- Email: [email protected]
Stepwise erosion as a method for determining the mechanisms of wear in gas borne particulate erosion
Wear 255 (2003) 44–54
Stepwise erosion as a method for determining the mechanisms of wear in gas borne particulate erosion M.G. Gee∗ , R.H. Gee, I. McNaught National Physical Laboratory, Centre for Materials Measurement, Queens Road, Teddington, Middlesex TW11 0LW, UK
Abstract In erosion and abrasion, the detailed processes that cause material removal are still poorly understood. This means that with a few notable exceptions, good models for the prediction of materials behaviour are not still readily available. Stepwise testing is a new approach that has been recently developed as a way of providing information on the build up of damage in erosion. The technique uses a combination of exposing a test sample to repeated very low duration abrasion or erosion followed by accurate relocation in the SEM. This enables the real wear processes within a material to be followed in considerable detail, giving an excellent basis for developing new and improved models. In this preliminary study, the technique has been applied to the gas borne particulate erosion wear of a high speed steel and a WC/Co hardmetal. For the hardmetal, it was observed that in the initial stage of wear although gross damage to the structure of the material occurred over the localised area corresponding to the impact area of the particle, some damage occurred over a much wider area. This was thought to be due to immediate fragmentation of impacting particles followed by scouring of the surface near the impact site by fragments. Further damage was caused by a combination of further removal of the binder phase, build up of plastic deformation in single WC grains followed by fracture and fragmentation and removal, and intergranular fracture of the WC skeleton. Multi-grain cracking was not observed, suggesting that models based on indentation cracking are not realistic. © 2003 Elsevier Science B.V. All rights reserved. Keywords: Wear; Erosion; Wear mechanisms; WC/Co; Hardmetals
1. Introduction WC based hardmetals and high speed tools are widely used in applications where erosion and abrasion resistance are important. Because of this, there has been considerable interest in measuring the wear performance of these materials in the laboratory, and developing an understanding of the mechanisms of wear that occur in abrasion and erosion [1–24]. These studies have often used tests that have been standardised so that comparisons in results between laboratories can be facilitated [25–27]. Often these studies have been combined with an evaluation of the microstructure and mechanical performance of the materials, so that information on how the wear resistance alters with changes in material make-up can be gathered. Attempts have also been made to model the wear so that equations linking the wear to material properties such as hardness, toughness and grain size can be developed [12,13,28–30].
∗
Corresponding author. Tel.: +44-208-943-6374; fax: +44-208-943-2989. E-mail address: [email protected] (M.G. Gee).
The aim of this type of approach is that more knowledge about the wear mechanisms can be used to guide the development of new materials, and also to provide better ways of selecting materials for particular applications. Despite this, the understanding of the wear mechanisms in abrasion and erosion is by no means complete. This paper describes a new method that can be used to determine the detailed wear mechanisms that take place in erosion and abrasion. This uses stepwise erosion or abrasion combined with relocation scanning electron microscopy. This preliminary report describes work that was carried out on gas borne particulate erosion. In this type of experiment, a stream of erodent particles borne in a gas flow impinges on a sample giving linear increases of wear for WC/Co hardmetals and high speed steels (Fig. 1).
2. Experimental method The essence of the method is to incrementally expose a sample to erosion from small quantities of erodent, and then be able to accurately relocate the test sample in the scanning electron microscope to enable the same area to be examined
0043-1648/03/$ – see front matter © 2003 Elsevier Science B.V. All rights reserved. doi:10.1016/S0043-1648(03)00090-5
M.G. Gee et al. / Wear 255 (2003) 44–54
45
Fig. 1. Typical erosion results for tool steels and hardmetals.
at high magnification, allowing for the development in damage to be followed at specific points on the sample surface. 2.1. Materials In this preliminary investigation, two materials were examined. The first material was a WC/Co hardmetal that had 11% of Co as binder, a 5.16 m mean linear intercept grain size, and a hardness of 912 HV30. The second material was an ASP30 high speed steel with a hardness of 850 HV30. The samples were 40 mm × 20 mm × 5 mm. One of the 40 mm × 20 mm faces of each sample was polished by standard metallographic techniques. Before testing the
polished face of the sample was cleaned by washing in detergent solution, rinsing in very hot water, and then blow drying with compressed air. After every exposure this cleaning procedure was repeated before examination in the SEM. 2.2. Erosion test system The erosion test system that was used was a conventional gas borne particulate erosion system (Fig. 2). The nozzle bore size was 5 mm and the nozzle length was 250 mm. Erodent is drawn into the gas stream through a venturi from a rotating table that controls the feed rate of erodent.
