Sliding temperature and wear behaviour of cast Al–Si–Mg alloys

Sliding temperature and wear behaviour of cast Al–Si–Mg alloys

Materials Science and Engineering A 382 (2004) 328–334 Sliding temperature and wear behaviour of cast Al–Si–Mg alloys Dheerendra Kumar Dwivedi∗ Mecha...

517KB Sizes 0 Downloads 29 Views

Materials Science and Engineering A 382 (2004) 328–334

Sliding temperature and wear behaviour of cast Al–Si–Mg alloys Dheerendra Kumar Dwivedi∗ Mechanical Engineering Department, National Institute of Technology, Hamirpur (HP) 177005, India Received 9 January 2004; received in revised form 28 April 2004

Abstract In present paper the influence of sliding interface temperature on friction and wear behaviour of cast Al–(4–20%) Si–0.3% Mg has been reported. Wear and friction tests were performed under dry sliding conditions using pin on disc type of friction and wear monitor with the data acquisition system conforming to ASTM G99 standard. It was found that sliding interface temperature has close relation with wear and friction response of these alloys. Initial rise in interface temperature reduces the wear rate and as soon as a critical temperature (CT) is crossed, wear rate abruptly increases in case of all the compositions used in this investigation. Friction coefficient during the sliding of all aluminium alloys (irrespective of silicon content) first decreases with the rise in interface temperature and then abruptly increases beyond certain critical temperature. Critical temperature was found to be a function of alloy composition, i.e. silicon content. Hypoeutectic alloys showed lower critical temperature than the hypereutectic alloys. © 2004 Elsevier B.V. All rights reserved. Keywords: Cast Al–Si alloys; Sliding wear; Silicon %; Microstructure; Interface temperature; Wear rate; Sliding conditions

1. Introduction Relative motion between the two sliding surfaces generates frictional heat as most of the work done against friction is converted into heat, which in turn increases the interface temperature [1,2]. Rise in interface temperature affects the metallic intimacy between the sliding surfaces in two ways: (1) increase the contaminations at the sliding surface due to oxidation, and (2) adversely affecting the mechanical properties of sliding surfaces. Oxide layer reduces the metal to metal contact between the sliding surfaces. Whereas thermal softening of surface and near surface layers increases the metallic intimacy due to deformation of peaks at the surface under external load. These two aspects significantly influence the wear response of most of the metallic materials [3–6]. Influence of frictional heating on tribological behaviour and failure of sliding components, the surface and near surface temperature has been an attractive topic for many years. Unfortunately, it is very difficult to obtain the sliding surface temperature or temperature field during the

Tel.: +91-1972-222530. E-mail address: [email protected] (D.K. Dwivedi).

0921-5093/$ – see front matter © 2004 Elsevier B.V. All rights reserved. doi:10.1016/j.msea.2004.05.014

sliding by means of experimental methods. It has been reported that transition in wear mechanism has close relationship with temperature gradient in sub-surface region [1,3]. The effect of sliding temperature on wear behaviour of two multi-components Al–Si alloys (LM13 and LM28) in as cast and heat treated conditions was published earlier [4,5]. It was found that there is a critical temperature for each alloy at which wear rate suddenly increases and transition from mild to severe wear takes place. Maustafa [7] have also reported that there is a critical temperature at which mode of wear suddenly changes from mild to severe. However, there are not many reports, which relate the wear mechanism to the change in interface temperature for cast Al–Si alloy with varying silicon percentage. In general, there are two types of adhesive wear namely mild and severe wear. Transitions from one wear regime to another takes place as contact load or speed is gradually changed, and occurs quite abruptly [4]. The size, shape, micro-cracking tendency and thermal stability of different micro constituents greatly control the mechanical and tribological properties of Al–Si alloys [8–10]. In the present work, variation of wear rate and friction coefficient with interface temperature rise and silicon content have been investigated.

