Wear 260 (2006) 1356–1360
Sliding wear behaviour of Ca ␣-sialon ceramics at 600 ◦C in air Z.-H. Xie a,∗ , M. Hoffman a , R.J. Moon a , P.R. Munroe a , Y.-B. Cheng b a
School of Materials Science and Engineering, University of New South Wales, Sydney NSW 2052, Australia b School of Physics and Materials Engineering, Monash University, Melbourne Vic. 3800, Australia Received 13 May 2005; received in revised form 24 August 2005; accepted 22 September 2005 Available online 21 November 2005
Abstract As an extension of a previous investigation on the wear behaviour of Ca ␣-sialon ceramics of differing microstructures at room temperature, wear testing was conducted at 600 ◦ C in air to explore the effects of microstructure, contact pressure and sliding speed on the wear behaviour. Under all loading conditions from 1 MPa to 1 GPa, a constant high friction coefficient was observed and a severe wear process was dominant, in which the sliding contact induced cracks were observed in different microstructures. Wear particles were generated along the wear track, but no tribofilm was detected. Increasing the sliding speed from 10 to 23 cm/s was found to significantly increase wear rate. However, variations in microstructure had little impact. That is, large elongated-grained ␣-sialon exhibited only a slightly lower wear rate than fine equiaxed-grained ␣-sialon. © 2005 Elsevier B.V. All rights reserved. Keywords: ␣-sialon; Microstructure; Severe wear; Friction coefficient; Sliding speed
1. Introduction ␣-Sialon ceramics have high hardness and are potential materials for abrasion-resistant applications. In recent years, considerable effort has been invested in the development of in-situ toughened ␣-sialon ceramics , and a typical case is the development of calcium ␣-sialons with elongated grains [2,3]. The room temperature sliding wear behaviour of Ca ␣sialons has been investigated [4–6]. Results revealed that a large elongated-grained ␣-sialon had a higher mild-to-severe wear transition threshold than a fine equiaxed-grained material. Moreover, within the severe wear regime the large elongatedgrained ␣-sialon exhibited a reduced wear rate, compared to the fine equiaxed-grained material, due to a greater resistance to crack extension arising from the coarser microstructure. Within the mild wear regime material removal for the large elongated-grained material was controlled by transgranular fracture, whereas grain pull-out prevailed in the fine equiaxedgrained material. Consequently, a greater wear rate was observed for the finer material. In addition, an increase in sliding speed caused a slight increase in wear rate for both microstructures, more evidently in the mild wear regime. Sliding wear models
Corresponding author. Tel.: +61 402 723 248; fax: +61 2 9385 5956. E-mail address: [email protected]
0043-1648/$ – see front matter © 2005 Elsevier B.V. All rights reserved. doi:10.1016/j.wear.2005.09.011
suggest that grain aspect ratio plays a more important role than grain diameter in the control of sliding wear behaviour . Wear applications of ␣-sialon ceramics may involve high temperatures and oxidizing gaseous environments, where degradation of mechanical properties and chemical reactions may take place on the wear surface. Therefore, further work is required to examine the wear behaviour of ␣-sialons at elevated temperatures. In this present work, unlubricated sliding wear tests were carried out at 600 ◦ C in air, using a ball (SiC)-on-disc configuration, to investigate the effects of microstructure, contact pressure and sliding speed on the sliding wear behaviour of ␣-sialon ceramics. Additionally, this present study is an extension of a previous room temperature wear investigation of Ca ␣-sialon ceramics and will allow the effect of temperature on the wear behaviour to be assessed. The wear tracks and wear particles were examined using a focused ion beam (FIB) system and X-ray photoelectron spectroscopy (XPS) to identify the wear mechanisms of Ca ␣-sialons at high temperature. 2. Experimental procedure Processing of fine equiaxed-grained (EQ) and large elongated-grained (EL) Ca ␣-sialon samples was described in detail in an earlier work . The microstructural parameters and mechanical properties of each are listed in Table 1. The same
Z.-H. Xie et al. / Wear 260 (2006) 1356–1360
Table 1 Microstructural and mechanical properties of ␣-sialon EQ and EL  Sample identification
Average grain diameter [m]
Density [g/cm3 ]
HV [GPa] (load = 98 N)
KIC [MPa m1/2 ] (load = 98 N)a
Young’s modulus [GPa]b
12.8 ± 0.5 12.0 ± 0.2
3.7 ± 0.3 7.5 ± 0.3
Obtained from Vickers indentation test. Obtained from nanoindentation test using a spherical tipped conical indenter (5 m in radius).