Fig. 2. Gas borne particulate erosion test system [27].
46
M.G. Gee et al. / Wear 255 (2003) 44–54
Table 1 Weight of aliquots used for stepwise erosion studies Aliquot number
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18
WC/Co
High speed steel
Aliquot weight (g)
Cumulative weight (g)
Aliquot weight (g)
Cumulative weight (g)
0.05 0.05 0.06 0.15 0.21 0.23 0.20 0.21 0.40 0.80 1.60 1.20 1.60 1.60 1.60 1.60 2 2
0.05 0.10 0.16 0.32 0.53 0.76 0.95 1.16 1.56 2.36 3.96 5.16 6.76 8.36 9.96 11.56 13.56 15.56
0.05 0.1 0.05 0.2 0.2 0.2 0.2 0.2 0.2 0.2 0.2
0.05 0.15 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8
The test conditions that were employed were: • Erodent of rounded quartz 220 m in size. • Particle velocity of 75 m s−1 . • Stand-off distance between end of nozzle and sample of 20 mm. • Angle of impingement 90◦ . The samples were exposed to aliquots of erodent at these conditions. In early experiments, estimates of the size of aliquot were calculated on the basis that the probability of contact of any point on the surface should be 50% for a single exposure, but it was found that the damage from a single exposure with this size of aliquot was too great. Instead an empirical approach was used with a very small initial aliquot that was increased in weight as the experiment proceeded. Table 1 shows the aliquots that were used on the two samples. At the start of the experiment, the samples were carefully placed in the erosion test system so that the area chosen for the repeated erosion was at the centre of the expected wear scar. The position of the sample was carefully marked so that the sample could be replaced in the same position in the erosion system. The sample was then exposed to the first aliquot of erodent. It was then taken from the test system, cleaned as described earlier, and examined in the SEM. After the appropriate images had been taken, the sample was exposed to a further aliquot of erodent. 2.3. Relocation scanning electron microscopy The relocation relied on marking the samples so that the same area on the sample could be found repeatedly. This was carried out by marking the samples with lines of Vickers indentations made with a 30 kg load is shown in Fig. 3.
Fig. 3. Sample marking technique.
A number of areas for the mechanistic study were chosen close to a specific marker indentation. Care was taken when inserting the sample in the scanning electron microscope to ensure that the orientation of the sample was the same after every exposure to erodent. As a secondary measure, the digital stage of the SEM was zeroed on a specific indentation and the position of the chosen areas measured with a nominal resolution of 1 m relative to the zero position. A high resolution, high brightness field emission scanning electron microscope was used for the imaging. A careful strategy was employed to enable the same area to be identified visually. Thus micrographs of the general area were taken at 500×, 1000×, and a montage of the working area was taken at 2000×. It was originally planned to use the digital framestore associated with the field emission SEM to acquire large pixel size images of the area of interest, but unfortunately the framestore proved inadequate to the task so that a return to the use of film proved essential.
3. Results A series of montages that resulted from the stepwise erosion experiments on the WC/Co sample is shown in Fig. 4. On the first exposure, considerable damage to the surface in the form of removal of much of the cobalt binder phase on the surface has already occurred. As exposure proceeds, cracking around individual WC grains occurs until the grains become unsupported. Removal of individual grains occurs as the exposure continues with a concomitant roughening of the sample surface. Examination of the exposed surfaces at higher magnification reveals more details of the wear mechanisms that are occurring (Figs. 5 and 6). In Fig. 5b, considerable evidence of plastic deformation through the presence of slip lines on the large triangular grain can be seen (A in Fig. 5). In this figure some of the larger WC grains can be seen to be cracked (B), with some cracks also occurring between grains. Fragmentation of a narrow sliver of WC grain has taken place towards the bottom right of the image (C), and grooving has occurred to the Co binder phase to the bottom of the image (D). Between Fig. 5d and e the large triangular grain has been removed, and it is perhaps significant that internal cracking can be seen at the base of this grain in Fig. 5d (E). In Fig. 5e material is smeared across some of the underlying WC grains, and a grain at the base of the image is
M.G. Gee et al. / Wear 255 (2003) 44–54
47
Fig. 4. Maps of relocated area on WC/Co sample. Cross marks same point on each image: (a) unexposed sample, 0.05 g cumulative exposure; (b) 0.53 g cumulative exposure; (c) 2.36 g cumulative exposure; (d) 6.76 g cumulative exposure; (e) 9.96 g cumulative exposure; (f) 15.56 g cumulative exposure.
