**3.3 The correlation between the MML and the wear debris**

As for wear debris, its formation appeared to occur by two principal mechanisms, namely, the physical displacement of material from the worn surface by the ploughing action of the hard tool steel or alumina asperities, and secondly, delamination of large sheets (up to 1mm in extent) at particularly at high load like 140N. The thickness of the delamination sheets was found broadly consistent with the thickness of the MML, although it could not be defined with certainty whether the delamination occurred within the MML or at the MML/substrate interface. However, the longitudinal cross-sections suggested that both mechanisms were probable. Moreover, the cracks in the MML can give rise to delamination wear as a result of subsurface shear in a manner proposed by Suh (1977) where plate-like wear debris is produced. As one of the main principal wear mechanism in the present study was the delamination of the MML (part or whole), it would be reasonable to expect a correlation between MML thickness and specific wear rate.

### **3.4 The correlation between MML thickness and specific wear rate**

A detailed comparison between several commercial wrought aluminium alloys, namely; A2124, A3004, A5056 and A6092 was carried out for this purpose. For this Al/M2 system, the specific wear rate was relatively insensitive to MML thickness for the A3004 and A5056, although the specific wear rate decreased in a linear manner with increasing MML thickness (refer to Fig. 4a).

In contrast, for the A2124 and particularly the A6092, the specific wear rate was a strong function of the MML thickness. Although a reasonable linear fit was possible for the A2124,

dispersed into the mixed surface layer and act as a pinning source of the grain boundaries in

Fig. 3. Subsurface damage of longitudinal cross sections of A2124 alloys against M2 at 140N after sliding 10.8 km. Black arrows indicate the direction of sliding. A corresponding EDS

As for wear debris, its formation appeared to occur by two principal mechanisms, namely, the physical displacement of material from the worn surface by the ploughing action of the hard tool steel or alumina asperities, and secondly, delamination of large sheets (up to 1mm in extent) at particularly at high load like 140N. The thickness of the delamination sheets was found broadly consistent with the thickness of the MML, although it could not be defined with certainty whether the delamination occurred within the MML or at the MML/substrate interface. However, the longitudinal cross-sections suggested that both mechanisms were probable. Moreover, the cracks in the MML can give rise to delamination wear as a result of subsurface shear in a manner proposed by Suh (1977) where plate-like wear debris is produced. As one of the main principal wear mechanism in the present study was the delamination of the MML (part or whole), it would be reasonable to expect a

A detailed comparison between several commercial wrought aluminium alloys, namely; A2124, A3004, A5056 and A6092 was carried out for this purpose. For this Al/M2 system, the specific wear rate was relatively insensitive to MML thickness for the A3004 and A5056, although the specific wear rate decreased in a linear manner with increasing MML thickness

In contrast, for the A2124 and particularly the A6092, the specific wear rate was a strong function of the MML thickness. Although a reasonable linear fit was possible for the A2124,

analysis are shown in A, B and C areas (Ghazali, 2005)

**3.3 The correlation between the MML and the wear debris** 

correlation between MML thickness and specific wear rate.

(refer to Fig. 4a).

**3.4 The correlation between MML thickness and specific wear rate** 

the ultrafine mixture in the MML and in the wear debris (Li and Tandon, 1999).

Fig. 4. The relationship between MML thickness and (a) wear rate and (b) load after 10.8km against M2 counterface (Ghazali, 2005)

Effects of Dry Sliding Wear of Wrought Al-Alloys on Mechanical Mixed Layers (MML) 325

case of Al-alloys, it is readily combined with oxygen to form a stable oxide layer. Oxidation, may have opposing effects on the wear process; one, it degraded the surface by removing metal atoms and second, it plays protective role in reducing metallic contact and decrease the wear rate (Degnan, 1995). However, whether or not the environment reaction has a beneficial and detrimental effect on wear rate, it depends strongly on the mechanical interaction of the reaction product with the substrate, particularly under surface plasticity

Moisture in the environment also can have major effect on wear of metals. Endo and Goto (1978) reported the high humidity had a dentrimental effect on the fretting of aluminium alloys but negligible on carbon steels. Moreover, humidity can control the friction at room temperatures, particularly ceramics as higher coefficient of friction may occurr at

Beside humidity, all wear processes are influenced by temperature. The temperature reached at the surface of the contact is strongly influenced by the width of the contact (Johnson, 1985) and flash temperature is responsible for many wear and friction effects (Gecim and Winer, 1986). Wear occurs in conjunction with the dissipation of frictional energy in the contact and this is always accompanied by a rise in temperature. The frictional energy is generated by the combination of load and sliding speed and its distribution and dissipation is influenced by other contact conditions such as size and relative velocity. In regards to temperature effects on sample size and mass, contact spots have a tendency to remain in one place much longer on the smaller pin (alloy) than the larger side (counterface), causing stronger local heating in the former. Moreover, in this work, the rotating counterface will mostly experience extra cooling convention than the stationary alloy, which constantly hot due to repeated passage during the test. Local heating at contact spots also has other effect. Most obviously, the local hardness is reduced and thereby the

At high loads like the one used in the present study, 140N, friction heating can induce an increase in temperature, resulting a thermal softening beneath the worn surface, and may affect the wear mechanism (Zhang and Alpas, 1997 and Wang and Rack 1991). Maupin et al., (1992, 1993) studied that large grains were replaced by fine nanocrystalline grains which were relatively free of dislocations underneath the worn surface. Such microstructures could develop only if the temperature of the surface due to friction is very high. In addition, the deformed layer beneath the worn surface could result in higher plastic flow and work hardening resulting in increased wear resistance. At such high temperature, oxidation of the

