- Email: [email protected]
Smearing-type wear behavior of Al62Cu25.5Fe12.5 quasicrystal abrasive on soft metals
Intermetallics 68 (2016) 23e30
Contents lists available at ScienceDirect
Intermetallics journal homepage: www.elsevier.com/locate/intermet
Smearing-type wear behavior of Al62Cu25.5Fe12.5 quasicrystal abrasive on soft metals Yongjun Chen, Jianbin Qiang, Chuang Dong* Key Lab of Materials Modification by Laser, Ion, and Electron Beams (Ministry of Education), School of Materials Science and Engineering, Dalian University of Technology, Dalian 116024, China
a r t i c l e i n f o
a b s t r a c t
Article history: Received 20 July 2015 Received in revised form 28 July 2015 Accepted 1 September 2015 Available online 14 September 2015
The abrasive polishing behavior of Al62Cu25.5Fe12.5 quasicrystal on Cu, Al and austenite stainless steel alloys were investigated, to compare with commonly used hard abrasives such as diamond, alumina and silica. The quasicrystal abrasive showed a dominating smearing-type wear mechanism, in sharp contrast to all the other three abrasives, as reflected by large indent size shrinking with respect to surface removal depth. The quasicrystal abrasive polishing, producing a flattened surface with minor depth removal, may open new application fields where low-wearing and fine surface finishing are demanded. © 2015 Elsevier Ltd. All rights reserved.
Keywords: Intermetallics Quasicrystal Surface finishing Tribological properties
1. Introduction Quasicrystals (QCs) have unique surface properties such as low coefficient of friction, low surface energy, and high corrosion resistance [1e4]. As measured by Kang et al. using pin-on-disk against diamond indenter [4], their coefficient of friction varies from 0.05 to 0.2 [4,5], while under the same tribological testing conditions, it is 0.42 for copper, 0.37 for aluminum and 0.32 for lowcarbon steel. The low friction behavior is related to the intrinsic low surface energy property, resulted from the low electron density of states at Fermi level [6,7], which is much lower than that of a typical clean metal and significantly lower than that of an oxidized metal such as oxidized quasicrystals. Their Vickers hardness is about 6.5e11 GPa [1,4], higher than those of normal high-strength steels but close to that of silica. A high H/E ratio is generally indicative of good wear resistance [8]. Their H/E ratios are close to 0.05 GPa, which is similar to that of alumina and is among the highest in metallic compounds. In many aspects, QCs behave like covalent compounds. The combination of all these properties makes them suitable for non-stick oven coating [2]. Generally speaking, QCs are good wear resistant materials, and at the same time, do not wear
* Corresponding author. E-mail address: [email protected] (C. Dong). http://dx.doi.org/10.1016/j.intermet.2015.09.001 0966-9795/© 2015 Elsevier Ltd. All rights reserved.
much their counter parts principally due to the low friction property. Low-friction abrasives generally cause smearing-type wear during polishing [9,10], where plastic deformation smears out the irregular roughness and leaves a smoothened surface, without much surface material removal. On the other hand, via cutting-type wear, hard abrasive particles cut away the asperities on the worn surface, producing a fine surface at the expense of severe surface removal. It was especially noticed that when the frictional coefficient is low, only plastic deformation occurs without producing any wear debris [11]. Consequently, QCs should be useful in fields where low-wearing and fine surface finishing are demanded. In fact, QC particles have already been used as solid lubricant additives in engine oil [12], where the friction is reduced by avoiding intense abrasion while at the same time minimizing local scratching. Their applications can also be envisaged in harsh wearing environments such as oil pipe connectors and motors. The present work explores the abrasive polishing behavior of AleCueFe QC particles on soft metals, to compare with commonly used hard abrasives such as diamond, alumina, and silica. The Al62Cu25.5Fe12.5 QC [13] is used in our work for the low cost and easy availability. The wearing mechanism can be unveiled by following changes in wear rate and in surface morphology.
24
Y. Chen et al. / Intermetallics 68 (2016) 23e30
were machined to a cylindrical form, F40 mm in diameter and 12 mm in height. The initial workpiece surfaces were grounded by SiC abrasive papers with 800, 1500, and finally 2000 grits, successively. In order to study the wear mechanism of the different abrasives, micro-indentation was carried out by using normal loads of 0.25 N, 0.49 N, 0.98 N, 1.96 N, 2.94 N, 4.91 N and 9.81 N on an HV1000 Vivtorinox hardness tester for each workpiece materials. Rhombohedral indents of various sizes were produced with the different loads. Indent morphology was observed by using scanning electron microscope (Zeiss Supra 55 (VP)).
2. Material and methods 2.1. Fabrication of quasicrystals abrasive Ingots of the composition Al62Cu25.5Fe12.5 were prepared by arc melting commercially available pure Al (99.99%), Cu (99.99%) and Fe (99.99%) metals in argon atmosphere. The master alloys were remelted three times in order to achieve chemical homogeneity. From the alloy ingots, cylindrical rod samples of 3 mm in diameter were fabricated by cooper-mould suction-casting in argon atmosphere. The rod samples were then sealed in quartz tubes under vacuum and were annealed at 800 C for 8 h, then furnace cooled. The ingots were further crushed by ball-milling, using cemented carbide balls of 10 mm in diameter as the grinding medium. A ballto-powder weight ratio of 15:1 was employed in this experiment. The rotation speed was 200 r/min. The grinding was carried out for 5 h in argon protection. Finally, the powders were sieved through 700 meshes (~20 mm). Structure, morphology and composition of the powders were characterized using X-ray diffraction (BRUKER D8 Focus, Cu Ka radiation) and scanning electron microscopy (JSM5600LV). Table 1 lists the particle diameters measured by different methods.
2.4. Polishing tests Polishing tests were carried out on a commercial polishing machine PhoenixBeta/Vector, Buehler Ltd., using 10 N vertical load and Buehler-TriDent™ polishing pad (Fig. 1). To enhance the weight loss, a high rotational speed of 350 rpm was chosen. The workpiece holder had a fixed rotational speed of 60 rpm. The average linear sliding speed of the workpiece with respect to the polishing pad was 2 m s1. The duration of each polishing test was 1.5 min and a fixed amount of the abrasive paste was added, 0.2 g. After each 3 min interval, the weight loss of the workpiece was measured by an electronic balance with sensitivity reaching 0.0001 mg. The indent morphologies were observed on SEM. The roughness of the worn surface was measured by a non-contact optical profiler, NewView 5022 (Zygo, USA).
2.2. Abrasive powders and polishing pastes In order to reveal the characteristics of the QC abrasive, three kinds of common polishing abrasives were compared with, including diamond, alumina and silica. The diamond paste was produced by Zhengzhou Research Institute for Abrasives & Grinding Co. Ltd., and the diamond particle size is about 10 mm as given by the company (Table 1). Alumina and silica particles were the products of Shenyang Shihua Powder Material Co., Ltd.. Their sizes are shown in Table 1. The size distributions of the four kinds of abrasives particles were measured using Mastersizer 2000 (Malvern, UK) laser diffractometer. The alumina, QC and silica polishing pastes were prepared by ourselves, using the same lubricant paste (containing mainly stearic acid and Vaseline) from Zhengzhou Research Institute for Abrasives & Grinding Co. Ltd. The abrasiveover-paste concentration was fixed to 10 weight percent (wt. %), the same as in the bought diamond paste.