procedure was used here for ␣-sialon disc specimen preparation as that used at room temperature . Unlubricated sliding wear tests were performed on a ballon-disc type high temperature tribometer (CSEM Instruments, Switzerland) at 600 ◦ C (measured at the base of the specimen) in air. A 4.5 mm diameter SiC ball was used, which is the same as that used for the room temperature study in . A series of non-stop tests (maximum sliding distance of 1 km) were run at a constant applied normal load of 5 N and at two sliding speeds of 10 and 23 cm/s, with a track radius of 4 mm. The friction coefficient was measured in real time by the tribometer. After each sliding test, loose wear debris scattered on the worn disc were collected for chemical analysis (X-ray Photoelectron Spectroscopy, ESCALAB220i-XL, VG Scientific, UK). Subsequently, the sample surface was gold coated, without any surface cleaning, and examined using a focused ion beam (FIB) system (FEI xP200 focused ion beam miller, FEI Company, Portland, USA). The normalized wear rate, wn , and the apparent contact pressure between the SiC ball and the ␣-sialon disc, p, were calculated at successive distances according to the following equations: wn = p=
2(V2 − V1 ) (l2 − l1 )(A1 + A2 )
2P A1 + A 2
where V is the wear volume, l is the sliding distance, A is the apparent contact area calculated by the wear scar diameter of the ball, P is the normal load, the subscripts 1 and 2 represent measurements at successive distances. The wear volume is calculated as: V = 2πR(Save )
pressure and wear data at room temperature is given for comparison . Firstly, it can be seen that the wear rates of EQ and EL at 600 ◦ C are much higher than at room temperature for all apparent contact pressures. As the apparent contact pressure reduces, the wear rates of both ␣-sialon microstructures decrease slowly, but no wear transition can be distinguished. Secondly, the microstructure had only a small influence on the wear rate; EL has marginally greater wear resistance (i.e. slightly lower wear rate) than EQ. Thirdly, the wear rates of EQ and EL increased with increasing sliding speed, and this effect was more evident as the apparent contact pressure decreases. Measurements of wear scar diameter of the SiC ball at particular sliding distances revealed that the apparent contact pressure was the same for both microstructures as a function of sliding distance. The wear tracks of both EQ and EL materials as a function of sliding distance for a speed of 10 cm/s were examined (Fig. 2). Significant material removal took place continuously on the sliding surfaces of both ␣-sialon materials. Fine wear debris particles were trapped in the wear tracks, and in general, the size of wear particles decreased as the sliding distance increased. There were no identifiable differences between the worn surfaces of the two samples at the same distance in terms of wear debris and surface cracking. The evolution of subsurface damage as the sliding distance increases was revealed by FIB milling and imaging. At high contact pressures (∼800 MPa), large subsurface cracks formed (Fig. 3(a)). As the contact pressure decreased (∼0.8 MPa), the size of cracks was observed to decrease (Fig. 3(b)). When the sliding speed was increased to 23 cm/s, the number and size of cracks increased and more wear debris appeared, as compared to the 10 cm/s sliding speed at
where R is the sliding radius and Save is the average wear track cross-sectional area measured by a contact type profilometer (SURFTEST SV-600, Mitutoyo, Japan). 3. Results The measured friction coefficients of both ␣-sialon samples tested at 600 ◦ C in air were similar and in a range of 0.6–0.8 under both the 10 and 23 cm/s sliding speeds. The influence of both microstructure and sliding speed on wear rates of the two ␣-sialons during sliding tests at 600 ◦ C is shown in Fig. 1. The wear rates are plotted as a function of apparent contact
Fig. 1. Normalized wear rate, wn , of EQ and EL vs. apparent contact pressure, p, at sliding speeds of 10 and 23 cm/s. Note that the line on the left corner shows the normalized wear rate measured for EQ at room temperature, 50 cm/s and 5 N as a function of the apparent contact pressure. This represents the maximum room temperature wear rate of the Ca ␣-sialon material studied by the authors [5,6].
