact between the coupled bodies will be also changed. That is, $\mu$ will characterize the tribological properties of the materials with a new structure that is changed during the friction test. Only the value of N remains unchanged under these conditions. Moreover, in order to compare the value of load with the mechanical characteristics of the tested materials, it is better to use stresses per unit of contact area, i.e., the value of pressure (P). That is, in this case, the value of friction coefficient can be defined as: $\mu$ = f/P, where f—specific (per unit of contact area) friction force. If P = 1 MPa, as it was in our case, the automatically measured value of $\mu$ will be equal to the numerical value of f, which is a physical characteristic of the friction process.

The wear intensity ${\mathrm{I}}_h$ ($\mathsf{\mu }$m/m) of the samples was defined as the ratio of reduction in the specimen height $\Delta$h ($\mathsf{\mu }$m) to sliding distance L (m).

3. Results and Discussion

The structure of the sintered Al-30Sn composite is shown in Figure 2a. It can be seen that dark Al grains form a continuous net, and light tin interlayers are located on their boundaries. If the composite with such a structure is processed by ECAP with route A, the grains of aluminum matrix and tin inclusions have a fibrous shape in the flow plane of the sample. The thickness of the fibers is decreased with an increasing number of ECAP passes (Figure 2b,c). The reduction in the distance between tin fibers is favorable for the self-lubricating process when the sliding of a counterbody is carried out in a direction perpendicular to the elongated fibers.

Simultaneously, the ECAP processing leads to a significant increase in the strength of the Al-30Sn composite (Figure 3). The most significant increase in the yield strength (${\sigma }_{0.2}$) and flow stress at compression by 4% (${\sigma }_4$) values was after the first ECAP pass. The increase in the number of passes from two to four leads to a slight increase in the strength of the composite. However, at the same time, its ductility is significantly decreased. The same results were reported in [23,25] in the case of Al-Sn composites with other Sn content. Therefore, two passes by ECAP with route A are enough for a significant increase in the strength and wear resistance of the Al-30Sn composite, especially at high pressures on the friction surface (

Table 1. Effect of the number of ECAP passes (N) on the mechanical properties and linear wear intensity ${\mathrm{I}}_h$ at different pressures (P) under dry friction against steel of an Al-30Sn composite; ${\mathrm{V}}_{sl}$ = 0.07 m/s; L = 500 m.

N	Mechanical Properties		Linear Wear Intensity ${\mathrm{I}}_{\mathit{h}}$, $\mathsf{\mu }$m/m
	${\mathit{\sigma }}_{0.2}$	${\mathit{\sigma }}_4$	1 MPa	3 MPa	5 MPa
0	44	67	0.12	0.30	0.42
1	106	114	-	-	-
2	120	136	0.10	0.20	0.28
4	136	145	-	-	-

).

The value of $\mu$ is continuously increased with an increase in the sliding distance during the running-in process. Subsequently, this stage is gradually replaced by a stage of steady state friction that is characterized by fluctuations of $\mu$ around some average value (≥0.6). A sliding distance of approximately 300 m is required to reach a maximum $\mu$ for the sintered and ECAP-ed composites (Figure 4). At the same time, it follows from the graphs of $\mu$(L) that the steady state friction coefficient and specific friction force of the Al-30Sn composite are almost not changed after the ECAP processing. The measured value of $\mu$ is higher than that of traditional antifriction materials under the dry friction process. Therefore, the investigated friction pair “Al-30Sn composite—steel” is not an antifriction one under the test conditions.

It was found from the study of the friction surfaces of sintered and ECAP-ed Al-30Sn samples that their relief and composition are similar during the stage of steady state friction. The images of the friction surfaces are shown in Figure 5. Defective areas filled with wear debris and relatively flat areas with parallel grooves can be seen. The chemical composition of the friction surfaces consists of initial (Al and Sn) and additional (O and Fe) elements (

Table 2. Elemental composition (EDX) of the friction track on the steel counterbody shown in Figure 6a.