2.5. AFM Only pure aluminum samples were scanned with atomic force
2.3. Workpiece materials For the workpieces, four relatively soft metals were selected, Ale5Cu (wt.%), pure Cu, Cue6Sn (wt.%), and 18Cre8Ni (wt.%) austenite stainless steel. Their properties are shown in Table 1. They
Fig. 1. Schematic diagram of the polishing test.
Table 1 Properties of workpieces and abrasives. Material
ASTM
Hardness (GPa)
Elastic modulus (GPa)
H/E
KIC (MPa m1/2)
Diameter measured by mastersizer 2000 (mm)
Diameter estimated by SEM (mm)
Pure aluminum Pure copper Ale5Cu 18Cre8Ni Cue6Sn Diamond
e C11000 2024 S30400 C51900 e
0.25a 1.0a 1.6a 2.1a 2.4a 88e100 [14]
68 110 73.1 200 109 800e925 [14]
0.004 0.009 0.022 0.010 0.022 0.10e0.13
6e7 [15]
e
Al2O3
e
15.2e20.3 [16]
36 6[16]
0.04e0.06
2.2e3.5 [17,18]
SiO2
e
8.5 [19]
70e94 [20]
0.12e0.17
1.1e1.5 [19,20]
QC
e
6.5e11 [1,4]
168 [21]
0.05e0.06
1.5 [1,2]
Ave:9e10 90% < 13.3 Ave:8e10 90% < 13.7 Ave:10e16 90% < 36.8 Ave:8e12 90% < 26.3
a
Vickers hardness measured by the present authors at a load of 200 g.
~8 ~10 ~8
Y. Chen et al. / Intermetallics 68 (2016) 23e30
microscopy (Agilent PicoPlusⅡ) to measure the roughness of the worn surface. The workpieces, F6 mm in diameter and 5 mm in height, were grounded by SiC abrasive papers with 800, 1500, and finally 2000 grits, successively. Then they were polished in a solution of perchloric acid and ethanol with a volume ratio of 1:4 under the voltage 15 V for 2 min. Finally they were polished with the Al2O3, SiO2 and QC abrasives for 5 min with load 5 N. 3. Results and discussion 3.1. Structure and morphology of QC abrasive Fig. 2 shows the X-ray diffraction patterns of the Al62Cu25.5Fe12.5 powder alloys prepared under different conditions. In the as-cast state, the alloy contains a mixture of B2-type b-Al(Cu, Fe) and icosahedral QC. In order to increase the QC content, it is necessary to anneal the sample at 800 C for 8 h [22,23]. The XRD pattern shown in Fig. 2(b) corresponds to the thus-annealed sample, where only QC peaks can be observed. The annealed sample was passed to a ball-milling process to fabricate fine abrasive particles, whose diffraction peaks were nearly identical with the as-annealed bulky sample, as shown in Fig. 2(c). The size distribution of diamond and Al2O3 particles are more homogeneously than SiO2 and QC, with the latter two powders showing larger size discrepancies, though average sizes of the four abrasives are similar, about 10 mm. 3.2. Indent morphology evolution of soft Ale5Cu alloy worn by QC and Al2O3 abrasives In order to unveil the abrasive wear mechanism, a soft substrate Ale5Cu was polished by the four kinds of abrasive pastes for the same duration of 3 min. The samples were all initially indented at the seven different loads to leave rhombohedral indent marks of different sizes, as shown in Fig. 3(a). An enlarged initial indent morphology produced from the load of 1.96 N is again shown in Fig. 3(b). Fig. 3(c)e(f) compare the worn surface changes after being polished respectively by the diamond, Al2O3, QC and SiO2 abrasive pastes. It is obvious that the indent morphology issued from the QC abrasive polishing shown in Fig. 3(e) is different from all the others, where visible plastic deformation occurs only around the indent edges, while the indent bottom is not really scratched as reflected by the clean crosslines there, signifying the weakest surface
Fig. 2. X-ray diffraction patterns of the Al62Cu25.5Fe12.5 QC powders prepared by copper-mould suction-casting, post annealed, and ball-milling crushed.
25
removal in this case. Other abrasives induce obvious substrate removal from the edges down to the bottom of the indents. Therefore, the QC abrasive generates the most severe plastic deformation and the smallest surface removal (this will be further discussed later as shown in Fig. 6), indicating a dominating smearing-type wear mechanism in the QC polishing. However, cutting wear is also involved because surface removal is inevitable. In order to quantify the abrasive wear, considering both the surface removal and the smearing effect, a geometrical model is constructed below. 3.3. Geometrical model for cutting and smearing wears The abrasive wearing has been classified according to the movement patterns of particles at the interface, which include mainly cutting and smearing. Free abrasive polishing is a very complex interfacial process. The interactions between the particles and the surfaces involved are highly complicated due to the irregular shapes and sizes of the abrasive particles, generating complex surface topographies on the worn surfaces. Accordingly to the literature [24], round particles produce a wear rate similar to that produced by angular particles in free abrasive polishing, because the particles may orient themselves during contact to assume small attack angles on the leading edges with respect to the polished surface. A geometrical model is then proposed to distinguish the two major polishing mechanisms, cutting and smearing, as shown in Fig. 4, which ignores the shape and size of the abrasive particles and the pressure exerted on per particle. The initial indent morphology is a rhombohedral indent mark produced by a diamond indent tip with the facet angle 136 , as is presented in Fig. 4(a), including a top-view (top) and a cross-section view (bottom). As shown in the top view, the indent diagonal length D is 2√2 times the center-toedge distance R, and after polishing the initial indent shrinks roughly into a dot-lined square, with the center-to-edge distance change of DR ¼ DD/2√2. In order to increase the measurement precision, largely affected by the broad deformation zone inside the indents, the two diagonal lengths changes DD along the crosslines were measured and averaged. The calculated DR quantifies the indent edge-length change. The substrate removal depth after polishing, Dh, can be easily deducted from the mass loss divided by mass density and by exposed surface area. The morphology evolution of an indent is then characterized by the relative change of the edge-length with respect to the surface removal depth, DR/Dh. In Fig. 4, three polishing modes are presented, including ideal cutting, smearing-dominating, and cutting-dominating. Fig. 4(a) presents the ideal cutting mode, which features with pure surface removal while leaving the indent bottom intact. This means that the abrasives do not polish the inside part of the indent so that the indent bottom profile (for instance the angle of 136 ) is kept unchanged during the polishing process. Also no smearing occurs. This mode is presumably produced when very hard abrasives are used. Using the substrate workpiece removal height Dh and the indent width change DR, a parameter tanq ¼ DR/Dh is constructed, which is a constant in the pure cutting mode, tanq1 ¼ DR/Dh ¼ tan (136 /2) ¼ 2.48, q1 being the angle of the indent surface facet with respect to the substrate normal. This mode is only ideal and in practice a mixed cutting and smearing mechanism is generally involved [25e28]. Fig. 4(b) shows the model for the smearing-dominating mode, corresponding to the polishing situation where abrasive particles smear out the indent edges (and any protrusions) into the indent bottom (or any intruding parts) via plastic deformation. The substrate surface layer is then partially squeezed into the indent, making the indent surface curved out. Consequently, in comparison
26
Y. Chen et al. / Intermetallics 68 (2016) 23e30
Fig. 3. Surface morphology changes of the rhombohedral indents on Ale5Cu alloy as observed by SEM. The original indents produced with loads of 0.25 N, 0.49 N, 0.98 N, 1.96 N (the dotted outlined), 2.94 N, 4.91 N and 9.81 N are shown in (a). The enlarged morphology of the outlined indent is shown in (b). The morphologies of the same-size indents were followed after 3 min polishing by the diamond (c), Al2O3 (d), QC (e) and SiO2 (f) abrasives.