D.K. Dwivedi / Materials Science and Engineering A 382 (2004) 328–334


Table 1 Mechanical test results Alloy

Al–4% Si

Al–8% Si

Al–12% Si

Al–16% Si

Al–20% Si

Tensile strength [MPa] Hardness [VHN]

134 70

189 84

210 98

180 125

160 138

2. Experimental procedure 2.1. Material Experimental alloys were prepared by controlled melting of high purity aluminium, Al–28% Si and Al–10% Mg master alloys in a graphite crucible using a muffle furnace and cast in metallic mould of size 25 mm × 25 mm × 150 mm. Nominal compositions of experimental alloys are shown in Table 1. Hypoeutectic alloys were modified by addition of 0.015% strontium, in form of Al–10% Sr and hypereutectic alloys were modified by addition of 0.01% P in form of red phosphorous. Wear test pins (cylindrical) of 6 mm diameter and 20 mm length were prepared by turning and one end of pin was polished. 2.2. Friction and wear behaviour A pin on disc type wear monitor (DUCOM, TL-20, Bangalore) with data acquisition system was used to evaluate the wear friction behaviour of aluminium alloys against hardened ground steel (En-31) disc having hardness of RC60 and surface roughness (Ra ) 0.5 ␮m. A schematic diagram is shown in Fig. 1. Load was applied on pin by dead weight through pulley string arrangement. The system had maximum loading capacity of 200 N. Disc was rotated by dc motor, having speed range of 0–2000 rev min−1 to yield sliding speed 0–10 m/s. Weight loss was used as a measure of wear. Weight loss was measured after 50, 100, 200, 500, 1000, 1500 and 2000 m sliding distances. Wear rate was calculated using weight loss per unit sliding distance (g/m). Counter surface was abraded against carbide polishing papers and cleaned with acetone and dried before each sliding test. Friction and wear data were acquired at rate of 5 samples/s during the 60 min of sliding at various velocities 0.3, 1.0, 2.0, 3.0, 4.0, 5, and 6 m/s and constant normal load of order of 30 N. Disc revolution per minute (at 80 mm track diameter)

Fig. 1. Schematic diagram of pin on disc wear and friction monitor.

was varied to change the sliding speed. Sliding conditions when a lot of vibration, noise and gross metal transfer during the sliding take place, has been considered as severe metallic wear conditions. The friction force was recorded during the experiment by using a load cell (accuracy 0.1 N and capacity 200 N). 2.3. Temperature measurement Temperature measurements of wear pin during the sliding were carried out with three chromel–alumel thermocouples. These thermocouples were placed into the holes each of 1 mm diameter at 1.5, 3.0 and 4.5 mm away from sliding surface drilled up at axis of cylindrical wear pin. To find out the temperature exactly at the sliding surface extrapolation method was used. This sliding surface temperature has been referred as interface temperature in the forgoing section. Temperatures were also noted with help of digital temperature indicators after 50, 100, 500, 1000, 1500, and 2000 m of sliding distance. Temperature at sliding surface can be important in explaining the friction and wear because mechanical properties of material and surface oxidation are affected by temperature. 2.4. Mechanical test Hounsfield computerized tensile testing machine (20 kN) was used to carry out tensile test at 1 mm/min strain rate for tensile strength and ductility by using diameter 6.2 mm and gauge length 23.2 mm tensile sample. Vickers hardness test was performed at 5 kg load to study the effect of silicon content on hardness and relate it with wear behaviour.

3. Results 3.1. Microstructure study Optical microphotographs of cast Al–(4–20)% Si alloys are shown in Fig. 2(a–e). It is observed that the primary ␣-aluminium grains are predominantly present in structure of cast Al–4% Si (Fig. 2a). Little amount of eutectic silicon can also be seen in the inter-dendritic region. Optical microphotograph of cast Al–8% Si alloy shows that primary ␣-aluminium grains are embed in the eutectic (Fig. 2b). Eutectic silicon can be seen in the inter-dendritic region. It is observed that the addition of silicon has increased the eutectic amount present in the inter-dendritic region. Al–12%


D.K. Dwivedi / Materials Science and Engineering A 382 (2004) 328–334

Fig. 2. Optical microphotographs of as cast: (a) Al–4% Si, (b) Al–8% Si, (c) Al–12% Si, (d) Al–16% Si, and (e) Al–20% Si alloy (200×).

Si alloy also shows that primary ␣-aluminium grains are present in the inter-dendritic region (Fig. 2c). Moreover, addition of silicon increases the amount of eutectic present along the grain boundaries. Optical microphotograph of cast Al–16% Si alloy shows coarse polyhedral shaped primary silicon crystals in matrix of eutectic (Fig. 2d). Increase in silicon content from 16 to 20% increases the volume fraction of primary silicon crystals rest of the features are not significantly affected (Fig. 2e). It is expected that increase in the proportion of eutectic and primary silicon crystals with the addition of silicon would increase the hardness and affect the wear behaviour.