Sample	Elements, % mass.
	Oxygen	Aluminium	Iron	Tin
Point 1	4.33	3.44	91.89	0.34
Point 2	28.70	24.90	40.53	5.87
Point 3	9.43	7.95	80.41	2.21
Point 4	3.90	1.96	93.98	0.16
Friction surface of the Al-30Sn sample (Figure 5b)	42.80	34.94	9.71	12.55

).

It is clear that the parallel grooves on the friction surface of the samples are formed by some irregularities on the steel disk or hard particles fixed on the friction track surface. The structure and elements distribution map of the friction track surface are shown in Figure 6. The concentration of the elements is given in Table 2.

It follows from Figure 6a that parallel grooves are also present on the friction track. That is, fixed hard particles having high strength and hardness are also located on the friction surface of the sample. Nevertheless, in spite of their abrasive effect, some quantity of transferred Al particles remains on the friction track surface. Tin content on this surface is negligible, and it is mainly located in the transferred Al particles (points 2 and 3 in Table 2 and Figure 6). Oxide films cover the entire friction track surface, but higher content was also found in the transferred particles.

4. Discussion

According to the data in Table 1, the wear intensity of the sintered Al-30Sn composite is significantly higher than that of the ECAP-ed one. The enhanced wear resistance of the processed composite may be because of the following reasons: formation of a layered structure with a short distance between the sources of solid lubricant (Figure 2); structural strengthening and an increase in ductility of the composite due to elimination of pores and improvement of the interphase bonds; strain hardening of the aluminum matrix (Figure 3).

As to the first reason, it was shown in [23,25] that the wear resistance of Al-Sn composites was almost the same after ECAP processing with routes A and C. The difference in ${\mathrm{I}}_h$ was due to a difference in the size of subgrains in the aluminum matrix and its strength according to the Hall–Petch law, respectively. That is, the formation of a layered structure during ECAP with route A has little effect on the wear intensity of Al-Sn composites. Moreover, the increase in the wear resistance of cast Al-30Sn alloy from forming a nanostructured state after ECAP with route BC [26] was the same as in our case (route A).

Usually, every sintered material has structural defects such as pores and areas of poor adhesion due to the presence of oxide films on the surface of the raw powders. The elimination of these defects increases the ductility of the composite. As a result, it can be deformed up to higher acting stresses before beginning softening processes such as localization of deformation in shear bands or cracking. For this purpose, it is sufficient to subject the composite sample to compression in a closed press mold. In this case, the deformation of the sample is small, but it is carried out at high pressure and temperature. As a result, their density becomes close to the theoretical one. The wear intensity of sintered Al-Sn composites processed by this method is significantly decreased [27]. That is, an increase in the ductility and strengthening of composites due to removing of structural defects during the densification of the composite samples give the main contribution in their wear resistance under a dry friction process.

ECAP processing has the same effect on the structural strengthening of Al-30Sn composites. The porosity of the sintered composite was 5.1%, while the density after the ECAP processing is 0.1%. The pores are eliminated, and adhesive bonds become stronger due to deformation of the grains and fragmentation of the surface oxide films. The strength of the ECAP-ed samples is higher than that of the samples compressed in a closed press mold, but the degree of ordering of the structure in the ECAP-ed materials is also higher. The latter fact is important because after the removing of other structural defects, tin interlayers on the aluminum grain boundaries become weak places in the composites and contribute to a localization of plastic flow. Hence, the level of shear stresses required for the beginning of a localization of plastic flow is lower in the case of Al-Sn composites having the layered structure. Moreover, metals with FCC lattices such as Al have a high tendency for severe strain hardening during the linear stage of plastic deformation with length of approximately 2%–3% [25]. The stresses of plastic flow at the end of this stage are the main contributor to the ultimate strength of the samples. This fact also provides the rapid reaching of stresses of plastic flow localization in the composites unprocessed by ECAP. In turn, for the same reasons, the stress of shear localization in Sn interlayers is not much different for the ECAP-ed and compressed Al-Sn composites having high content of the soft phase (Figure 3).