48
M.G. Gee et al. / Wear 255 (2003) 44–54
Fig. 4. (Continued ).
M.G. Gee et al. / Wear 255 (2003) 44–54
49
Fig. 5. Area A in Fig. 4a: (a) unexposed sample; (b) 0.05 g exposure; (c) 0.53 g exposure; (d) 2.36 g exposure; (e) 3.96 g exposure; (f) 5.16 g exposure; (g) 6.76 g exposure; (h) 8.36 g exposure; (i) 9.96 g exposure; (j) 11.56 g exposure.
50
M.G. Gee et al. / Wear 255 (2003) 44–54
Fig. 5. (Continued ).
Fig. 6. Area B in Fig. 4a: (a) unexposed sample; (b) 0.05 g exposure; (c) 0.53 g exposure; (d) 2.36 g exposure; (e) 3.96 g exposure; (f) 6.76 g exposure; (g) 8.36 g exposure; (h) 9.96 g exposure; (i) 11.56 g exposure.
M.G. Gee et al. / Wear 255 (2003) 44–54
51
Fig. 6. (Continued ).
fragmented (F). In Fig. 5f the large grain at the top of the image has a large internal crack; in Fig. 5g this grain has been removed. Fig. 6 shows another area. Again, the initial exposure removes most of the surface binder from the sample surface. Progressive damage occurs to single grains of WC until they are removed from the sample. Cracks grow between the different WC grains also weakening their support. Often when grains are removed, other grains are revealed below. Thus a large grain below the triangular grain was progressively damaged in the wear until it was removed between Fig. 6e and f revealing another grain beneath the surface (H). Several grains are removed from the top of the image in Fig. 6h (I), leaving the large triangular grain proud of the surface— note also the near vertical large surface that is revealed to the base of the grain in Fig. 6h. This grain shows evidence for considerable plastic deformation and some cracking. The plastic deformation is visible both on the top face and the vertical face of this grain (J). The large grain is removed from the structure after the next exposure (Fig. 6i). The relocated area on the high speed steel is shown in Fig. 7. Some erosion damage had taken place to this area from an earlier erosion experiment to the side of the reloca-
tion area. Initially, much of the damage that is visible comes from relatively small discrete grooves. Close examination of these grooves shows that carbides in the microstructure of the steel often act to arrest grooves. However, as exposure continues areas of gross plastic deformation occur that spread masses of material across the surface of the steel (Fig. 7d). These areas are then subject to further plastic deformation that further damages the surface. Removal of material occurs when fragments of material are broken off layers of deformed material (Fig. 7e).
4. Discussion 4.1. Mechanisms of erosion The results of this preliminary study shed light on the mechanisms of wear that occur in the gas borne particulate erosion of WC/Co hardmetals and high speed steel. The stepwise erosion of the WC/Co sample clearly showed that wear occurred by the accumulation of damage, fracture and removal of single grains of WC. The events that lead up to this process are:
52
M.G. Gee et al. / Wear 255 (2003) 44–54
Fig. 7. High speed steel relocated area. Note area was subject to some damage from an adjacent erosion exposure before experiment started: (a) 0.051 g exposure; (b) 0.153 g exposure; (c) 0.204 g exposure; (e) 0.607 g exposure.
• Removal of binder phase from the surface layer of the sample. • Plastic grooving of binder phase. • Accumulation of plastic strain in WC grains. • Fracture and fragmentation of individual WC grains. • Cracking between WC grains. • Breakaway of unsupported WC grains. The removal of binder phase from the surface is very immediate. Much of the cobalt binder phase is removed even after only 0.05 g of erodent has been used. To put this figure in perspective, for the 200 m sand used in this study, the area that would be covered by the approximately 500 sand
particles in this first aliquot would only cover just over 20% of the area of the erosion scar (and the particles are not flat so only contact initially over a small area). This suggests that when a single sand particle contacts the surface it breaks up into many fragments that move across the surface gouging out binder phase as they proceed. The build up of plastic strain in individual WC grains often reaches the point at which fracture occurs, weakening and fragmenting the individual WC grain. In addition, the growth of cracks between one WC grain and the next breaks down the strong WC inter-grain network or skeleton in the material leading to a general weakening of the structure and increasing the likelihood of breakaway of WC grains.