with the former ideal cutting mode, for a given surface removal height Dh, the indent edge change DR is larger, which leads to a larger tanq2 ¼ DR/Dh ratio and a larger q2. The q2 angle also qualitatively indicates the degree of curvature of the indent surface due to the squeezing filling, i.e. the indent bottom becomes more flattened that in the ideal cutting mode. Fig. 4(c) shows the model for the cutting-dominating mode, where the cutting removes most of the smearing effect, and what is more, the cutting also removes the indent bottom part (the inclining faces in the figure), though to a less degree than the flat surface. The indent bottom removal means that the indent facets incline further downward, making a smaller q3 than q1 and a smaller tanq2 ¼ DR/Dh ratio. As aforementioned, the deformation modes of the polishing process can then be followed by the tanq ¼ DR/Dh parameter, with tanq2 (smearing-dominating) > tanq1 (pure cutting) > tanq3 (cutting-dominating). This geometrical model provides a simple parameter, tanq ¼ DR/Dh, that distinguishes clearly the occurrence of smearing- or cutting-dominating polishing modes in the complicated abrasive wear processes. In the following, we will
show the experimental results of the polishing processes on relatively soft workpieces, such as pure copper, Ale5Cu, Cue6Sn bronze, and 18Cre8Ni austenite steel, using the four abrasives QC, Al2O3, SiO2, and diamond. The tanq ¼ DR/Dh values, easily obtainable by following the morphology evolutions during the polishing wear, will be used to identify the principal wear mechanism for each combination of abrasive and workpiece. 3.4. Abrasive polishing mechanism evaluated by tanq ¼ DR/Dh parameter In this section, the tanq ¼ DR/Dh experimental results will be shown in the sequence of increasing workpiece hardness, from the softest pure copper to the hardest Cue6Sn alloy. For the softest workpiece, pure copper, as Fig. 5(a) shows, the tanq-vs.-time curves using the four kinds of abrasives all fall above the dotted line, indicating a dominating smearing mechanism; furthermore, the QC abrasive exhibits the most striking dominating smearing in sharp contrast to all the other abrasives. Fig. 5(b) shows the relationship between tanq ¼ DR/Dh and the
Y. Chen et al. / Intermetallics 68 (2016) 23e30
polishing time for the workpiece Ale5Cu polished with the four kinds of abrasives. In this figure, the dotted line represents the value of 2.48 for the ideal cutting mode. Above this line the polishing process is smearing-dominating and below this line, cuttingdominating. For the Ale5Cu workpiece, the QC abrasive presents the most prominent smearing mechanism. Next to it comes the diamond abrasive, which presents relatively strong smearing polishing at the beginning, and with increasing polishing time, the cutting mode gradually dominates the process, with tanq falling close to the ideal value of 2.48. In contrast, the silica and especially Al2O3 abrasives share the cutting-dominating mechanism.
27
Fig. 5(c) and (d) show the same curves but for a harder workpieces, 18Cre8Ni stainless steel and Cue6Sn alloy, respectively. Again, all the wearing processes are dominated by smearing, as indicated by the tanq values all above the horizontal dotted line. Again, the QC abrasive singles out among the four abrasives, and the smearing dominance becomes even more prominent for longer polishing. The above experimental results indicate strongly that the QC abrasive shows prominent smearing wear mechanism against all kinds of substrate workpieces, just as anticipated. This feature makes the QC abrasive distinctively superior to common abrasives
Fig. 4. Geometrical models for ideal cutting (a, including a top-view of the indent), smearing (b), and cutting (c) by following the indent surface profile changes. In each figure, the solid lines mark the initial indent surface, the dotted line the worn surface after polishing.
28
Y. Chen et al. / Intermetallics 68 (2016) 23e30
Fig. 5. Evolutions of tanq ¼ DR/Dh vs. time on pure copper (a), Ale5Cu (b), 18Cre8Ni austenite steel (c), and Cue6Sn(d) polished with different abrasives. The dotted horizontal line indicates tanq ¼ 2.48 for the ideal cutting mode. The error bars arise from averaging the tanq values measured on the five large indents.
Fig. 6. Relationships between mass loss and roughness after polishing for 9 min. The surface roughness is presented by both Ra (shaded bars) and Rpv (open bars, average values marked).
Y. Chen et al. / Intermetallics 68 (2016) 23e30
29
addressed, which further supports the smearing mechanism and unveils the relevant wearing phenomena, especially in relation to the QC abrasive wear. 3.5. Wear rate and roughness
Fig. 7. Relationship between tanq parameter and roughness Ra after polishing for 9 min.
in the sense of smearing-type of polishing, which might open up new application fields for QC materials. In the following, the wear rate and surface roughness will be
The relationships between measured wear rate (workpiece mass loss per unit surface area) and roughness (Ra, average surface roughness and Rpv, peak-to-valley roughness) after polishing for 9 min are presented in Fig. 6, shown in the sequence of increasing hardness, from the softest pure copper to the hardest Cue6Sn alloy. In this figure, wear rate is expressed as filled square symbols (linked with solid lines), and roughness is shown as vertical bars, with shaded ones representing Ra and the open ones representing Rpv. In all the cases the QC abrasive produces the lowest wear rate and a fairly low surface roughness, in good support of the predominant smearing mechanism as revealed in the previous section. The relative low surface roughness also points out that the QC abrasive generates finer surfaces than normal hard abrasives for avoiding deep scratches, which constitutes another unique advantage of this abrasive material. For instance, while polishing Cue6Sn alloy, the QC abrasive produces a wear rate of 2.3 g/m2, which is nearly twenty times smaller than that by alumina, 42.3 g/m2, but their respective roughnesses as measured by Rpv are not so much different, respectively 1007 nm and 591 nm. A low wear rate usually results in a rough surface [29]. In contrary, the QC abrasive
Fig. 8. AFM surface topographies of pure aluminum workpiece polished with abrasives Al2O3 (a), SiO2 (b), and QC (c) for 5 min. The roughness distributions (percentage for a given roughness value) of the three cases are summarized in (d).
