Fretting wear behavior of AZ91D magnesium alloy
CHEN An-hua(陈安华)1, HUANG Wei-jiu(黄伟九)2, LI Zhao-feng(李兆峰)2
1. School of Mechanical Engineering, Hunan University of Science and Technology, Xiangtan 411201, China;
2. School of Materials Science and Engineering, Chongqing Institute of Technology, Chongqing 400050, China
Received 28 July 2006; accepted 15 September 2006
Abstract: The fretting behaviour of the AZ91D magnesium alloy was investigated. The influence of the number of cycles, normal load (contact pressure) and the amplitude of slip on the fretting behavior of the material were focused. Fretting tests were performed under various running conditions with regard to normal load levels and slip amplitudes. The friction coefficient between the surfaces at the fretting junction was continuously recorded. The fretting damage on the magnesium specimens was studied by SEM. The results show that the wear volume increases with the increase of slip amplitude, and linearly increases with the increase of normal load in the mixed and gross slip regime, but the normal load has no obvious effect on the wear volume in the partial slip regime. The predominant fretting wear mechanism of magnesium alloy in the slip regime is the oxidation wear, delaminated wear and abrasive wear.
Key words: AZ91D magnesium alloy; fretting; friction coefficient; wear mechanism; wear behavior
1 Introduction
Fretting always occurs when the small tangential displacement acts repeatedly between surfaces with an applied quasi-static normal force. Fretting damage is a very complex phenomenon which involves adhesion, abrasion, fatigue and corrosion on the contact surfaces. Under different conditions, it appears as fretting fatigue, fretting wear and fretting corrosion[1-3].
The mass reduction of automobiles is one of the most effective ways for improving fuel consumption since the resistances of a vehicle to rolling, climbing and acceleration are directly dependent on the vehicle mass. Therefore, the application of magnesium alloys whose density is only roughly 25% as that of steel and 66% as that of aluminum is expected to substantially increase in this decade. The widely used magnesium alloys are Mg-Al series, such as AZ91D and AM60B,which have been mainly used to produce the engineering parts such as transmission gearbox housings, seat frames, accelerator pedal bracket, instrument boards of cars and steering wheel[4-6]. The accessories in automobile are always subjected to the vibration environment. The fretting damages of these accessories are unavoidable. So it is of a great theoretical and practical importance to carry out a fretting tribological investigation of magnesium alloys. So in this study, the fretting wear properties of AZ91D magnesium alloy and the influence of the number of cycles, normal load (contact pressure) and the amplitude of slip on the fretting behavior of the material were systematically investigated, and the fretting damage mechanism was also discussed.
2 Experimental
The fretting wear tests were carried out using a hydraulic Deltalab-Nene testing system, as described in Ref.[7]. The material used in this study was AZ91D alloy, and the tests were made on the same metal combinations in ambient laboratory air. The magnesium alloy balls with a diameter of 30 mm were used as the sphere specimens. The flat specimens with dimensions of
10 mm×10 mm×20 mm were polished to a roughness (Ra) of approximately 0.04 μm by diamond paste and alumina powder. All specimens were cleaned in an ultrasonic bath using alcohol before tests. All tests were conducted at room temperature of (23±2) ℃ and relative humidity of (50±10)%.
After testing, variations of tangential force Ft with the displacement amplitude (D) as a function of the number of cycles (Ft—D—N curves or friction log) were analyzed. The profiles of fretting scares were measured by a Talysurf 6 profilometer, and their morphologies and wear debris were examined by JSM-5600LV scanning electron microscope with an energy dispersive X-ray spectroscope.
3 Results and discussion
3.1 Friction logs and fretting maps
Fretting logs of the AZ91D magnesium alloy at different displacement amplitudes are shown in Fig.1.