Thus, the main contribution of ECAP with route A to increase the wear resistance of sintered Al-30Sn composite under a dry friction process is the elimination of structural defects and, as a consequence, an increase in ductility. This fact allows for a decrease in the wear intensity of subsurface layers of the sample due to their larger deformation before the formation of wear particles. Other factors, such as formation of the layered structure as well as strengthening of the Al matrix, can lead to only a slight increase in the ductility of the composite and thus improve its wear resistance under the dry friction process. The deformation is concentrated on the friction surface of the composite sample under the dry friction process. As a result, the upper layer of the matrix grains is shifted and automatically involves the deformation lower layers of the grains in the presence of cohesive bonds between them. When there are many necks between the grains, their relative shift is difficult, and thus the material flow width is increased up to several layers of the deformed grains and can reach more than 100 $\mathsf{\mu }$m [9,24,28,29].

The sintered Al-30Sn composite contains pores and areas with poor adhesive bonds. The negative influence of porosity on the wear resistance of aluminum composites was also discussed in [26,27,30]. When the structure of Al-Sn composites does not contain pores and is not ordered, as, for example, after compression in the closed press mold, a significant change in the grain shape before a localization of plastic flow along the tin interlayers under a deformation is required. Such a transformation of the structure is accompanied by strengthening of the matrix grains without their cracking.

However, in the case of Al-Sn composites subjected to ECAP with route A, their structure is ordered, and plastic flow can be localized along tin interlayers during the small deformations under friction test due to the high value of shear stresses. During such a plastic shear, the macrostructure of the composites will be ordered even more, while the stress of their plastic flow will be low. This shear can be carried out until the plasticity limit of the tin interlayers is exhausted. Only junctions in the form of coherent boundaries between the aluminum matrix grains can resist the localization of such a flow.

In our case, the matrix grains form a skeleton structure, and their junctions are not fractured after the ECAP processing (Figure 2). Thin flattened Al grains in the processed composites are perpendicular to the friction surface of the sample and separated by the interlayers of soft tin. Under the dry friction test, the friction force tries to shift the tops of such grains in the sliding direction, and they easily bend due to sliding along the tin interlayers located on the grain boundaries. As a result, the matrix grains in the upper layer of the sample have the shape of comets with tails parallel to the friction surface (Figure 7). When the plasticity limit of the tin interlayers is exhausted, a delamination of the Al grains takes place, and wear particles are formed.

Their thickness is smaller than that in the case of sintered Al-30Sn composites, and the formed pits on the friction surface are also smaller (Figure 5). That is, the formation of ordered and flattened structure in the sintered Al-30Sn sample by ECAP with route A provides an increase in its ductility due to easier change in the shape of the matrix grains in the subsurface layers under the dry friction process. Hence, such grains can be subjected to higher deformations, and the wear intensity of the ECAP-ed composite is decreased.

Another important characteristic of antifriction materials is the value of the friction coefficient. This value makes it possible to divide all materials, with antifriction having $\mu$ < 0.3 and friction ones having a higher value of $\mu$. Therefore, according to Figure 4, the sintered and ECAP-ed Al-30Sn composite are antifriction materials at the beginning of the friction test, but then they very quickly become friction ones. Usually, the law of Amontons–Coulomb is used to determine the value of $\mu$. This law implies that the composition of the friction surfaces of the investigated friction pair is not changed during the measurement of $\mu$. However, it can be seen in Figure 3 that the values of $\mu$ are not constant and change every second. The growth of $\mu$ during the running-in period under the dry friction process can be caused by numerous factors.

According to the graphs in Figure 4, the number of ECAP passes practically does not affect the average value of friction coefficient during the stage of steady state friction. For example, the decrease in $\mu$ was the same ($\Delta \mu$ ≈ 0.1) after two and four ECAP passes in comparison with the value for the initial sintered sample. The value of $\Delta \mu$ is not changed with increasing sliding velocity from 0.07 to 0.6 m/s. The same weak influence of strain hardening on the value of $\mu$ was reported in [26], where the cast Al-30Sn sample processed by ECAP was subjected to a dry friction test at sliding velocity of 0.07 m/s. Hence, the weak sensitivity of $\mu$ to the sample strength can be explained if we assume that secondary structures having properties which differ significantly from that of the initial materials are formed on the friction surfaces of the coupled bodies.