30
Y. Chen et al. / Intermetallics 68 (2016) 23e30
polishing produces the smallest mass loss in combination with a smoothened surface, which makes the QC abrasive special from normal abrasives. In order to further unveil the smearing effect in association with surface roughness, the tanq vs. Ra relationship is plotted in Fig. 7. Three zones are clearly identified as classified by their tanq values, a cutting zone A with tanq ¼ 2.48, a transition zone B with tanq ¼ 2.48e10, and a smearing zone C with tanq > 10. The B and C zone boundary at tanq ¼ 10 is set roughly at the periphery of the zone where the roughness data scatter widely, though its exact location cannot be determined at present. In zones A and B, the roughness scatters in an irregular manner, spreading from 10 nm up to 50 nm, while in zone C consisting entirely of the roughness data from the QC abrasive polishing, the roughness Ra shows a decreasing tendency with increasing tanq, from 20 nm down to 7 nm, implying the increasingly dominating smearing mechanism that produces a flattened surface only in the special case of the QC abrasive polishing of the relative hard Cu alloys and the stainless steel. This result also supports the proposed geometrical model, as tanq well reflects the smearing effect in comparison with the cutting effect for free abrasive wear. This tanq parameter can then reasonably be used to predict surface roughness and scratch damage. According to the abrasive materials properties shown in Table 1, QC has hardness values similar to SiO2 and H/E ratios close to Al2O3. To unveil the special surface polishing morphology by the QC abrasive, in comparison with Al2O3 and SiO2, in the next part, initially fine-polished flat surfaces on pure aluminum are observed with AFM after being polished with the three abrasives. 3.6. Fine morphologies of worn surfaces using AFM In this part, the worn surfaces on initially electro-polished pure aluminum were observed for topography and roughness. The samples were polished for 5 min with a load pressure of 5 N using the three abrasives. Topographies and surfaces roughness distributions were obtained from the AFM images shown in Fig. 8. The Al2O3 abrasive produces a large roughness span, centering around 120 nm and extending up to 200 nm. The SiO2 abrasive generates a roughness profile centered by around 70 nm and extending up to 100 nm. In sharp contrast, the QC abrasive polishing results in a profile similar to that of silica but with a broader roughness plateau, ranging from 50 to 70 nm. This result further confirms that the QC abrasive produces a flattened surface finishing in a more efficient manner than the other two abrasives, as a result of the smearing-type wear mechanism. 4. Conclusions After comparing the free abrasive polishing behavior of QCAlCuFe, alumina, silica, and diamond abrasives on soft metals such as Cu and Al alloys and an austenite stainless steel, focusing on the smearing-type plastic deformation, it is pointed out that the QC abrasive shows dominating smearing mode, in sharp contrast to all the other three hard abrasives, as reflected by large indent size shrinking with respect to surface removal depth in accordance with the proposed geometrical model for cutting and smearing wear. The QC abrasive wear produces a flattened surface with the smallest surface removal when used to polish soft metals, thus opening up new application fields for QCs in particle form, where low-wearing and fine surface finishing are demanded.
Acknowledgments The present work is supported by Natural Science Foundation of China under grant number 51131002.
References €ster, W. Liu, H. Liebertz, M. Michel, Mechanical properties of quasicrys[1] U. Ko talline and crystalline phases in AleCueFe alloys, J. Non-Cryst. Solids 153 (1993) 446e452. [2] J.M. Dubois, P. Weinland, Coating Materials for Metal Alloys and Metal and Method, in: United States of America Patent and Trademark Office, U.S. No. 5,204,191, 20 Apr., 1993. [3] J.M. Dubois, Properties-and applications of quasicrystals and complex metallic alloys, Chem. Soc. Rev. 41 (2012) 6760e6777. [4] S. Kang, J. Dubois, J. Von Stebut, Tribological properties of quasicrystalline coatings, J. Mater. Res. 8 (1993) 2471e2481. [5] C.J. Jenks, P.A. Thiel, Quasicrystals: A short review from a surface science perspective, Langmuir 14 (1998) 1392e1397. , J.-M. Dubois, V. Fourne e, P. Brunet, D.J. Sordelet, L. Zhang, About [6] E. Belin-Ferre the Al 3p density of states in AleCueFe compounds and its relation to the compound stability and apparent surface energy of quasicrystals, Mater. Sci. Eng. A 294 (2000) 818e821. [7] C. Dong, A. Perrot, J.M. Dubois, E. Belin, Hume-Rothery phases with constant e/ a value and their related electronic properties in AleCueFe(eCr) quasicrystalline systems, in: Materials Science Forum, Trans. Tech. Publ, 1994, pp. 403e416. [8] T. Oberle, Properties influencing wear of metals, J. Met. 3 (1951) 438e439. [9] G. Beilby, Aggregation and Flow of Solids, Macmillan Publishing, London, 1921. [10] L.E. Samuels, Metallographic Polishing by Mechanical Methods, third ed., ASM, 1982. [11] T. Kayaba, K. Hokkirigawa, K. Kato, Analysis of the abrasive wear mechanism by successive observations of wear processes in a scanning electron microscope, Wear 110 (1986) 419e430. [12] Z. Tao, Y. Wenjie, G. Zhili, AlcCuaXb alloy powder engine oil additive applicable to engine and preparation method thereof, in: State Intellectual Property Office of the P.R.C, China, 21 Mar. 2012. No. 101,570,711 B. [13] E. Huttunen-Saarivirta, Microstructure, fabrication and properties of quasicrystalline AleCueFe alloys: a review, J. Alloys Compd. 363 (1) (2004) 154e178. [14] M. Bauccio, ASM Engineered Materials Reference Book, ASM, Ohio, 1994. [15] F. Lu, Z. Jiang, W. Tang, T. Huang, J. Liu, Accurate measurement of strength and fracture toughness for miniature-size thick diamond-film samples by threepoint bending at constant loading rate, Diam. Relat. Mater. 10 (2001) 770e774. [16] I.D. Marinescu, Handbook of Advanced Ceramics Machining, CRC Press, Boca Raton, USA, 2006. [17] R. Chauhan, Y. Ahn, S. Chandrasekar, T. Farris, Role of indentation fracture in free abrasive machining of ceramics, Wear 162 (1993) 246e257. [18] L. Fang, Y. Gao, S. Si, Q. Zhou, Effect of lubricants on the friction and wear of Al2O3 against gray cast iron, Wear 210 (1997) 145e150. [19] P.J. Blau, ASM Handbook, Friction, Lubrication and Wear Technology, vol. 18, ASM, 1992. ASM International, 1992. [20] J.S. Lyons, T.L. Starr, Strength and toughness of slip-cast fused-silica composites, J. Am. Ceram. Soc. 77 (1994) 1673e1675. [21] K. Tanaka, Y. Mitara, M. Koiwa, Elastic constants of Al-based icosahedral quasicrystals, Philos. Mag. A 73 (1996) 1715e1723. [22] Y. Calvayrac, A. Quivy, M. Bessiere, S. Lefebvre, M. Cornier-Quiquandon, D. Gratias, Icosahedral AlCuFe alloys: towards ideal quasicrystals, J. Phys. 51 (1990) 417e431. [23] G. Rosas, R. Perez, On the transformations of the j-AlCuFe icosahedral phase, Mater. Lett. 47 (2001) 225e230. [24] M.A. Moore, P.A. Swanso, The effect of particle shape on abrasive wear: a comparison of the theory and experiment, in: K.C. Ludema (Ed.), Proc. Int. Conf. on Wear of Materials, ASME, New York, 1983, pp. 1e11. [25] L. Fang, X. Kong, J. Su, Q. Zhou, Movement patterns of abrasive particles in three-body abrasion, Wear 162 (1993) 782e789. [26] L. Fang, Q. Zhou, Y. Li, An explanation of the relation between wear and material hardness in three-body abrasion, Wear 151 (1991) 313e321. [27] F. Liang, K. Xianglong, Z. Qingde, A wear tester capable of monitoring and evaluating the movement pattern of abrasive particles in three-body abrasion, Wear 159 (1992) 115e120. [28] W. Da Silva, H. Costa, J. De Mello, Transitions in abrasive wear mechanisms: effect of the superimposition of interactions, Wear 271 (2011) 977e986. [29] Y. Xie, B. Bhushan, Effects of particle size, polishing pad and contact pressure in free abrasive polishing, Wear 200 (1996) 281e295.