According to the variation of Ft—D—N curves, three fretting regimes, namely, partial slip, mixed and gross slips regimes are observed. For example, under a given normal load of 200 N, at a very low displacement amplitude such as 5 μm, the Ft—D loop was closed. A few parallel epipedic cycles occurring at the beginning are caused by the initial friction of superficial contamination layers (Fig.1(a)). On the contrary, for the higher displacement amplitude (60 μm), all Ft—D curves are parallel epipedic cycles and relative slip takes place during a whole fretting process in slip regime (Fig.1(c)). For an intermediate displacement amplitude such as
20 μm, the Ft—D curves are composed of partial slip and gross slip cycles, depending on the number of fretting cycles (Fig.1(b)). Based on these friction logs and the observation of wear scars, the fretting maps of AZ91D magnesium alloy are drawn as shown in Fig.2, which indicates that there are a obvious mixed regime and a cracking domain in the running conditions fretting map(RCFM) and material response fretting map(MRFM), respectively. This implies that AZ91D magnesium alloy can be damaged by fretting fatigue under certain fretting contact conditions, and further indicates that the application of AZ91D magnesium alloy should be avoided under these contact conditions.
Fig.1 Friction logs of AZ91D magnesium alloy at different amplitudes: (a)D=5 μm; (b) D=20 μm; (c) D=60 μm
(Normal load Fn=200 N, Frequency f= 5 Hz, Number of cycles N=30 000)
Fig.2 Fretting maps of AZ91D magnesium alloy: (a) RCFM; (b) MRFM
3.2 Friction coefficient
The variation of friction coefficient of AZ91D magnesium alloy with number of cycles at a range of slip amplitude under the same experimental conditions (normal load 200 N, frequency 5 Hz) is shown in Fig.3. It can be seen that the friction coefficient at the slip amplitude of 5 μm increases rapidly at the beginning and reaches the largest value after 400 fretting cycles. During the following cycles, it shows a little fluctuation, but it still maintains the level of the largest value around 0.28. While the slip amplitude is 20 μm, the variation trend of
friction coefficient is similar, but the cycles to attain the maximum value is about 5 000, which is far more than that at the slip amplitude of 5 μm. However, the friction coefficient at the slip amplitude of 60 μm varies in a quite wide range, the friction coefficient shows a gradual rise tendency, and attains the largest value of 0.88 at about 4 000 cycles. With a further increase in testing cycles, it shows a gradual descent and still maintains the level around 0.75 after 10 000 cycles. The difference of the variation tendency of the friction coefficient during fretting process could be attributed to different fretting regimes.
The influence of slip amplitude on the friction coefficient is shown in Fig.4. It can be seen that the friction coefficient increases with the slip amplitude while at a small displacement amplitude (e.g. D≤20μm), and then shows the tendency of approaching a constant value with the increase of slip amplitude. The friction coefficient becoming constant could be attributed to that the total slip occurring in the fretting contact and the wear becomes the predominant characteristic of the process.
Fig.3 Dependence of friction coefficient on number of cycles (Fn=200 N; f=5 Hz)
Fig.4 Effect of displacement on friction coefficient and wear volume(Fn=200 N; f=5 Hz; N=30 000)
3.3 Wear behavior
The variation of wear volume of magnesium alloy with the slip amplitude under the same experimental conditions (normal load 200 N, frequency 5 Hz, number of cycle 30 000) is also shown in Fig.4. The results show that the wear volume rapidly increases with an increase in the slip amplitude when the slip amplitude is lower than 40 μm. After that, the variation of wear volume becomes moderate with an increase in the slip amplitude.
Fig.5 shows the variation of wear volume of magnesium alloy with the normal load under different slip amplitudes. It can be found that the normal load has little effect on the wear volume of magnesium alloy while at a small displacement amplitude (e.g. D=5 μm). However the wear volume linearly increases with normal load while at larger slip amplitude (e.g. D=20 μm or D=60 μm).
Fig. 5 Effect of load on wear volume( f=5 Hz; N=30 000)
3.4 Wear scar morphology
The variations of wear scar morphologies of magnesium alloy with number of cycles are shown in Fig.6. At 1 000 cycles, the wear scar morphology displays a few adhesive trace and abrasive grooves, and there are a few fine flake-like wear debris presented on the wear scar (Fig.6(a)). In the following 4 000 cycles, the wear scar further increases and a large quantity of debris spread over the wear surface, and a few micro cracks are resulted from the severe contact stress (Fig.6(b)). When the fretting process reaches 30 000 cycles, the whole contact surface is full of grooves and detached debris (Fig.6(c)), which indicates that the abra-sive wear takes place at the interface. The abrasives come from the detached particles and materials ploughed.
