A study on tribological characterization of Al-Cu-Mg-B composites subjected to mechanical wear

2.1 Materials selection and processing

The composite fabricated consisted of a matrix made of a ternary Al-Cu-Mg alloy reinforced with hard aluminum diboride particles. In the production of the composites through gravity casting, three commercial master alloys were used: Al-5 wt% B, Al-10 wt% Mg, and Al-33 wt% Cu. Magnesium and copper were used as alloying elements because of their high solubility in the Al solid solution [5], which contributes greatly to the overall wear resistance of the Al-B-Cu-Mg composites. In each composite the aluminum matrix was composed of Al-2.5 wt% Cu-1 wt% Mg. The AlB₂ particles (contained in the Al-B master alloy) were used as reinforcement for the composite matrix, with the amount of boron ranging from 0 to 4 wt%. The size of the AlB₂ particles, the volume percent of particles, and volume percent of pores were studied in a prior research on AMCs with the same composite compositions [6]. These AlB₂ particles ranged in size from 1 to 20 μm. The volume percent of particles increases with the weight percent of boron, varying from 1.4% to 16%. The same tendency is shown in the volume percent of pores that varies from 0.37% to 2.30% with the increase in the amount of boron [6].

The charge material was mixed and melted in an electrical furnace at 750°C to avoid the transformation from AlB₂ to AlB₁₂ [2, 7, 8]. In the process, AlB₂ particles tend to be at the bottom of the graphite crucible because of higher density (3.2 g/cm³) than liquid aluminum (∼2.4 g/cm³) at the furnace operating temperature. Therefore, to attain more homogeneity of the samples, mechanical agitation was applied with a stirring rod, while the molten composite, containing solid boride particles, was poured into a cylindrical preheated graphite mold. Graphite was used because of its high thermal conductivity to favor high solidification rates and smaller grains.

After solidification, the 2-inch-diameter ingots were annealed at 350°C for 4 h in an electrical furnace. The power of the furnace was turned off to allow the ingots to cool to room temperature. The purpose of the annealing treatment was to homogenize the chemical composition of the matrix and to stabilize the microstructure of the fabricated composites, avoiding the influence of the stress concentrations due to uneven cooling rates in the castings [6]. Al-Cu-Mg matrix composites could have been strengthened by means of an aging treatment. However, samples were annealed only because we wanted to study the wear rate solely as a function of their chemical composition and not their heat treatability.

The ingots were sectioned with a low-speed saw and disk specimens were obtained. The specimens were mounted, ground, and polished to produce a flat, scratch-free surface for metallographic analysis. The grinding steps included 320, 400, 600, 800 grit SiC papers. The samples were polished using 3-μm diamond suspension followed by a 0.05-μm SiO₂ emulsion. Finally, the samples were cleaned with distilled water and alcohol and dried to be characterized.

2.2 Materials characterization

The composite microstructure was investigated using optical microscopy (OM), scanning electron microscopy (SEM), and energy-dispersive spectroscopy (EDS). For OM, the samples were observed using a Nikon (Melville, NY, USA) Epiphot 200 inverted optical microscope, and the SEM analysis was performed with a JEOL (Peabody, MA, USA) 6390 scanning electron microscope.

Dry sliding wear test was conducted using a pin-on-disk tester where the composite samples were the disks and a 440 martensitic stainless steel ball with a diameter of 3 mm was the indenting pin. The setup consisted of the indenting pin resting on the composite sample disk, rotating under the influence of a load. The apparatus was constructed in our research laboratory on the basis of specifications of the ASTM G99 standard. Low contact loads and small sliding velocity were selected to offset excessive heating in the samples. The test was conducted twice at each condition with a 1 N constant load at a 0.004-m/s sliding speed and a total sliding distance of 2.5 m, using a new pin for each test. The worn surfaces and wear debris were analyzed using SEM (equipped with an EDS system) and OM. Optical profilometry using a Zygo (Middlefield, CT, USA) New View 6300 profilometer system allowed determination of the geometry (depth and width) of the pin-on-disk tracks, which permitted us to determine the material volume removed due to wear.

3.1 Microstructure analysis

Microstructure analysis of the composites without etching allowed the identification of the phases, using the microstructure analyses of previous research with similar compositions as a reference. A homogeneous distribution of particles was obtained and the thermodynamically stable θ phase was identified [6]. The five optical micrographs in Figure 1 evince the distribution of reinforcement particles in the matrix as boron level increases in the composite from 0 to 4 wt%. These micrographs show diborides embedded in the aluminum matrix that appear as darker particles and distributed in clusters throughout the entire matrix [9]. In Figure 1A, the intermetallic phase, Al₂Cu (θ), homogenized through annealing, is evident throughout the matrix but mainly in the matrix grain boundaries. The metastable θ′ is not present because aging was not performed in these specimens [10].

Figure 1

Optical micrographs showing the AlB₂ distribution as boron levels increase from 0 to 4 wt%: (A) Al-2.5 wt% Cu-1 wt% Mg-0 wt% B, (B) Al-2.5 wt% Cu-1 wt% Mg-1 wt% B, (C) Al-2.5 wt% Cu-1 wt% Mg-2 wt% B, (D) Al-2.5 wt% Cu-1 wt% Mg-3 wt% B, (E) Al-2.5 wt% Cu-1 wt% Mg-4 wt% B.

Then, the samples were observed with a SEM equipped with an EDS system (Figure 2). It should be noted that the boron Kα peak was at the fringe of the EDS detection limit (low-energy side of the EDS spectrum). However, the microstructure analysis revealed the presence of boron in the composites. Figure 2A shows a secondary electron image of the microstructure of Al-2.5 wt% Cu-1 wt% Mg-2 wt% B composite, where an elliptical cluster of AlB₂ is apparent. Figure 2B shows a global EDS spectrum revealing the presence of copper, magnesium, and aluminum in the Al-2.5 wt% Cu-1 wt% Mg-2 wt% B composite.