M.G. Gee et al. / Wear 255 (2003) 44–54
No evidence for binder phase extrusion [9] was observed in this study, but it was not expected as the wear processes that are taking place would remove any binder phase that was forced to the surface. There was also no evidence at all for long multi-grain cracks on the exposed wear surface; this contradicts some theories of the wear of these materials that expect that wear occurs by the lateral crack mechanism that occurs in indentation [12,29,30]. If these cracks existed they would extend across many grains. In fact all the cracks that were observed were only a single grain or less in size. This confirms similar observations by other studies of the erosive wear of coarse grained hardmetals [1,8,9]. Erosion in the high speed steel was observed to occur by the expected mechanisms of plastic grooving and deformation, with eventual material loss through the fracture of heavily deformed layers of material on the exposed surface. The microstructure of the material seemed to have little effect on the erosion, with the possible exception that some grooves were caused by the erodent debris were clearly impeded by carbide particles. It should be noted that although the material removal mechanisms demonstrated to occur in this paper take place by a locally intermittent process, a macroscopic linear wear rate is observed because these local events are averaged over many hundreds of sites across the wear zone on the sample. 4.2. Possible extensions to this work There are several areas where more work could usefully be carried out. Thus when the exposure used in this study is compared with the typical exposures used in an erosion test (Fig. 1), it can be seen that the exposure in this study is very small and that if there was an incubation period for erosion, as has been reported [31], the experiments carried out here have all occurred in this initial period. Thus even at the end of the stepwise exposure experiments reported here, the original polished grains have only just been removed from the surface. An obvious extension of the experiments reported here is to perform stepwise erosion on a sample that has already been exposed to substantial erosion. This would confirm if the same set of mechanisms takes place when the surface has already been damaged compared to a pristine surface. It is also important to carry out the same type of study on a range of different hardmetals where the composition and structure is varied to include a range of grain sizes and cobalt binder fractions. The hardmetal used in this study was chosen specifically to give a large scale microstructure where the evolution of damage would be easier to follow. This certainly made the experiments relatively simple, but other experiments on finer grained hardmetals, particularly ultra-fine grained hardmetals would confirm the transition in wear mechanism that has been reported to occur from a process dominated by fracture in specific WC grains in the large grain sized material to a process where gross plastic
53
deformation is the key mechanism for fine grained materials [23,24,31,32]. Another useful extension of the work would be to carry out measurements of the surface form of the relocated area as exposure takes place. This can be carried out most conveniently using stereo images combined with subsequent image analysis. This would enable measurements of the volume loss due to wear to be monitored for the exposed area, and would enable models to be formulated relating the wear rate to the individual wear events.
5. Conclusions The stepwise erosion technique has proved to be an effective tool to investigate the mechanisms of wear that occur in WC/Co hardmetals and high speed steel. The preliminary results for the coarse grained hardmetal that was examined show that wear occurs by progressive damage process that involves: • Surface binder phase removal almost instantly. • Build up of considerable plastic deformation in WC grains. • When further plastic deformation cannot be accommodated, or when local impact forces are too high, individual grains fracture and fragment. • Growth of cracking between individual grains in the WC skeleton of the material. • Final breakaway of single WC grains. No evidence for large scale lateral cracking was found. In the high speed steel, there was some evidence for cracking and local grooving, but the main mechanism for removal of material was fracture of unsupported and grossly deformed layers of material on the surface.
Acknowledgements The authors would like to acknowledge the support of the materials measurement programme of DTI for the work reported here. References [1] J. Larsen-Basse, Wear of hard metals in rock drilling: a survey of the literature, Powder Metall. 16 (1973) 1–32. [2] J. Larsen-Basse, C.M. Shishido, P.A. Tanouye, Some features of abraded WC–Co surfaces, J. Aust. Inst. Met. 19 (1974) 271–275. [3] J. Larsen-Basse, C.M. Perrott, P.M. Robinson, Abrasive wear of WC–Co composites. I. Rotary drilling tests, Mater. Sci. Eng. 13 (1974) 83–91. [4] R.I. Blomberg, C.M. Perrot, The wear of WC–Co composites during abrasion on silicon carbide, Wear 19 (1974) 266–269. [5] R.I. Blomberry, C.M. Perrot, P.M. Robinson, Abrasive wear of tungsten carbide–cobalt composites. II. Wear mechanisms, Mater. Sci. Eng. 13 (1974) 93–100.