3.2. Interface temperature Typical temperature variation during the sliding of cast Al–20% Si alloys with sliding distance at different sliding speeds and 30 N load is shown in Fig. 3a. Figure shows that there are at the most three trends of temperature variation with sliding distance in the entire range of sliding speeds used in the work. At low sliding speeds two regimes of temperature variation with sliding distance were observed whereas at high sliding speed three regimes of temperature variation were observed. Out of the two regimes of temperature variation with sliding distance at low speeds, the first

D.K. Dwivedi / Materials Science and Engineering A 382 (2004) 328–334


220 0.3m/s 2.0m/s 4.0m/s 6.0m/s

Temperature [oC]


1.0m/s 3.0m/s 5.0m/s

140 100 60 20 0



Interface Temperature [ C]

(a) 180


Fig. 5. Wear rate vs. interface temperature relationship at constant load 3 kgf and varying sliding speed condition for different cast Al–Si alloys.

Al-4%Si Al-8%Si Al-12%Si Al-16%Si Al-20%Si

140 100 60 20 0.3


800 1200 1600 Sliding Distance [m]







Sliding speed [m/sec]

Fig. 3. (a) Interface temperature vs. sliding distance relationship for cast Al–20% Si alloys at different sliding speeds. (b) Interface temperature vs. sliding speed relationship for different cast Al–Si alloys.

one corresponds to initial steep rise in temperature during the run in period and the second one corresponds to the steady state sliding shows constancy in temperature. It is similar to earlier studies [4,8,9]. Variation in interface temperature (after 1000 m of sliding or before seizure) with sliding speed for cast Al–(4–20)% Si is shown in Fig. 3b. It is observed that increase in sliding speed increases the interface temperature irrespective of alloy composition. In the low speed range (depending upon the alloy composition) the relationship between interface temperature and sliding speed is almost linear. Above certain sliding speed (i.e. transition speed) abrupt rise in interface temperature takes place after traverse of some sliding distance. Transition speed is also a function of alloy composi-

tion, i.e. silicon content. Increase in silicon content increases the transition speed. In the low speed range (0.3–2.0 m/s) silicon content does not affect the interface temperature appreciably. Moreover, minor reduction in the interface temperature was noticed with increasing silicon content. Abrupt rise in temperature takes place as soon as seizure like conditions are attained in all Al–(4–20)% Si alloys. As and when temperature reaches a critical value depending upon the alloy composition transfer of pin material to disc by melting/fusion starts and gross metal transfer takes place in very short distance of sliding. Increase in silicon content increases the temperature at which this gross metal transfer takes place. 3.3. Friction behaviour Variation of friction coefficient with the sliding temperature for cast Al–Si is shown in Fig. 4. It appears that friction coefficient in all the cases initially decreases with increase in sliding temperature and beyond certain critical temperature it increases. Moreover, temperature above which rise in friction coefficient takes place increases with silicon content. Silicon addition lowers the friction coefficient under identical sliding condition. 3.4. Wear behaviour

Fig. 4. Friction coefficient vs. interface temperature relationship at constant load 3 kgf and varying sliding speed condition for different cast Al–Si alloys.

Interface temperature significantly affects the wear behaviour of cast Al–Si alloys (Fig. 5). It is observed that wear rate of all the alloys decreases initially with increase in temperature up to certain temperature and thereafter it increases abruptly. Reduction in wear rate may be attributed to oxidation of sliding surface with increase in temperature. The temperature above which sudden increase in wear rate takes place can be termed as critical temperature (CT). It appears that silicon content has dominating effect on the critical temperature corresponding to the transition in wear. Increase in silicon content increases the critical temperature. Critical temperature for hypoeutectic Al–4% Si alloy (70 ◦ C) was lower than that for hypereutectic Al–20% Si alloy (200 ◦ C).


D.K. Dwivedi / Materials Science and Engineering A 382 (2004) 328–334

Fig. 6. SEM images of wear debris of generated on sliding: (a) Al–4% Si alloy at 0.3 m/s, (b) Al–4% Si alloy at 3.0 m/s, (c) Al–20% Si alloy at 0.3 m/s sliding speed, and (d) Al–20% Si alloy at 6.0 m/s.