Fig.6 SEM images of worn surfaces of AZ91D alloys at different cycles: (a) 1 000; (b) 5 000; (c) 30 000
3.5 Fretting damage mechanism of AZ91D magne-
sium alloy
Based on the fretting test results and the morphologies of wear scar, the fretting damage mechanism of AZ91D magnesium alloy in gross slip regime was discussed.
With regard to the evolution of the friction coefficient curves (as shown in Fig.3, D=60 μm), the fretting wear process of magnesium alloy could be divided into three stages. In the first stage, the friction coefficient increases significantly with the number of fretting cycles. The increase in the friction coefficient is attributed to an increase in the contact area and the combined effects of the cyclic plastic deformation and work hardening on the surface and sub-surface layers[8]. This stage is also associated with adhesion between the magnesium alloys in fretting contact (see Fig.6(a)), which results in that the adhesive wear becomes the predominant wear mechanism. In the second stage, the descent and fluctuation of the friction coefficient are observed, which is mainly caused by a series of changes takning place on the contact surface continuously or alternatively, including the crack nucleation and propagation, delamination of deformed materials (see Fig.6(b)), movement and oxidation of wear debris, brittle fracture and degradation of oxide film, and so on. One important result could be regarded as the reasonable evidence for the above supposition. As listed in Table 1, the content of oxygen in wear debris on the central contract region is far higher than that in the magnesium alloy substrate, which indicates that the oxidation of metal substrate (or wear debris) occurs in the process of fretting. The formation of metal oxide will result in the improvement of the brittleness of metal surface layer, and also provide the source for the formation of wear debris[9]. So the fatigue wear, oxidation wear and abrasive wear are the predominant wear mechanism in this stage. The third stage is characterized by stabilized fretting condition, the contact surfaces are completely separated by the detached wear debris, the formation and ejection of wear debris attain dynamic balance, and the friction coefficient remains a fairly steady value, the abrasive wear is the predominant wear mechanism in this stage. From the whole fretting process, the predominant wear damage mechanism is the oxidation wear, delaminated wear and abrasive wear.
Table 1 Chemical composition of metal substrate and wear debris
4 Conclusions
1) Three kinds of force-displacement cycle were observed for AZ91D magnesium alloy, depending on the experimental conditions, which indicates that all of the three fretting regimes, namely, partial slip, mixed and gross slips regimes occur in the fretting process of AZ91D magnesium alloy.
2) The wear volume increases with an increase of slip amplitude and linearly increases with an increase of normal load in the mixed and gross slip regime, but normal load has no obvious effect on the wear volume in the partial slip regime.
3) In the gross slip regime, the fretting damage mechanism of AZ91D magnesium alloy is the oxidation wear, delaminated wear and abrasive wear.
References
[1] WATERHOUSE R B. Fretting Corrosion[M]. Oxford: Pergamon Press, 1972.
[2] WATERHOUSE R B. Fretting Fatigue[M]. London: Applied Science Press, 1981.
[3] BERTHER Y, VINCENT L, GODEL M. Fretting fatigue and fretting wear[J]. Tribology International, 1989, 22(4): 235-242.
[4] DECKER R F. The renaissance in magnesium[J]. Advanced Mater & Proc, 1998(9): 31-35.
[5] FROES F H, ELIEZER D, AGHION E. The science, technology and application of magnesium[J]. J Mine Metals and Mater Soc, 1998, 5(9): 30-34.
[6] CLOW B B. Magnesium industry overview[J]. Advanced Mater & Proc, 1996, 150(4): 33-36.
[7] ZHOU Z R, VINCENT L. Fretting Wear[M]. Beijing: Science Press, 2002.
[8] PUKL B, VODOPIVEC F. The fretting behaviour of AlSiMg-T6[J]. Wear, 2002, 212: 173-182.
[9] LI Shi-zhuo, DONG Xiang-lin. Erosion Wear and Fretting Wear of Materials[M]. Beijing: China Machine Press, 1987.(in Chinese)
(Edited by CHEN Wei-ping)
Foundation item: Project(KFJJ0302) supported by the Hunan Provincial Key Laboratory of Health Maintenance for Mechanical Equipment, China
Corresponding author: CHEN An-hua; Tel: +86-732-8290177; E-mail: ahchen@hnust.edu.cn