Z.-H. Xie et al. / Wear 260 (2006) 1356–1360
Fig. 2. FIB images of the sliding wear tracks of the EQ and EL microstructures for sliding distances of 1 m (∼800 MPa) and 1 km (∼0.8 MPa) at a sliding speed of 10 cm/s showing that crack-induced severe wear occurs, even at a pressure of ∼0.8 MPa. Arrows indicate the sliding direction of the SiC ball.
the same sliding distance. This is consistent with the increase in measured wear rates seen in Fig. 1. Formation of a tribofilm was not observed on the wear tracks of either EQ or EL following testing, which is in contrast to
those observed at room temperature [4–6]. X-ray photoelectron spectroscopy (XPS) analysis of wear debris collected from both ␣-sialon samples showed no difference in binding energies with the untested sample. Specifically, no oxidation product of Si3 N4 SiO2 could be detected, indicating that there was no oxidation reaction occurring during wear testing at 600 ◦ C. 4. Discussion The current study demonstrates that sliding process of the ␣-sialon ceramics at 600 ◦ C, at both 10 and 23 cm/s sliding speed, was dominated by severe wear even as the apparent contact pressure decreased below 1 MPa (Fig. 1). The severe wear was induced by surface and subsurface cracks, in which wear rates were influenced by sliding speed, but only slightly affected by the ␣-sialon microstructure. These results are in contrast to observations at room temperature [4–6] and it is believed that the increase in temperature facilitated material removal by both thermal and mechanical effects. 4.1. Thermal effect
Fig. 3. FIB cross-sectional micrograph showing sliding contact-induced cracks in the fine-grained ␣-sialon at the sliding distance of (a) 1 m (∼800 MPa) and (b) 1 km (∼0.8 MPa). The sliding speed was 10 cm/s. Note that the size of subsurface cracks is reduced in (b), compared to those observed in (a). Arrow indicates the sliding direction of the SiC ball.
In the present study the bulk temperature of the samples was 600 ◦ C. During the sliding wear test, however, surface temperatures will be higher at the sliding contact points, due to local heat pulses generated by frictional heating. The temperature rise at these sliding contact points, T, can be estimated by considering a circular contact with a circular heat source and the conduction
Z.-H. Xie et al. / Wear 260 (2006) 1356–1360
of heat across the boundary between two bodies with different thermal properties  by: T =
where µ is the friction coefficient, P is the applied normal load, v is the sliding speed, s is the radius of circular contact and keff is the effective thermal conductivity, expressed as : 2 + 0.4vsρ (5) keff = 2.667 kball + kdisc c k disc p,disc disc where kball and kdisc represents the thermal conductivity of the ball and disc, respectively, ρdisc is the density of the disc and cp,disc is the specific heat capacity of the disc. The above analysis suggests that during sliding tests the local temperature would increase with sliding speed, friction coefficient and decrease with increasing local contact area. The local temperature generated during sliding contact was calculated from Eqs. (4) and (5), by taking the friction coefficient of 0.7 (at 600 ◦ C), the normal load of 5 N, the radius of local circular contact of 4 m, the thermal conductivity of SiC of 90 W m−1 K−1 , the thermal conductivity of ␣-sialon of 21 W m−1 K−1 , the density of ␣-sialon of 3200 kg/m3  and the specific heat capacity of ␣-sialon of 680 J/kg . The calculated local sliding contact temperature was 895 ◦ C for v = 10 cm/s and 1277 ◦ C for v = 23 cm/s. The increased local temperature would reduce the local hardness of the samples [10,11], increasing material removal via plastic deformation. Moreover, the increased local temperature would increase the thermal stress at the local contact area, and facilitate the formation of thermally induced cracks, leading to significant material removal observed in this study. As the sliding speed increased, more frictional heating would be generated and accumulated, resulting in an increased local temperature rise in the wear track, and may be responsible for the increasing difference in wear rates measured at 10 and 23 cm/s as the apparent contact pressure decreased (Fig. 1). For comparison purposes, the calculated local sliding contact temperatures for wear tests conducted at 25 ◦ C, and using a friction coefficient of 0.4 , were 189 ◦ C for v = 10 cm/s and 407 ◦ C for v = 23 cm/s. 4.2. Mechanical effect The mechanical processes for causing severe wear are considered to result from localized sliding contact points which interact with each other and lead to an abrasion/indentation mechanism for initiating and then further extending cracks. The influence of this mechanism on severe wear can be estimated by considering the critical conditions necessary for a sliding contact indenter to initiate surface microcracks [12–15]. As a result, the mechanical severity parameter of contact, Sm , which describes the critical condition for the initiation of mechanically-induced severe wear was adopted here as : √ (1 + 10µ)pmax a (6) Sm = Kc