In our case, a sliding distance of approximately 250–300 m is necessary for the formation of such structures. Usually, in the case of metals, most of the input mechanical energy is converted into heat, while a smaller part is in the form of elastic fields of deformation defects. Therefore, the running-in period of the Al-30Sn samples processed by the ECAP method is slightly shorter than that of the sintered ones (Figure 4). This is due to the fact that the ECAP-ed material has a higher density of subgrain boundaries and others deformation defects before the tribological test.

In order to determine the features of changes in the structure of the friction surfaces, it is necessary to analyze the graphs of the dependence of $\mu$ on the sliding distance of the sample. Three such graphs $\mu$ (L) are shown in Figure 8. Graph 1 was plotted during the passing of the Al-30Sn sample a sliding distance of 5 m along a clean friction track. Next, this sample was transferred on a new clean track and slid along it during a sliding distance of 20 m (graph 2). Then, the sample was subjected to the next test with an L of 50 m along a new clean friction track for plotting graph 3. The surface of the friction track was investigated at the end of each test.

Thus, graph 1 shows an increase in the friction force under the conditions of mutual transformation of the friction surfaces of the sample and steel counterbody. It grows relatively quickly in this case due to two parallel processes: an adhesive wear of the sample with the formation of transferred layer on the friction track surface and strengthening of the friction surface of the sample due to its deformation and oxidation. Image of the friction track surface after this period is shown in Figure 8.

It can be seen in Figure 9a that the first contacts of the counterbody with the sample surface take place in some areas where the surface irregularities have a maximum height. They perceive the main load and are pressed. The material of such irregularities is strengthened, and so-called rubs are formed. A great effort is required to overcome such rubs by sliding irregularities of the counterbody. The tops of the rubs are quickly cut off and wear out (Figure 9b), and the number of contacts between the Al-30Sn sample and the counterbody increases. As a result, stresses of the dry friction contact increase (first friction curve in Figure 8). The transfer and adhesion of wear particles to the counterbody surface also contribute to an increase in the friction force.

When this sample has contact with a new clean friction track (curve 2), the initial temperature of the sample surface is equal to the ambient temperature. Moreover, the friction surface of the sample is smoothed and oxidized. Hence, its adhesive interaction with the steel counterbody is not as active as in the previous case (curve 1). Due to the lower friction force, the temperature of the friction surface and intensity of the adhesive interaction increase more slowly. For this reason, the slope of the friction curve is lower in this case. A constant angle of the curve inclination indicates that the mechanism of the frictional interaction is the same during the test, and changes on the friction surface have an extensive character. Specifically, the density of the transferred layer increases (Figure 8c).

When the sample has contact with yet another new friction track, the aforementioned situation is repeated, and the same wear mechanism takes place. The value of the friction force is gradually increased, and a transferred layer containing small particles is formed on the friction track surface (Figure 8). However, with an increase in the duration of the friction test, large wear particles are formed on the sample surface by delamination of the highly deformed matrix grains. As a result, the friction force is rapidly increased with an increasing number of such wear particles. The number of wear particles is not constant at any given time, even at the stage of steady state friction. Therefore, the value of the friction force is also not constant and fluctuates around some average value. The frictional interaction with high-amplitude fluctuations of f begins after the passing of the sample a sliding distance of approximately 65 m (in total) along the friction track (friction curve 3 in Figure 8).

Evolution of the structure of the transferred layer can be seen in more detail in Figure 10. After several revolutions of the disk (passes of the sample), the friction surface of the Al-30Sn sample is significantly heated, and tin is melted and extruded. The oxidized friction surface of the sample is poorly wetted by liquid tin. As a result, tin is transferred on the counterbody surface (Figure 10a). Therefore, the temperature and heating depth of the friction surface increase. The amount of extruded tin also increases and reaches a maximum (Figure 10b).