3.5. SEM study Wear debris study under SEM has shown that at low sliding speeds, i.e. 0.3 m/s, comparatively small laminated particles are produced after sliding of Al–4% Si and Al–20% Si alloys indicating the occurrence of mild wear (Fig. 6a and b). At high sliding speed (3.0 m/s) Al–4% Si alloy generates long metallic strings, which indicate the occurrence of severe metallic wear (Fig. 6c) whereas metallic debris with Al–20% Si alloy was generated during the sliding at 6 m/s speed only (Fig. 6d). Scanning electron microscopy (Fig. 7a) of wear surface of Al–20% Si alloy after sliding speeds at 0.3 m/s reveals crater, oxidized surface, delamination and scoring are the main mechanisms responsible for material loss indicating the occurrence of mild oxidative wear while in severe wear region gross plastic deformation and damages to surface can

be notice on wear surface after sliding of and Al–20% alloys at 6 m/s speeds (Fig. 7b).

4. Discussion Archard [11] classified the wear broadly into two categories, i.e. mild wear and severe wear. Torabian et al. [12] have reported three regimes of wear, i.e. low, mild and severe wear, whereas So [13] have noticed mild oxidative, mild-cum-severe (mixed), severe wear. Adhesive wear of aluminium alloys under dry sliding conditions takes place by oxidation, metallic failures or combination of two [14,15]. Sliding conditions such as sliding speed, load and counter surface primarily governs active wear mechanism at any stage. Under low load and sliding speed conditions when frictional heating does not cause significant rise in the inter-

Critical Temperature [ 0C]

D.K. Dwivedi / Materials Science and Engineering A 382 (2004) 328–334


190 170 150 130 110 90 70 50 4





Silicon [%]

Fig. 8. Influence of silicon on critical temperature.

Fig. 7. SEM microphotographs of worn out surface of Al–20% Si alloy after sliding at (a) 0.3 m/s, and (b) 6.0 m/s sliding speed.

face temperature of wear pin, removal of material is mainly controlled by formation of oxide on the sliding surface and its fracture due to surface traction forces [16]. Therefore, rate of wear depends on oxidation and fracture of these of oxides from sliding surface. Analysis of worn surface and debris produced after the sliding test under these sliding conditions (when interface temperature remains stable) reveals that the mode of wear is mild oxidative (Fig. 7a). Increase in sliding speed increases the steady state interface temperature (Fig 4b). If interface temperature in steady state remains more or less constant during the sliding, weight loss occurs at constant rate (Fig. 4a). As interface temperature crosses a critical value weight loss suddenly increases. This is due to the softening of material in the subsurface region. The high temperature generated at the interface causes severe plastic deformation (Fig. 7b). Under these conditions, surface layer is destabilized and severe metallic wear takes place, which causes the abrupt increase in weight loss. Das et al. [9] have also reported a similar behaviour. Below critical temperature, loss

of material takes place primarily due to oxidation. This is due to the fact that below critical temperature heat generated during the sliding at the interface is not sufficient to cause plastic deformation (Fig. 7a). It appears that the rise in temperature of sliding surface due to frictional heating controls the wear mechanism and wear rates. Transition from mild oxidative to severe metallic wear is due to two opposing dynamic processes, i.e. rate of formation of, (1) surface oxide film caused, and (2) the exposure of fresh metal surface during the sliding. If the first process dominates mild oxidative wear takes place, while the domination of second one causes severe metallic wear [16,17]. Increase in silicon percentage increases the resistance to thermal softening and ability to support the surface oxide film owing to higher hardness. Similar observations were also made by Sarkar and Clarck [16], Davis and Eyre [17], Razavizadeh and Eyre [18]. Razavizadeh et al. [65] have reported that the transition load increases in direct proportion to silicon content (up to 20%) in cast aluminium alloys. Silicon reduces the tendency to form adhesion welds with the counterface [19,20]. It can be seen that a lower temperature corresponds to mild wear and high temperature corresponds to severe wear. Transition in wear mechanism is a result of joint action of applied load and frictional heat. There should be no doubt that when temperature exceeds a value, it reduces the flow strength/hardness of material and therefore increases the wear rate and cause transition from mild to severe wear. This limiting temperature may be termed as critical temperature (CT). Accordingly mild and severe can be called low and high heat effects, respectively. Below the critical temperature change in wear rate is small with the temperature variation. Properties of sliding metal surface control the friction and wear behaviour, therefore mechanical properties, microstructure, composition, work hardening characteristics, thermal stability of phases related with the cast Al–Si base alloy are important. Shift in transition point towards the high temperature/speed with an increase of silicon content (Fig. 8) indicates the dependence of transition point on composition and may be attributed to increased stability of subsurface material and oxide layer [4,8]. Increased presence of primary silicon particles in hypereutectic alloys may resist