54
M.G. Gee et al. / Wear 255 (2003) 44–54
[6] H. Feld, P. Walter, Powder Metall. Int. 7 (1975) 188–190. [7] H. Conrad, D. McCabe, G.A. Sargent, Effects of microstructure on the erosion of WC–Co alloys, in: R.K. Viswanadham, D.J. Rowcliffe, J. Gurland (Eds.), Science of Hard Materials, Plenum Press, New York, 1981, pp. 775–796. [8] J. Larsen-Basse, Binder extrusion of sliding wear of WC–Co alloys, in: Proceedings of the International Conference on Wear of Materials, ASME, New York, 1985, pp. 39–44. [9] J. Larsen-Basse, Binder extrusion in sliding wear of WC–Co alloys, Wear 105 (1985) 247–256. [10] E.A. Almond, L.A. Lay, M.G. Gee, Comparison of wear mechanisms in ceramics and cemented carbides, in: Proceedings of the Second International Conference on Science of Hard Materials, Conference Series No. 75, Institute of Physics, Adam Hilger, 1986, p. 919. [11] J. Larsen-Basse, N. Devnani, Binder extrusion as the controlling mechanism in abrasion of WC–Co cemented carbides, in: Proceedings of the Second International Conference on Science of Hard Materials, Conference Series No. 75, Institute of Physics, 1986, pp. 883–895. [12] K. Anand, C. Morrison, R.O. Scattergood, H. Conrad, J.L. Routbort, R. Warren, Erosion of multiphase materials, Inst. Phys. Conf. Ser. 75 (1986) 949–961. [13] S.F. Wayne, J.G. Baldoni, S.T. Buljan, Abrasion and erosion of WC–Co with controlled microstructures, Tribol. Trans. 33 (1990) 611–617. [14] D.G.F. Quigley, S. Luyckz, M.N. James, An empirical ranking of a wide range of WC–Co grades in terms of their abrasion resistance measured by the ASTM standard B611-85 test, Int. J. Refract. Hard Mater. 15 (1997) 73–79. [15] M.G. Gee, A.J. Gant, L.P. Byrne, B. Roebuck, Abrasion and reciprocating wear of hardmetals and ceramics, NPL Report CMMT(A) 166, May 1999. [16] M.G. Gee, A.J. Gant, Rotating wheel abrasion tests on hardmetals and ceramics, Measurement Note CMMT (MN)46, National Physical Laboratory, Teddington, Middlesex, UK, 1999. ISSN 1366-4506. [17] M.G. Gee, L. Fang, B. Roebuck, Mechanisms of wear in the high stress abrasion of WC/Co hardmetals, in: Proceedings of the EURO PM 1999, 1999, pp. 319–325.
[18] V.A. Pugsley, C. Allen, Microstructure/property relationships in the slurry erosion of tungsten carbide–cobalt, Wear 225–229 (1999) 1017–1024. [19] V.A. Pugsely, C. Allen, Microstructure/property relationships in the cavitation erosion of tungsten carbide–cobalt, Wear 233–235 (1999) 93–103. [20] Z. Feng, A. Ball, The erosion of four materials using seven erodents— towards an understanding, Wear 233–235 (1999) 674–684. [21] X. Ma, R. Liu, D.Y. Li, Abrasive wear behavior of D2 tool steel with respect to load and sliding speed under dry sand/rubber wheel abrasion condition, Wear 241 (2000) 79–85. [22] A.J. Gant, M.G. Gee, Wear of tungsten carbide–cobalt hardmetals and hot isostatically pressed high speed steels under dry abrasive conditions, Wear 251 (2001) 908–915. [23] C. Allen, M. Sheen, J. Williams, V.A. Pugsley, The wear of ultrafine WC–Co hardmetals, Wear 250 (2001) 604–610. [24] U. Beste, L. Hammersstrom, H. Engqvist, S. Rimlinger, S. Jocabson, Particle erosion of cemented carbides with low Co content, Wear 250 (2001) 809–817. [25] ASTM B611-85, Standard test method for the abrasive wear resistance of cemented carbides, in: Annual Book of ASTM Standards, vol. 02.05, ASTM, 1995, pp. 326–327. [26] ASTM G65-94, Standard test for measuring abrasion using the dry sand/rubber wheel apparatus, in: Annual Book of ASTM Standards, vol. 03.02, ASTM, 1996. [27] ASTM G76-95, Standard test for measuring abrasion using the dry sand/rubber wheel apparatus, in: Annual Book of ASTM Standards, vol. 03.02, ASTM, 1996. [28] N. Axén, S. Jacobsen, A model for the abrasive wear resistance of multiphase materials, Wear 174 (1994) 187–199. [29] E.A. Almond, Indentation phenomena and the wear of surfaces and edges, Mater. Sci. Technol. 2 (1986) 641–646. [30] A.G. Evans, D.B. Marshall, Wear mechanisms in ceramics, in: D.A. Rigney (Ed.), Fundamentals of Friction and Wear of Materials, ASM, Cleveland, OH, 1980, pp. 439–452. [31] I.M. Hutchings, Tribology: Friction and Wear of Engineering Materials, Arnold, Paris, 1992. [32] K. Jia, T.E. Fischer, Wear 200 (1996) 206–214.