Contents lists available at ScienceDirect
Intermetallics journal homepage: www.elsevier.com/locate/intermet
Smearing-type wear behavior of Al62Cu25.5Fe12.5 quasicrystal abrasive on soft metals Yongjun Chen, Jianbin Qiang, Chuang Dong* Key Lab of Materials Modification by Laser, Ion, and Electron Beams (Ministry of Education), School of Materials Science and Engineering, Dalian University of Technology, Dalian 116024, China
a r t i c l e i n f o
a b s t r a c t
Article history: Received 20 July 2015 Received in revised form 28 July 2015 Accepted 1 September 2015 Available online 14 September 2015
The abrasive polishing behavior of Al62Cu25.5Fe12.5 quasicrystal on Cu, Al and austenite stainless steel alloys were investigated, to compare with commonly used hard abrasives such as diamond, alumina and silica. The quasicrystal abrasive showed a dominating smearing-type wear mechanism, in sharp contrast to all the other three abrasives, as reflected by large indent size shrinking with respect to surface removal depth. The quasicrystal abrasive polishing, producing a flattened surface with minor depth removal, may open new application fields where low-wearing and fine surface finishing are demanded. © 2015 Elsevier Ltd. All rights reserved.
Keywords: Intermetallics Quasicrystal Surface finishing Tribological properties
1. Introduction Quasicrystals (QCs) have unique surface properties such as low coefficient of friction, low surface energy, and high corrosion resistance [1e4]. As measured by Kang et al. using pin-on-disk against diamond indenter [4], their coefficient of friction varies from 0.05 to 0.2 [4,5], while under the same tribological testing conditions, it is 0.42 for copper, 0.37 for aluminum and 0.32 for lowcarbon steel. The low friction behavior is related to the intrinsic low surface energy property, resulted from the low electron density of states at Fermi level [6,7], which is much lower than that of a typical clean metal and significantly lower than that of an oxidized metal such as oxidized quasicrystals. Their Vickers hardness is about 6.5e11 GPa [1,4], higher than those of normal high-strength steels but close to that of silica. A high H/E ratio is generally indicative of good wear resistance [8]. Their H/E ratios are close to 0.05 GPa, which is similar to that of alumina and is among the highest in metallic compounds. In many aspects, QCs behave like covalent compounds. The combination of all these properties makes them suitable for non-stick oven coating [2]. Generally speaking, QCs are good wear resistant materials, and at the same time, do not wear
* Corresponding author. E-mail address: [email protected] (C. Dong). http://dx.doi.org/10.1016/j.intermet.2015.09.001 0966-9795/© 2015 Elsevier Ltd. All rights reserved.
much their counter parts principally due to the low friction property. Low-friction abrasives generally cause smearing-type wear during polishing [9,10], where plastic deformation smears out the irregular roughness and leaves a smoothened surface, without much surface material removal. On the other hand, via cutting-type wear, hard abrasive particles cut away the asperities on the worn surface, producing a fine surface at the expense of severe surface removal. It was especially noticed that when the frictional coefficient is low, only plastic deformation occurs without producing any wear debris [11]. Consequently, QCs should be useful in fields where low-wearing and fine surface finishing are demanded. In fact, QC particles have already been used as solid lubricant additives in engine oil [12], where the friction is reduced by avoiding intense abrasion while at the same time minimizing local scratching. Their applications can also be envisaged in harsh wearing environments such as oil pipe connectors and motors. The present work explores the abrasive polishing behavior of AleCueFe QC particles on soft metals, to compare with commonly used hard abrasives such as diamond, alumina, and silica. The Al62Cu25.5Fe12.5 QC [13] is used in our work for the low cost and easy availability. The wearing mechanism can be unveiled by following changes in wear rate and in surface morphology.
24
Y. Chen et al. / Intermetallics 68 (2016) 23e30
were machined to a cylindrical form, F40 mm in diameter and 12 mm in height. The initial workpiece surfaces were grounded by SiC abrasive papers with 800, 1500, and finally 2000 grits, successively. In order to study the wear mechanism of the different abrasives, micro-indentation was carried out by using normal loads of 0.25 N, 0.49 N, 0.98 N, 1.96 N, 2.94 N, 4.91 N and 9.81 N on an HV1000 Vivtorinox hardness tester for each workpiece materials. Rhombohedral indents of various sizes were produced with the different loads. Indent morphology was observed by using scanning electron microscope (Zeiss Supra 55 (VP)).
2. Material and methods 2.1. Fabrication of quasicrystals abrasive Ingots of the composition Al62Cu25.5Fe12.5 were prepared by arc melting commercially available pure Al (99.99%), Cu (99.99%) and Fe (99.99%) metals in argon atmosphere. The master alloys were remelted three times in order to achieve chemical homogeneity. From the alloy ingots, cylindrical rod samples of 3 mm in diameter were fabricated by cooper-mould suction-casting in argon atmosphere. The rod samples were then sealed in quartz tubes under vacuum and were annealed at 800 C for 8 h, then furnace cooled. The ingots were further crushed by ball-milling, using cemented carbide balls of 10 mm in diameter as the grinding medium. A ballto-powder weight ratio of 15:1 was employed in this experiment. The rotation speed was 200 r/min. The grinding was carried out for 5 h in argon protection. Finally, the powders were sieved through 700 meshes (~20 mm). Structure, morphology and composition of the powders were characterized using X-ray diffraction (BRUKER D8 Focus, Cu Ka radiation) and scanning electron microscopy (JSM5600LV). Table 1 lists the particle diameters measured by different methods.
2.4. Polishing tests Polishing tests were carried out on a commercial polishing machine PhoenixBeta/Vector, Buehler Ltd., using 10 N vertical load and Buehler-TriDent™ polishing pad (Fig. 1). To enhance the weight loss, a high rotational speed of 350 rpm was chosen. The workpiece holder had a fixed rotational speed of 60 rpm. The average linear sliding speed of the workpiece with respect to the polishing pad was 2 m s1. The duration of each polishing test was 1.5 min and a fixed amount of the abrasive paste was added, 0.2 g. After each 3 min interval, the weight loss of the workpiece was measured by an electronic balance with sensitivity reaching 0.0001 mg. The indent morphologies were observed on SEM. The roughness of the worn surface was measured by a non-contact optical profiler, NewView 5022 (Zygo, USA).
2.2. Abrasive powders and polishing pastes In order to reveal the characteristics of the QC abrasive, three kinds of common polishing abrasives were compared with, including diamond, alumina and silica. The diamond paste was produced by Zhengzhou Research Institute for Abrasives & Grinding Co. Ltd., and the diamond particle size is about 10 mm as given by the company (Table 1). Alumina and silica particles were the products of Shenyang Shihua Powder Material Co., Ltd.. Their sizes are shown in Table 1. The size distributions of the four kinds of abrasives particles were measured using Mastersizer 2000 (Malvern, UK) laser diffractometer. The alumina, QC and silica polishing pastes were prepared by ourselves, using the same lubricant paste (containing mainly stearic acid and Vaseline) from Zhengzhou Research Institute for Abrasives & Grinding Co. Ltd. The abrasiveover-paste concentration was fixed to 10 weight percent (wt. %), the same as in the bought diamond paste.