where pmax represents the maximum Hertzian contact pressure induced by a local spherical asperity, a is the length of a surface crack and Kc is the intrinsic fracture toughness. The crack growth and resulting severe wear will start when Sm > 6 . Eq. (6) shows that friction coefficient has a significant influence on the initiation of cracks and subsequent material removal. Taking the intrinsic fracture toughness as 3.7 MPa m1/2 , assuming maximum Hertzian contact pressure equal to the hardness, i.e. 12 GPa , and the friction coefficient of 0.4 (at room temperature) [4–6] and 0.7 (at 600 ◦ C), according to Eq. (6) by increasing the temperature from 25 to 600 ◦ C the initial crack size that is able to cause severe wear is reduced by ∼67%. This suggests that the increase in friction coefficient would increase the wear rates of Ca ␣-sialons at high temperature. The increase in friction coefficient when increasing the temperature from 25 to 600 ◦ C may have resulted from an increased actual contact area between the sliding ball and the sample resulting from the reduction of hardness with increasing temperatures [10,11] and the lack of a lubricating tribofilm. 4.3. Microstructural inﬂuence The effect of microstructure on material removal in the ␣sialons is shown in Fig. 1, in which EL had a slightly lower wear rate than EQ. As discussed in a previous paper , the microstructural variations in Ca ␣-sialon ceramics have little influence on their intrinsic fracture toughness due to the presence of a significant amount of grain boundary glassy phase in these materials. As such, severe wear initiates in both EQ and EL under the same critical conditions. However, following crack initiation, the propagation of the cracks is influenced by the microstructure of the Ca ␣-sialon ceramics and it is expected that the large elongated grains of EL should exert a greater resistance to crack extension than the fine equiaxed grains of EQ and thus reduce the wear rate of EL. This effect is less pronounced in this work than has been observed at room temperature . This may be attributed to significant softening of the glassy phase, which facilitates pull-out of the elongated grains. Additionally, it was observed that the friction coefficient of EL is lower than that of EQ and this slight difference may also contribute to the decreased wear rate of EQ through reduced frictional heating and mechanical loading. This work shows that the variation of sliding speed exhibited a much stronger influence on wear rate than changes in microstructure during high temperature wear of these materials, suggesting that the wear damage incurred by thermal effects has exceeded the contribution of microstructural variation under current test conditions. 5. Conclusions This study of the sliding wear behaviour of ␣-sialon ceramics with fine equiaxed- and large elongated-grain microstructures at a nominal temperature of 600 ◦ C leads to the following conclusions:
Z.-H. Xie et al. / Wear 260 (2006) 1356–1360
1. Crack-induced severe wear occurs in both fine equiaxedgrained and large elongated-grained microstructures, at 10 and 23 cm/s sliding speeds, and over contact pressures ranging from 1 MPa to 1 GPa. 2. The large elongated-grained microstructure exhibits a slightly lower wear rate than fine equiaxed-grained microstructure during testing. This observation is in contrast to the large effect that microstructure had on sliding wear rates at room temperature [4–6]. 3. Higher wear rates were observed by increasing the sliding speed from 10 to 23 cm/s, however, the accumulation of frictional heat seems to exert a strong influence on wear rates. At high contact pressures (∼800 MPa) there was no increase in wear rate with the increased sliding speed, but as the apparent contact pressure decreased the deviation between the wear rates steadily increased. At low contact pressures (∼0.8 MPa) wear rate at 23 cm/s was ∼17% higher than at 10 cm/s. 4. Unlike room temperature testing there was no tribofilm formation at 600 ◦ C. Increase in coefficient of friction facilitated the localized heating of the sample surface and resulted in a smaller critical size flaw to initiate severe wear at 600 ◦ C. Acknowledgements This work was supported by an Australian Research Council (ARC) Large Grant entitled “High Temperature Wear of ␣-Sialon Ceramics”. ␣-Sialon samples were prepared with the assistance of Dr. Y. Zhang of Monash University, Melbourne, Australia. X-ray photoelectron spectroscopy analysis of wear debris was conducted by Dr. Bill Gong at School of Chemistry, University of New South Wales, Australia. The SiC balls used in the sliding wear tests were provided by Dr. George Collins at ANSTO, Australia. References  I.-W. Chen, A. Rosenflanz, A tough sialon ceramic based on ␣-Si3 N4 with a whisker-like microstructure, Nature 389 (6652) (1997) 701–704.