In general, the dry sliding of Al/M2 systems showed the following responses as a result of

• Elements present in the Al-alloy with high solubility in steel promoted a thick mechanically mixed layer, with higher Fe content. The effect was marked even for small

• The solubility of these elements in α-Fe is in the order of Si, Mn, Cu, Mg, which roughly

condition, (Rainforth et al., 2002), which is in line with the present work.

load-bearing area is increased (Kuhlmann-Wilsdorf, 1987).

surface is also a possibility, as observed in earlier results.

approximates the thickness of the MML formed.

repeated stress and frictional heat cycle:

contents in the Al-alloy.

temperatures above 800°C.

**4. Conclusion** 

the A6092 data was better represented by an exponential fit. Since this data did not fit the same trend as the other alloys, the experiment was repeated and measurements re-taken, but with essentially the same result. Thus, the difference in behaviour of this alloy appears to be reproducible. Interestingly, the two alloys where the specific wear rate was relatively insensitive to MML thickness also exhibited MML with the least Fe content (Table 2) and the most homogeneous structure. Conversely, the A6092 exhibited the highest Fe content, the most heterogeneous structure and the greatest influence on the specific wear rate. However, the thickness of the MML cannot explain the dramatic drop in specific wear rate with load observed for the A3004 alloy (Fig. 4b). The thickness of the MML is only one of several potential ways in which the MML can affect wear rate. Clearly for Al/M2 system, the mechanical properties of the MML (in particular hardness and fracture stress) and its adhesion to the substrate are contributing factors.


Table 2. Average quantitative EDS analysis on MML of Al-alloys against M2 (Ghazali, 2005)

In Al/M2 system, the A6092 exhibited the thickest MML and the highest Fe content. Since this was not replicated by the A5056, rather the reverse, it is clear that it is not the Mg content of the A6092 that promotes the formation of a thick MML. Thus, these results imply that stronger adhesion and transfer from the counterface is promoted by the Si in the alloy, while a high Mg content in the Al-alloy reduces adhesion. Similarly, the presence of Cu in the A2124 also appears to have promoted stronger adhesion than an equivalent amount of Mg, although the Cu was not as potent as the Si. The Mn in the A3004 also promoted a relatively thick MML, but one that was more homogeneous than for the A6092. The solubility of these elements in α-Al is in the order Si, Mn, Cu, Mg, which roughly approximates to the thickness of the MML formed. Thus, the observations are in-line with the Archard theory of adhesive wear, as might be expected. However, the level of alloy additions are small (e.g. Si) and it is surprising that the effect was as strong as observed. Thus, the wear performance is largely determined by the properties of the MML.

### **3.5 Effect of other variables**

The atmosphere under an unlubricated wear process can strongly influence sliding wear rates with oxygen content and humidity being probably one of the important factors. In the case of Al-alloys, it is readily combined with oxygen to form a stable oxide layer. Oxidation, may have opposing effects on the wear process; one, it degraded the surface by removing metal atoms and second, it plays protective role in reducing metallic contact and decrease the wear rate (Degnan, 1995). However, whether or not the environment reaction has a beneficial and detrimental effect on wear rate, it depends strongly on the mechanical interaction of the reaction product with the substrate, particularly under surface plasticity condition, (Rainforth et al., 2002), which is in line with the present work.

Moisture in the environment also can have major effect on wear of metals. Endo and Goto (1978) reported the high humidity had a dentrimental effect on the fretting of aluminium alloys but negligible on carbon steels. Moreover, humidity can control the friction at room temperatures, particularly ceramics as higher coefficient of friction may occurr at temperatures above 800°C.

Beside humidity, all wear processes are influenced by temperature. The temperature reached at the surface of the contact is strongly influenced by the width of the contact (Johnson, 1985) and flash temperature is responsible for many wear and friction effects (Gecim and Winer, 1986). Wear occurs in conjunction with the dissipation of frictional energy in the contact and this is always accompanied by a rise in temperature. The frictional energy is generated by the combination of load and sliding speed and its distribution and dissipation is influenced by other contact conditions such as size and relative velocity. In regards to temperature effects on sample size and mass, contact spots have a tendency to remain in one place much longer on the smaller pin (alloy) than the larger side (counterface), causing stronger local heating in the former. Moreover, in this work, the rotating counterface will mostly experience extra cooling convention than the stationary alloy, which constantly hot due to repeated passage during the test. Local heating at contact spots also has other effect. Most obviously, the local hardness is reduced and thereby the load-bearing area is increased (Kuhlmann-Wilsdorf, 1987).

At high loads like the one used in the present study, 140N, friction heating can induce an increase in temperature, resulting a thermal softening beneath the worn surface, and may affect the wear mechanism (Zhang and Alpas, 1997 and Wang and Rack 1991). Maupin et al., (1992, 1993) studied that large grains were replaced by fine nanocrystalline grains which were relatively free of dislocations underneath the worn surface. Such microstructures could develop only if the temperature of the surface due to friction is very high. In addition, the deformed layer beneath the worn surface could result in higher plastic flow and work hardening resulting in increased wear resistance. At such high temperature, oxidation of the surface is also a possibility, as observed in earlier results.