The extruded tin does not spread over the surface of the friction track because the steel counterbody is colder than the sample surface and also covered with oxide films and other contamination. A portion of the transferred tin particles is mechanically smeared along the surface of the friction track and remains in pits, although most of it is removed outside the track. That is, the transferred tin quickly wears out, as was reported in [15]. Aluminum wear particles are also transferred to the friction track surface. They are then heated under the friction process, and some quantity of tin that is contained in them is extruded and located on the friction track surface in the form of thin films, which can be seen as bluish spots in Figure 10d,e. Such a uniform distribution of tin is also confirmed by its distribution map (EDX) in Figure 6c.

The described mechanism of formation of the transferred layer on the friction track is confirmed by images obtained using a scanning whitelight interference microscope (Figure 11). It can be seen that small particles (sticks) of Sn are formed after a sliding distance of 5 m (Figure 11a). The number of these particles is increased with increasing sliding distance, up to 20 m (212 revolutions of the disk). As a result, large areas covered with a thin layer of transferred tin are formed on the friction track (Figure 11b). At the same time, some small Al particles are transferred to the friction track due to adhesion between the steel counterbody and the aluminum matrix. Large hard wear particles are then formed on the Al-30Sn sample surface and begin to scratch the friction track surface. As a result, deep scratches are formed (Figure 11c). Under these conditions, only hard aluminum particles that are strongly adhered to the steel counterbody can protrude above the friction track surface. Their abrasive action leads to formation of parallel grooves on the sample surface, which can be seen in Figure 5.

Thus, it can be concluded that the wear process of the Al-30Sn sample is carried out according to the mechanism of adhesive wear at the beginning of its dry friction contact against steel. Such a wear mechanism takes place regardless of the amount of preliminary deformation of the sample.

However, it can be seen in friction curve 3 (Figure 8) that, after a total sliding distance of approximately 65 m (5 + 20 + 40), the nature of the frictional interaction between the coupled bodies begins to change from an adhesive wear to an abrasive one. As a result, high-amplitude fluctuations of $\mu$ (or f) are present in this curve. That is, the wear mechanism of the Al-30Sn sample also begins to change. This fact is confirmed by the formation of large particles on the friction track that cannot be caused by adhesive wear generating only small sticks (Figure 10).

Thus, a transferred layer is formed on the friction track surface of a steel counterbody during the dry friction test. However, it does not effectively protect the friction track from the abrasive action of hard wear particles. The relief with scratches and large transferred particles is formed on the friction track (Figure 11c). These particles gradually wear out by hard oxidized friction surface of the sample, but new ones are formed nearby. Thus, their number is almost not changed under the constant conditions of the experiment. Achieving the dynamic equilibrium density of the transferred layer on the friction track means the beginning of the stage of steady state friction, which is characterized by a quasi-constant friction force between the coupled bodies.

5. Conclusions

Features of the frictional interaction between the sintered and subjected to ECAP with route A Al-30Sn composite samples and steel counterbody under dry friction conditions were studied. The obtained results allow for the following conclusions:

• The friction surface of the Al-30Sn sample is rapidly heated to the melting point of tin under the dry friction against steel. Liquid tin inclusions are extruded and transferred to the friction track surface of the steel counterbody. As a result, tin content in the subsurface layer of the sample is significantly decreased, and its further supply on the friction surface is limited.
• A thin transferred layer of tin formed on the counterbody surface wears out and is replaced by a discrete layer, which consists of wear particles of the sample transferred to the friction track. These particles severely deform the surface layer of the sample, and it is oxidized and strengthened under such a deformation. The significant strengthening of the surface layer leads to its embrittlement and formation of hard wear particles, which, in turn, causes wear process of the transferred layer.
• Achieving a balance between the numbers of transferred and wear particles on the friction track surface means the beginning of a stage of steady state friction of the friction pair “Al-30Sn sample–steel”. During this stage, the friction coefficient and wear intensity of the Al-30Sn sample have steady state values.
• Under the low pressure of 1 MPa, the effect of tin as a solid lubricant on the process of frictional interaction of the investigated friction pair during the stage of steady state friction is insignificant. As a result, the value of friction force between the coupled bodies is high and depends on the structure of the discrete transferred layer.