Stepwise erosion as a method for determining the mechanisms of wear in gas borne particulate erosion M.G. Gee∗ , R.H. Gee, I. McNaught National Physical Laboratory, Centre for Materials Measurement, Queens Road, Teddington, Middlesex TW11 0LW, UK
Abstract In erosion and abrasion, the detailed processes that cause material removal are still poorly understood. This means that with a few notable exceptions, good models for the prediction of materials behaviour are not still readily available. Stepwise testing is a new approach that has been recently developed as a way of providing information on the build up of damage in erosion. The technique uses a combination of exposing a test sample to repeated very low duration abrasion or erosion followed by accurate relocation in the SEM. This enables the real wear processes within a material to be followed in considerable detail, giving an excellent basis for developing new and improved models. In this preliminary study, the technique has been applied to the gas borne particulate erosion wear of a high speed steel and a WC/Co hardmetal. For the hardmetal, it was observed that in the initial stage of wear although gross damage to the structure of the material occurred over the localised area corresponding to the impact area of the particle, some damage occurred over a much wider area. This was thought to be due to immediate fragmentation of impacting particles followed by scouring of the surface near the impact site by fragments. Further damage was caused by a combination of further removal of the binder phase, build up of plastic deformation in single WC grains followed by fracture and fragmentation and removal, and intergranular fracture of the WC skeleton. Multi-grain cracking was not observed, suggesting that models based on indentation cracking are not realistic. © 2003 Elsevier Science B.V. All rights reserved. Keywords: Wear; Erosion; Wear mechanisms; WC/Co; Hardmetals
1. Introduction WC based hardmetals and high speed tools are widely used in applications where erosion and abrasion resistance are important. Because of this, there has been considerable interest in measuring the wear performance of these materials in the laboratory, and developing an understanding of the mechanisms of wear that occur in abrasion and erosion [1–24]. These studies have often used tests that have been standardised so that comparisons in results between laboratories can be facilitated [25–27]. Often these studies have been combined with an evaluation of the microstructure and mechanical performance of the materials, so that information on how the wear resistance alters with changes in material make-up can be gathered. Attempts have also been made to model the wear so that equations linking the wear to material properties such as hardness, toughness and grain size can be developed [12,13,28–30].
∗
Corresponding author. Tel.: +44-208-943-6374; fax: +44-208-943-2989. E-mail address: [email protected] (M.G. Gee).
The aim of this type of approach is that more knowledge about the wear mechanisms can be used to guide the development of new materials, and also to provide better ways of selecting materials for particular applications. Despite this, the understanding of the wear mechanisms in abrasion and erosion is by no means complete. This paper describes a new method that can be used to determine the detailed wear mechanisms that take place in erosion and abrasion. This uses stepwise erosion or abrasion combined with relocation scanning electron microscopy. This preliminary report describes work that was carried out on gas borne particulate erosion. In this type of experiment, a stream of erodent particles borne in a gas flow impinges on a sample giving linear increases of wear for WC/Co hardmetals and high speed steels (Fig. 1).