D.K. Dwivedi / Materials Science and Engineering A 382 (2004) 328–334

the gross plastic deformation, which is needed for metallic wear [9]. Friction coefficient beyond critical temperature increases possibly because of gross plastic deformation of surface layer and metal transfer to the counter face in form of visible patches. Large-scale plastic deformation of surface layer may increase the actual area of contact. Metal transfer to the counter surface creates a situation of sliding between similar metal. Therefore increased surface roughness and sliding between similar kinds of metal may be attributed to increase in friction coefficient beyond certain temperature. Reduction in wear rate with rise in temperature may also be attributed to increased ability of matrix to accommodate the hard silicon particles therefore cracking tendency of hard particle–matrix interface is reduced [8]. Contamination at sliding surfaces owing to oxidation with temperature rise may also reduce the metallic intimacy/adhesion, which is prerequisite for adhesive wear. Too high temperature rise may allow the thermal softening and so gross plastic flow of material and metallic failure [21].

5. Conclusions From the experimental results following point may be concluded: • Temperature of sliding surface does not change with the sliding distance in steady state under mild oxidative conditions. Unstable temperature rise takes place as soon as transition in wear from mild severe wear occurs. Temperature rise governs the wear rate and wear mechanism. • Increase in sliding speed increases the interface temperature almost linearly up to a critical speed (depending upon the silicon content) above which abrupt rise in interface temperature has been noticed for all the composition used in the present investigation. Increase in silicon content re-

duces the frictional heating and so temperature rise under oxidative wear conditions. • Increase in interface temperature initially reduces the wear rate and friction coefficient irrespective of silicon content. Both wear rate and friction coefficient abruptly increases beyond a certain temperature, which may be termed as critical temperature. Critical temperature increases with increase in silicon content. • Critical temperature for hypereutectic (Al–20% Si) alloy was found higher than the hypoeutectic alloy (Al–4% Si). References [1] F.E. Kennedy, Wear 95 (1984) 454–474. [2] Y. Wang, T. Lei, M. Yan, C. Gao, J. Phys. D Appl. A Phys. 25 (1991) A165–A169. [3] S.C. Lim, M.F. Ashby, Acta Metall. 35 (1) (1987) 1–24. [4] D.K. Dwivedi, A. Sharma, T.V. Rajan, Mater. Trans. 43 (9) (2002) 2256–2261. [5] D.K. Dwivedi, Mater. Sci. Technol. 19 (8) (2003) 1091–1096. [6] A.D. Sarkar, Wear of Metals, Pergamon Press, UK, 1976. [7] S.F. Maustafa, Wear 185 (1995) 189–195. [8] B.K. Prasad, K. Venkateswarlu, O.P. Modi, A.K. Jha, R. Dasgupta, A.H. Yegneswaran, Metall. Mater. Trans. A 29 (1998) 2747–2752. [9] S. Das, S.V. Prasad, T.R. Ramchandran, Wear (1989) 173–187. [10] C. Subramanain, Wear 151 (1991) 97–110. [11] J.F. Archard, J. Appl. Phys. 24 (1953) 981. [12] H. Torabian, J.P. Pathak, S.N. Tiwari, Wear 177 (1994) 47–54. [13] H. So, Wear 184 (1996) 161–167. [14] A. Somi Reddy, B.N. Pramila Bai, K.S.S. Murthy, S.K. Biswas, Wear 181–183 (1995) 658–667. [15] D.K. Dwivedi, Indian Foundry J. 47 (1) (2000) 18–22. [16] A.D. Sarkar, J. Clarck, Wear 75 (1982) 71–85. [17] F.A. Davis, T.S. Eyre, Tribol. Int. 27 (3) (1994) 171–181. [18] K. Razavizadeh, T.S. Eyre, Wear 79 (1982) 325–333. [19] S.K. Biswas, A. Somi Reddy, Aluminium India 18 (4) 1–13. [20] A. Somi Reddy, B.S. Murthy, S.K. Biswas, Wear 171 (1994) 115– 127. [21] S. Lingaurd, K.H. Fu, K.H. Chueng, Wear 96 (1984) 75–84.