Figure 2

(A) SEM image of Al-2.5 wt% Cu-1 wt% Mg-2 wt% B, (B) Global EDS spectrum of the worn surface detected in Al-2.5 wt% Cu-1 wt% Mg-2 wt% B composite.

3.2 Wear characteristics

Tribological experiments were focused on the determination of wear measured by the material volume removed, the wear coefficient value, and the material removal rates. The calculation of wear volume on each sample was based on topographical analysis via optical profilometry techniques, taking into consideration the track width, depth, and perimeter. Figure 3 evinces the differences in wear volume among samples. Noticeably, for higher boron concentrations (i.e., more AlB₂ particles), the wear volume decreases. Moreover, the wear volume of the unreinforced alloy is greater than that of the composites containing diboride particles.

Figure 3

Wear volume measured using optical profilometry as a function of boron percent.

For all of the composites under the same testing conditions, we calculated the values of the wear coefficient based on the worn volume obtained with the track geometry (maximum depths and widths), as measured by optical profilometry. The wear coefficient was evaluated using Eq. (1) proposed by Archard [11], where the wear volume (V) is directly proportional to the sliding distance (d) and the applied normal force (F), by the dimensional wear coefficient (k), conventionally in units of mm³/N·m. The resulting experimental wear coefficients are presented in Figure 4 as a function of boron percent. All the composites exhibited similar behavior with increasing content of AlB₂, where the wear coefficient decreases in the composites, obtaining higher wear resistance. This behavior is visibly related to the wear volume measured.

Figure 4

Wear coefficient measured on the tested samples as a function of boron content in the composites.

Archard’s wear equation can be modified to Preston’s equation [12–15] to calculate the material removal rate (RR) or dV/dt. According to Preston’s equation [Eq. (2)], the relative velocity (v_r) between the spherical pin and the specimen surface, the applied normal load (F), and the time (t) are the main parameters determining the amount of material removal (dV). Table 1 shows the variation of RR as a function of boron levels, showing the same tendencies as in wear volume.

3.3 Microscopy analysis of worn surfaces

SEM examination revealed detailed information on the wear tracks and debris, pointing to the formation of parallel scratches with plastic deformation edges. For instance, Figure 5 showed that the pin-on-disk wear test carried out on the Al-Cu-Mg-B composites induced plastic deformation, ploughing, and deposition of debris onto the surface. These features were accompanied by many grooves along the sliding direction, while the peeling of the matrix formed worn chips that adhered to the composite surface.

Figure 5

SEM micrographs of pin-on-disk worn surface on Al-2.5 wt% Cu-1 wt% Mg-0 wt% B composite.

SEM images reveal that AlB₂ particles are uniformly distributed throughout the matrix alloy. In our SEM observations coupled with EDS analysis, we discovered that the diboride particles hide themselves under the worn surface [16]. Other investigators have proposed that particle reinforcements are the most effective in improving the wear resistance of AMCs [17] by providing good interfacial bonding between the reinforcement and the matrix. The resistance to plastic flow becomes visible at places where clusters of AlB₂ are located, as seen in Figure 6. Figure 7 shows EDS spectrum analysis of the wear track where the presence of Fe was absent, resulting in no evidence of adhesive wear. The EDS analysis proved the nonappearance of debris coming from the pin, i.e., stainless steel, as no adhered pin material was found on the wear track, confirming the discovery in a previous research [18].

Figure 6

SEM images showing evidence of resistance to plastic flow of Al-2.5 wt%Cu-1 wt% Mg-1 wt% B composite.

Figure 7

EDS spectrum analysis of wear track: (A) Al-2.5% Cu-1% Mg-0% B composites, (B) Al-2.5% Cu-1% Mg-3% B composite.

Through further inspection of the SEM images, diverse wear mechanisms were identified: oxidative wear, abrasion wear, and delamination wear. Figure 8, A and B, show oxygen concentration of an X-ray mapping analysis on Al-2.5% Cu-1% Mg-0% B and Al-2.5% Cu-1% Mg-3% B wear tracks, respectively. These characteristics, along with the oxygen peaks present in the EDS spectrum shown in Figure 7, indicate an oxidative wear mechanism, in which unlubricated conditions of sliding cause relatively high “hot-spot” temperatures at the surface, producing oxidation [19]. Figure 9 corroborates the abrasive wear mechanism by illustrating the appearance of grooves and damage spots forming ploughs on the worn surface. Evidence of delamination wear on the composite surface can be seen in Figure 10, showing the presence of induced cracks when repeated and constant sliding introduces dislocations in the subsurface by plastic deformation that eventually shears the surface and forms long thin flake debris [20]. Furthermore, the wear tracks showed a change in the dominant wear mechanism present in the composite from abrasive wear to delamination wear when the level of boron rises from 3 to 4 wt%, as evidenced by the flake debris shown with arrows in Figure 11.

Figure 8

SEM micrographs and oxygen elemental map of: (A) Al-2.5% Cu-1% Mg-0% B, (B) Al-2.5% Cu-1% Mg-3% B.

Figure 9

SEM micrograph evidence of grooves and ploughs due to abrasive wear: (A) Al-2.5% Cu-1% Mg-1% B composite and (B) Al-2.5% Cu-1% Mg-3% B composite.

Figure 10

SEM micrographs showing induced cracks (arrows): (A) Al-2.5% Cu-1% Mg-1% B composite, (B) Al-2.5% Cu-1% Mg-3% B composites.

Figure 11

SEM showing evidence of delamination wear damage (arrows) in Al-2.5 wt% Cu-1 wt% Mg-3 wt% B composite.