2. Experimental method The essence of the method is to incrementally expose a sample to erosion from small quantities of erodent, and then be able to accurately relocate the test sample in the scanning electron microscope to enable the same area to be examined
0043-1648/03/$ – see front matter © 2003 Elsevier Science B.V. All rights reserved. doi:10.1016/S0043-1648(03)00090-5
M.G. Gee et al. / Wear 255 (2003) 44–54
45
Fig. 1. Typical erosion results for tool steels and hardmetals.
at high magnification, allowing for the development in damage to be followed at specific points on the sample surface. 2.1. Materials In this preliminary investigation, two materials were examined. The first material was a WC/Co hardmetal that had 11% of Co as binder, a 5.16 m mean linear intercept grain size, and a hardness of 912 HV30. The second material was an ASP30 high speed steel with a hardness of 850 HV30. The samples were 40 mm × 20 mm × 5 mm. One of the 40 mm × 20 mm faces of each sample was polished by standard metallographic techniques. Before testing the
polished face of the sample was cleaned by washing in detergent solution, rinsing in very hot water, and then blow drying with compressed air. After every exposure this cleaning procedure was repeated before examination in the SEM. 2.2. Erosion test system The erosion test system that was used was a conventional gas borne particulate erosion system (Fig. 2). The nozzle bore size was 5 mm and the nozzle length was 250 mm. Erodent is drawn into the gas stream through a venturi from a rotating table that controls the feed rate of erodent.
Fig. 2. Gas borne particulate erosion test system [27].
46
M.G. Gee et al. / Wear 255 (2003) 44–54
Table 1 Weight of aliquots used for stepwise erosion studies Aliquot number
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18
WC/Co
High speed steel
Aliquot weight (g)
Cumulative weight (g)
Aliquot weight (g)
Cumulative weight (g)
0.05 0.05 0.06 0.15 0.21 0.23 0.20 0.21 0.40 0.80 1.60 1.20 1.60 1.60 1.60 1.60 2 2
0.05 0.10 0.16 0.32 0.53 0.76 0.95 1.16 1.56 2.36 3.96 5.16 6.76 8.36 9.96 11.56 13.56 15.56
0.05 0.1 0.05 0.2 0.2 0.2 0.2 0.2 0.2 0.2 0.2
0.05 0.15 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8
The test conditions that were employed were: • Erodent of rounded quartz 220 m in size. • Particle velocity of 75 m s−1 . • Stand-off distance between end of nozzle and sample of 20 mm. • Angle of impingement 90◦ . The samples were exposed to aliquots of erodent at these conditions. In early experiments, estimates of the size of aliquot were calculated on the basis that the probability of contact of any point on the surface should be 50% for a single exposure, but it was found that the damage from a single exposure with this size of aliquot was too great. Instead an empirical approach was used with a very small initial aliquot that was increased in weight as the experiment proceeded. Table 1 shows the aliquots that were used on the two samples. At the start of the experiment, the samples were carefully placed in the erosion test system so that the area chosen for the repeated erosion was at the centre of the expected wear scar. The position of the sample was carefully marked so that the sample could be replaced in the same position in the erosion system. The sample was then exposed to the first aliquot of erodent. It was then taken from the test system, cleaned as described earlier, and examined in the SEM. After the appropriate images had been taken, the sample was exposed to a further aliquot of erodent. 2.3. Relocation scanning electron microscopy The relocation relied on marking the samples so that the same area on the sample could be found repeatedly. This was carried out by marking the samples with lines of Vickers indentations made with a 30 kg load is shown in Fig. 3.
Fig. 3. Sample marking technique.
A number of areas for the mechanistic study were chosen close to a specific marker indentation. Care was taken when inserting the sample in the scanning electron microscope to ensure that the orientation of the sample was the same after every exposure to erodent. As a secondary measure, the digital stage of the SEM was zeroed on a specific indentation and the position of the chosen areas measured with a nominal resolution of 1 m relative to the zero position. A high resolution, high brightness field emission scanning electron microscope was used for the imaging. A careful strategy was employed to enable the same area to be identified visually. Thus micrographs of the general area were taken at 500×, 1000×, and a montage of the working area was taken at 2000×. It was originally planned to use the digital framestore associated with the field emission SEM to acquire large pixel size images of the area of interest, but unfortunately the framestore proved inadequate to the task so that a return to the use of film proved essential.