 C.A. Wood, H. Zhao, Y.-B. Cheng, Microstructural development of calcium ␣-sialon ceramics with elongated grains, J. Am. Ceram. Soc. 82 (2) (1999) 421–448.  Z.-H. Xie, M. Hoffman, Y.-B. Cheng, Microstructural tailoring and characterisation of a Ca ␣-sialon composition, J. Am. Ceram. Soc. 85 (4) (2002) 812–818.  Z.-H. Xie, M. Hoffman, P. Munroe, Y.-B. Cheng, Friction and wear of ␣-sialon ceramics, in: E. Pereloma, K. Raviprasad (Eds.), Proceedings of Engineering Materials 2001, The Institute of Materials Engineering, Australasia Ltd., Melbourne, Australia, 2001, pp. 183– 189.  Z.-H. Xie, M. Hoffman, R. Moon, P. Munroe, Y.-B. Cheng, Effect of microstructure on sliding wear of calcium ␣-sialon ceramics, Key Eng. Mater. 280–283 (2005).  Z.-H. Xie, M. Hoffman, R.J. Moon, P.R. Munroe, Y.-B. Cheng, Sliding wear of ␣-sialon ceramics, Wear 260 (2006) 387–400.  J. Bos, H. Moes, Frictional heating of tribological contacts, J. Tribol. 117 (1) (1995) 171–177.  M.W. Barsoum, Fundamentals of Ceramics, McGraw-Hill Series in Materials Science and Engineering, The McGraw-Hill Companies Inc., 1997.  M.G. Gee, D. Butterfield, The combined effect of speed and humidity on the wear and friction of silicon nitride, Wear 162–164 (1993) 234– 245.  O. Unal, J.J. Petrovic, T.E. Mitchell, Mechanical properties of hot isostatically pressed Si3 N4 and Si3 N4 /SiC composites, J. Mater. Res. 8 (3) (1993).  W.P. Tai, T. Watanabe, Elevated-temperature toughness and hardness of a hot-pressed Al2 O3 -WC-Co composite, J. Am. Ceram. Soc. 81 (1) (1998) 257–259.  G.M. Hamilton, Explicit equations for the stress beneath a sliding spherical contact, Proc. Inst. Mech. Engrs. 197C (1983) 53–59.  H. Kong, M.F. Ashby, Wear mechanisms in brittle solids, Acta Metall. Mater. 40 (11) (1992) 2907–2920.  K.-H. Zum Gahr, Modeling and microstructural modification of alumina ceramic for improved tribological properties, Wear 200 (1996) 215– 224.  K. Adachi, K. Kato, N. Chen, Wear map of ceramics, Wear 203–204 (1997) 291–301.  Z.-H. Xie, M. Hoffman, R. Moon, P. Munroe, Y.-B. Cheng, Scratch damage in ceramics: role of microstructure, J. Am. Ceram. Soc. 86 (1) (2003) 141–148.