2.5. AFM Only pure aluminum samples were scanned with atomic force
2.3. Workpiece materials For the workpieces, four relatively soft metals were selected, Ale5Cu (wt.%), pure Cu, Cue6Sn (wt.%), and 18Cre8Ni (wt.%) austenite stainless steel. Their properties are shown in Table 1. They
Fig. 1. Schematic diagram of the polishing test.
Table 1 Properties of workpieces and abrasives. Material
ASTM
Hardness (GPa)
Elastic modulus (GPa)
H/E
KIC (MPa m1/2)
Diameter measured by mastersizer 2000 (mm)
Diameter estimated by SEM (mm)
Pure aluminum Pure copper Ale5Cu 18Cre8Ni Cue6Sn Diamond
e C11000 2024 S30400 C51900 e
0.25a 1.0a 1.6a 2.1a 2.4a 88e100 [14]
68 110 73.1 200 109 800e925 [14]
0.004 0.009 0.022 0.010 0.022 0.10e0.13
6e7 [15]
e
Al2O3
e
15.2e20.3 [16]
36 6[16]
0.04e0.06
2.2e3.5 [17,18]
SiO2
e
8.5 [19]
70e94 [20]
0.12e0.17
1.1e1.5 [19,20]
QC
e
6.5e11 [1,4]
168 [21]
0.05e0.06
1.5 [1,2]
Ave:9e10 90% < 13.3 Ave:8e10 90% < 13.7 Ave:10e16 90% < 36.8 Ave:8e12 90% < 26.3
a
Vickers hardness measured by the present authors at a load of 200 g.
~8 ~10 ~8
Y. Chen et al. / Intermetallics 68 (2016) 23e30
microscopy (Agilent PicoPlusⅡ) to measure the roughness of the worn surface. The workpieces, F6 mm in diameter and 5 mm in height, were grounded by SiC abrasive papers with 800, 1500, and finally 2000 grits, successively. Then they were polished in a solution of perchloric acid and ethanol with a volume ratio of 1:4 under the voltage 15 V for 2 min. Finally they were polished with the Al2O3, SiO2 and QC abrasives for 5 min with load 5 N. 3. Results and discussion 3.1. Structure and morphology of QC abrasive Fig. 2 shows the X-ray diffraction patterns of the Al62Cu25.5Fe12.5 powder alloys prepared under different conditions. In the as-cast state, the alloy contains a mixture of B2-type b-Al(Cu, Fe) and icosahedral QC. In order to increase the QC content, it is necessary to anneal the sample at 800 C for 8 h [22,23]. The XRD pattern shown in Fig. 2(b) corresponds to the thus-annealed sample, where only QC peaks can be observed. The annealed sample was passed to a ball-milling process to fabricate fine abrasive particles, whose diffraction peaks were nearly identical with the as-annealed bulky sample, as shown in Fig. 2(c). The size distribution of diamond and Al2O3 particles are more homogeneously than SiO2 and QC, with the latter two powders showing larger size discrepancies, though average sizes of the four abrasives are similar, about 10 mm. 3.2. Indent morphology evolution of soft Ale5Cu alloy worn by QC and Al2O3 abrasives In order to unveil the abrasive wear mechanism, a soft substrate Ale5Cu was polished by the four kinds of abrasive pastes for the same duration of 3 min. The samples were all initially indented at the seven different loads to leave rhombohedral indent marks of different sizes, as shown in Fig. 3(a). An enlarged initial indent morphology produced from the load of 1.96 N is again shown in Fig. 3(b). Fig. 3(c)e(f) compare the worn surface changes after being polished respectively by the diamond, Al2O3, QC and SiO2 abrasive pastes. It is obvious that the indent morphology issued from the QC abrasive polishing shown in Fig. 3(e) is different from all the others, where visible plastic deformation occurs only around the indent edges, while the indent bottom is not really scratched as reflected by the clean crosslines there, signifying the weakest surface
Fig. 2. X-ray diffraction patterns of the Al62Cu25.5Fe12.5 QC powders prepared by copper-mould suction-casting, post annealed, and ball-milling crushed.
25
removal in this case. Other abrasives induce obvious substrate removal from the edges down to the bottom of the indents. Therefore, the QC abrasive generates the most severe plastic deformation and the smallest surface removal (this will be further discussed later as shown in Fig. 6), indicating a dominating smearing-type wear mechanism in the QC polishing. However, cutting wear is also involved because surface removal is inevitable. In order to quantify the abrasive wear, considering both the surface removal and the smearing effect, a geometrical model is constructed below. 3.3. Geometrical model for cutting and smearing wears The abrasive wearing has been classified according to the movement patterns of particles at the interface, which include mainly cutting and smearing. Free abrasive polishing is a very complex interfacial process. The interactions between the particles and the surfaces involved are highly complicated due to the irregular shapes and sizes of the abrasive particles, generating complex surface topographies on the worn surfaces. Accordingly to the literature [24], round particles produce a wear rate similar to that produced by angular particles in free abrasive polishing, because the particles may orient themselves during contact to assume small attack angles on the leading edges with respect to the polished surface. A geometrical model is then proposed to distinguish the two major polishing mechanisms, cutting and smearing, as shown in Fig. 4, which ignores the shape and size of the abrasive particles and the pressure exerted on per particle. The initial indent morphology is a rhombohedral indent mark produced by a diamond indent tip with the facet angle 136 , as is presented in Fig. 4(a), including a top-view (top) and a cross-section view (bottom). As shown in the top view, the indent diagonal length D is 2√2 times the center-toedge distance R, and after polishing the initial indent shrinks roughly into a dot-lined square, with the center-to-edge distance change of DR ¼ DD/2√2. In order to increase the measurement precision, largely affected by the broad deformation zone inside the indents, the two diagonal lengths changes DD along the crosslines were measured and averaged. The calculated DR quantifies the indent edge-length change. The substrate removal depth after polishing, Dh, can be easily deducted from the mass loss divided by mass density and by exposed surface area. The morphology evolution of an indent is then characterized by the relative change of the edge-length with respect to the surface removal depth, DR/Dh. In Fig. 4, three polishing modes are presented, including ideal cutting, smearing-dominating, and cutting-dominating. Fig. 4(a) presents the ideal cutting mode, which features with pure surface removal while leaving the indent bottom intact. This means that the abrasives do not polish the inside part of the indent so that the indent bottom profile (for instance the angle of 136 ) is kept unchanged during the polishing process. Also no smearing occurs. This mode is presumably produced when very hard abrasives are used. Using the substrate workpiece removal height Dh and the indent width change DR, a parameter tanq ¼ DR/Dh is constructed, which is a constant in the pure cutting mode, tanq1 ¼ DR/Dh ¼ tan (136 /2) ¼ 2.48, q1 being the angle of the indent surface facet with respect to the substrate normal. This mode is only ideal and in practice a mixed cutting and smearing mechanism is generally involved [25e28]. Fig. 4(b) shows the model for the smearing-dominating mode, corresponding to the polishing situation where abrasive particles smear out the indent edges (and any protrusions) into the indent bottom (or any intruding parts) via plastic deformation. The substrate surface layer is then partially squeezed into the indent, making the indent surface curved out. Consequently, in comparison
26
Y. Chen et al. / Intermetallics 68 (2016) 23e30
Fig. 3. Surface morphology changes of the rhombohedral indents on Ale5Cu alloy as observed by SEM. The original indents produced with loads of 0.25 N, 0.49 N, 0.98 N, 1.96 N (the dotted outlined), 2.94 N, 4.91 N and 9.81 N are shown in (a). The enlarged morphology of the outlined indent is shown in (b). The morphologies of the same-size indents were followed after 3 min polishing by the diamond (c), Al2O3 (d), QC (e) and SiO2 (f) abrasives.
with the former ideal cutting mode, for a given surface removal height Dh, the indent edge change DR is larger, which leads to a larger tanq2 ¼ DR/Dh ratio and a larger q2. The q2 angle also qualitatively indicates the degree of curvature of the indent surface due to the squeezing filling, i.e. the indent bottom becomes more flattened that in the ideal cutting mode. Fig. 4(c) shows the model for the cutting-dominating mode, where the cutting removes most of the smearing effect, and what is more, the cutting also removes the indent bottom part (the inclining faces in the figure), though to a less degree than the flat surface. The indent bottom removal means that the indent facets incline further downward, making a smaller q3 than q1 and a smaller tanq2 ¼ DR/Dh ratio. As aforementioned, the deformation modes of the polishing process can then be followed by the tanq ¼ DR/Dh parameter, with tanq2 (smearing-dominating) > tanq1 (pure cutting) > tanq3 (cutting-dominating). This geometrical model provides a simple parameter, tanq ¼ DR/Dh, that distinguishes clearly the occurrence of smearing- or cutting-dominating polishing modes in the complicated abrasive wear processes. In the following, we will
show the experimental results of the polishing processes on relatively soft workpieces, such as pure copper, Ale5Cu, Cue6Sn bronze, and 18Cre8Ni austenite steel, using the four abrasives QC, Al2O3, SiO2, and diamond. The tanq ¼ DR/Dh values, easily obtainable by following the morphology evolutions during the polishing wear, will be used to identify the principal wear mechanism for each combination of abrasive and workpiece. 3.4. Abrasive polishing mechanism evaluated by tanq ¼ DR/Dh parameter In this section, the tanq ¼ DR/Dh experimental results will be shown in the sequence of increasing workpiece hardness, from the softest pure copper to the hardest Cue6Sn alloy. For the softest workpiece, pure copper, as Fig. 5(a) shows, the tanq-vs.-time curves using the four kinds of abrasives all fall above the dotted line, indicating a dominating smearing mechanism; furthermore, the QC abrasive exhibits the most striking dominating smearing in sharp contrast to all the other abrasives. Fig. 5(b) shows the relationship between tanq ¼ DR/Dh and the
Y. Chen et al. / Intermetallics 68 (2016) 23e30
polishing time for the workpiece Ale5Cu polished with the four kinds of abrasives. In this figure, the dotted line represents the value of 2.48 for the ideal cutting mode. Above this line the polishing process is smearing-dominating and below this line, cuttingdominating. For the Ale5Cu workpiece, the QC abrasive presents the most prominent smearing mechanism. Next to it comes the diamond abrasive, which presents relatively strong smearing polishing at the beginning, and with increasing polishing time, the cutting mode gradually dominates the process, with tanq falling close to the ideal value of 2.48. In contrast, the silica and especially Al2O3 abrasives share the cutting-dominating mechanism.
27
Fig. 5(c) and (d) show the same curves but for a harder workpieces, 18Cre8Ni stainless steel and Cue6Sn alloy, respectively. Again, all the wearing processes are dominated by smearing, as indicated by the tanq values all above the horizontal dotted line. Again, the QC abrasive singles out among the four abrasives, and the smearing dominance becomes even more prominent for longer polishing. The above experimental results indicate strongly that the QC abrasive shows prominent smearing wear mechanism against all kinds of substrate workpieces, just as anticipated. This feature makes the QC abrasive distinctively superior to common abrasives
Fig. 4. Geometrical models for ideal cutting (a, including a top-view of the indent), smearing (b), and cutting (c) by following the indent surface profile changes. In each figure, the solid lines mark the initial indent surface, the dotted line the worn surface after polishing.
28
Y. Chen et al. / Intermetallics 68 (2016) 23e30
Fig. 5. Evolutions of tanq ¼ DR/Dh vs. time on pure copper (a), Ale5Cu (b), 18Cre8Ni austenite steel (c), and Cue6Sn(d) polished with different abrasives. The dotted horizontal line indicates tanq ¼ 2.48 for the ideal cutting mode. The error bars arise from averaging the tanq values measured on the five large indents.
Fig. 6. Relationships between mass loss and roughness after polishing for 9 min. The surface roughness is presented by both Ra (shaded bars) and Rpv (open bars, average values marked).
Y. Chen et al. / Intermetallics 68 (2016) 23e30
29
addressed, which further supports the smearing mechanism and unveils the relevant wearing phenomena, especially in relation to the QC abrasive wear. 3.5. Wear rate and roughness
Fig. 7. Relationship between tanq parameter and roughness Ra after polishing for 9 min.
in the sense of smearing-type of polishing, which might open up new application fields for QC materials. In the following, the wear rate and surface roughness will be
The relationships between measured wear rate (workpiece mass loss per unit surface area) and roughness (Ra, average surface roughness and Rpv, peak-to-valley roughness) after polishing for 9 min are presented in Fig. 6, shown in the sequence of increasing hardness, from the softest pure copper to the hardest Cue6Sn alloy. In this figure, wear rate is expressed as filled square symbols (linked with solid lines), and roughness is shown as vertical bars, with shaded ones representing Ra and the open ones representing Rpv. In all the cases the QC abrasive produces the lowest wear rate and a fairly low surface roughness, in good support of the predominant smearing mechanism as revealed in the previous section. The relative low surface roughness also points out that the QC abrasive generates finer surfaces than normal hard abrasives for avoiding deep scratches, which constitutes another unique advantage of this abrasive material. For instance, while polishing Cue6Sn alloy, the QC abrasive produces a wear rate of 2.3 g/m2, which is nearly twenty times smaller than that by alumina, 42.3 g/m2, but their respective roughnesses as measured by Rpv are not so much different, respectively 1007 nm and 591 nm. A low wear rate usually results in a rough surface [29]. In contrary, the QC abrasive
Fig. 8. AFM surface topographies of pure aluminum workpiece polished with abrasives Al2O3 (a), SiO2 (b), and QC (c) for 5 min. The roughness distributions (percentage for a given roughness value) of the three cases are summarized in (d).
30
Y. Chen et al. / Intermetallics 68 (2016) 23e30
polishing produces the smallest mass loss in combination with a smoothened surface, which makes the QC abrasive special from normal abrasives. In order to further unveil the smearing effect in association with surface roughness, the tanq vs. Ra relationship is plotted in Fig. 7. Three zones are clearly identified as classified by their tanq values, a cutting zone A with tanq ¼ 2.48, a transition zone B with tanq ¼ 2.48e10, and a smearing zone C with tanq > 10. The B and C zone boundary at tanq ¼ 10 is set roughly at the periphery of the zone where the roughness data scatter widely, though its exact location cannot be determined at present. In zones A and B, the roughness scatters in an irregular manner, spreading from 10 nm up to 50 nm, while in zone C consisting entirely of the roughness data from the QC abrasive polishing, the roughness Ra shows a decreasing tendency with increasing tanq, from 20 nm down to 7 nm, implying the increasingly dominating smearing mechanism that produces a flattened surface only in the special case of the QC abrasive polishing of the relative hard Cu alloys and the stainless steel. This result also supports the proposed geometrical model, as tanq well reflects the smearing effect in comparison with the cutting effect for free abrasive wear. This tanq parameter can then reasonably be used to predict surface roughness and scratch damage. According to the abrasive materials properties shown in Table 1, QC has hardness values similar to SiO2 and H/E ratios close to Al2O3. To unveil the special surface polishing morphology by the QC abrasive, in comparison with Al2O3 and SiO2, in the next part, initially fine-polished flat surfaces on pure aluminum are observed with AFM after being polished with the three abrasives. 3.6. Fine morphologies of worn surfaces using AFM In this part, the worn surfaces on initially electro-polished pure aluminum were observed for topography and roughness. The samples were polished for 5 min with a load pressure of 5 N using the three abrasives. Topographies and surfaces roughness distributions were obtained from the AFM images shown in Fig. 8. The Al2O3 abrasive produces a large roughness span, centering around 120 nm and extending up to 200 nm. The SiO2 abrasive generates a roughness profile centered by around 70 nm and extending up to 100 nm. In sharp contrast, the QC abrasive polishing results in a profile similar to that of silica but with a broader roughness plateau, ranging from 50 to 70 nm. This result further confirms that the QC abrasive produces a flattened surface finishing in a more efficient manner than the other two abrasives, as a result of the smearing-type wear mechanism. 4. Conclusions After comparing the free abrasive polishing behavior of QCAlCuFe, alumina, silica, and diamond abrasives on soft metals such as Cu and Al alloys and an austenite stainless steel, focusing on the smearing-type plastic deformation, it is pointed out that the QC abrasive shows dominating smearing mode, in sharp contrast to all the other three hard abrasives, as reflected by large indent size shrinking with respect to surface removal depth in accordance with the proposed geometrical model for cutting and smearing wear. The QC abrasive wear produces a flattened surface with the smallest surface removal when used to polish soft metals, thus opening up new application fields for QCs in particle form, where low-wearing and fine surface finishing are demanded.
Acknowledgments The present work is supported by Natural Science Foundation of China under grant number 51131002.
References €ster, W. Liu, H. Liebertz, M. Michel, Mechanical properties of quasicrys[1] U. Ko talline and crystalline phases in AleCueFe alloys, J. Non-Cryst. Solids 153 (1993) 446e452. [2] J.M. Dubois, P. Weinland, Coating Materials for Metal Alloys and Metal and Method, in: United States of America Patent and Trademark Office, U.S. No. 5,204,191, 20 Apr., 1993. [3] J.M. Dubois, Properties-and applications of quasicrystals and complex metallic alloys, Chem. Soc. Rev. 41 (2012) 6760e6777. [4] S. Kang, J. Dubois, J. Von Stebut, Tribological properties of quasicrystalline coatings, J. Mater. Res. 8 (1993) 2471e2481. [5] C.J. Jenks, P.A. Thiel, Quasicrystals: A short review from a surface science perspective, Langmuir 14 (1998) 1392e1397. , J.-M. Dubois, V. Fourne e, P. Brunet, D.J. Sordelet, L. Zhang, About [6] E. Belin-Ferre the Al 3p density of states in AleCueFe compounds and its relation to the compound stability and apparent surface energy of quasicrystals, Mater. Sci. Eng. A 294 (2000) 818e821. [7] C. Dong, A. Perrot, J.M. Dubois, E. Belin, Hume-Rothery phases with constant e/ a value and their related electronic properties in AleCueFe(eCr) quasicrystalline systems, in: Materials Science Forum, Trans. Tech. Publ, 1994, pp. 403e416. [8] T. Oberle, Properties influencing wear of metals, J. Met. 3 (1951) 438e439. [9] G. Beilby, Aggregation and Flow of Solids, Macmillan Publishing, London, 1921. [10] L.E. Samuels, Metallographic Polishing by Mechanical Methods, third ed., ASM, 1982. [11] T. Kayaba, K. Hokkirigawa, K. Kato, Analysis of the abrasive wear mechanism by successive observations of wear processes in a scanning electron microscope, Wear 110 (1986) 419e430. [12] Z. Tao, Y. Wenjie, G. Zhili, AlcCuaXb alloy powder engine oil additive applicable to engine and preparation method thereof, in: State Intellectual Property Office of the P.R.C, China, 21 Mar. 2012. No. 101,570,711 B. [13] E. Huttunen-Saarivirta, Microstructure, fabrication and properties of quasicrystalline AleCueFe alloys: a review, J. Alloys Compd. 363 (1) (2004) 154e178. [14] M. Bauccio, ASM Engineered Materials Reference Book, ASM, Ohio, 1994. [15] F. Lu, Z. Jiang, W. Tang, T. Huang, J. Liu, Accurate measurement of strength and fracture toughness for miniature-size thick diamond-film samples by threepoint bending at constant loading rate, Diam. Relat. Mater. 10 (2001) 770e774. [16] I.D. Marinescu, Handbook of Advanced Ceramics Machining, CRC Press, Boca Raton, USA, 2006. [17] R. Chauhan, Y. Ahn, S. Chandrasekar, T. Farris, Role of indentation fracture in free abrasive machining of ceramics, Wear 162 (1993) 246e257. [18] L. Fang, Y. Gao, S. Si, Q. Zhou, Effect of lubricants on the friction and wear of Al2O3 against gray cast iron, Wear 210 (1997) 145e150. [19] P.J. Blau, ASM Handbook, Friction, Lubrication and Wear Technology, vol. 18, ASM, 1992. ASM International, 1992. [20] J.S. Lyons, T.L. Starr, Strength and toughness of slip-cast fused-silica composites, J. Am. Ceram. Soc. 77 (1994) 1673e1675. [21] K. Tanaka, Y. Mitara, M. Koiwa, Elastic constants of Al-based icosahedral quasicrystals, Philos. Mag. A 73 (1996) 1715e1723. [22] Y. Calvayrac, A. Quivy, M. Bessiere, S. Lefebvre, M. Cornier-Quiquandon, D. Gratias, Icosahedral AlCuFe alloys: towards ideal quasicrystals, J. Phys. 51 (1990) 417e431. [23] G. Rosas, R. Perez, On the transformations of the j-AlCuFe icosahedral phase, Mater. Lett. 47 (2001) 225e230. [24] M.A. Moore, P.A. Swanso, The effect of particle shape on abrasive wear: a comparison of the theory and experiment, in: K.C. Ludema (Ed.), Proc. Int. Conf. on Wear of Materials, ASME, New York, 1983, pp. 1e11. [25] L. Fang, X. Kong, J. Su, Q. Zhou, Movement patterns of abrasive particles in three-body abrasion, Wear 162 (1993) 782e789. [26] L. Fang, Q. Zhou, Y. Li, An explanation of the relation between wear and material hardness in three-body abrasion, Wear 151 (1991) 313e321. [27] F. Liang, K. Xianglong, Z. Qingde, A wear tester capable of monitoring and evaluating the movement pattern of abrasive particles in three-body abrasion, Wear 159 (1992) 115e120. [28] W. Da Silva, H. Costa, J. De Mello, Transitions in abrasive wear mechanisms: effect of the superimposition of interactions, Wear 271 (2011) 977e986. [29] Y. Xie, B. Bhushan, Effects of particle size, polishing pad and contact pressure in free abrasive polishing, Wear 200 (1996) 281e295.