3. Results A series of montages that resulted from the stepwise erosion experiments on the WC/Co sample is shown in Fig. 4. On the first exposure, considerable damage to the surface in the form of removal of much of the cobalt binder phase on the surface has already occurred. As exposure proceeds, cracking around individual WC grains occurs until the grains become unsupported. Removal of individual grains occurs as the exposure continues with a concomitant roughening of the sample surface. Examination of the exposed surfaces at higher magnification reveals more details of the wear mechanisms that are occurring (Figs. 5 and 6). In Fig. 5b, considerable evidence of plastic deformation through the presence of slip lines on the large triangular grain can be seen (A in Fig. 5). In this figure some of the larger WC grains can be seen to be cracked (B), with some cracks also occurring between grains. Fragmentation of a narrow sliver of WC grain has taken place towards the bottom right of the image (C), and grooving has occurred to the Co binder phase to the bottom of the image (D). Between Fig. 5d and e the large triangular grain has been removed, and it is perhaps significant that internal cracking can be seen at the base of this grain in Fig. 5d (E). In Fig. 5e material is smeared across some of the underlying WC grains, and a grain at the base of the image is
M.G. Gee et al. / Wear 255 (2003) 44–54
47
Fig. 4. Maps of relocated area on WC/Co sample. Cross marks same point on each image: (a) unexposed sample, 0.05 g cumulative exposure; (b) 0.53 g cumulative exposure; (c) 2.36 g cumulative exposure; (d) 6.76 g cumulative exposure; (e) 9.96 g cumulative exposure; (f) 15.56 g cumulative exposure.
48
M.G. Gee et al. / Wear 255 (2003) 44–54
Fig. 4. (Continued ).
M.G. Gee et al. / Wear 255 (2003) 44–54
49
Fig. 5. Area A in Fig. 4a: (a) unexposed sample; (b) 0.05 g exposure; (c) 0.53 g exposure; (d) 2.36 g exposure; (e) 3.96 g exposure; (f) 5.16 g exposure; (g) 6.76 g exposure; (h) 8.36 g exposure; (i) 9.96 g exposure; (j) 11.56 g exposure.
50
M.G. Gee et al. / Wear 255 (2003) 44–54
Fig. 5. (Continued ).
Fig. 6. Area B in Fig. 4a: (a) unexposed sample; (b) 0.05 g exposure; (c) 0.53 g exposure; (d) 2.36 g exposure; (e) 3.96 g exposure; (f) 6.76 g exposure; (g) 8.36 g exposure; (h) 9.96 g exposure; (i) 11.56 g exposure.
M.G. Gee et al. / Wear 255 (2003) 44–54
51
Fig. 6. (Continued ).
fragmented (F). In Fig. 5f the large grain at the top of the image has a large internal crack; in Fig. 5g this grain has been removed. Fig. 6 shows another area. Again, the initial exposure removes most of the surface binder from the sample surface. Progressive damage occurs to single grains of WC until they are removed from the sample. Cracks grow between the different WC grains also weakening their support. Often when grains are removed, other grains are revealed below. Thus a large grain below the triangular grain was progressively damaged in the wear until it was removed between Fig. 6e and f revealing another grain beneath the surface (H). Several grains are removed from the top of the image in Fig. 6h (I), leaving the large triangular grain proud of the surface— note also the near vertical large surface that is revealed to the base of the grain in Fig. 6h. This grain shows evidence for considerable plastic deformation and some cracking. The plastic deformation is visible both on the top face and the vertical face of this grain (J). The large grain is removed from the structure after the next exposure (Fig. 6i). The relocated area on the high speed steel is shown in Fig. 7. Some erosion damage had taken place to this area from an earlier erosion experiment to the side of the reloca-
tion area. Initially, much of the damage that is visible comes from relatively small discrete grooves. Close examination of these grooves shows that carbides in the microstructure of the steel often act to arrest grooves. However, as exposure continues areas of gross plastic deformation occur that spread masses of material across the surface of the steel (Fig. 7d). These areas are then subject to further plastic deformation that further damages the surface. Removal of material occurs when fragments of material are broken off layers of deformed material (Fig. 7e).
4. Discussion 4.1. Mechanisms of erosion The results of this preliminary study shed light on the mechanisms of wear that occur in the gas borne particulate erosion of WC/Co hardmetals and high speed steel. The stepwise erosion of the WC/Co sample clearly showed that wear occurred by the accumulation of damage, fracture and removal of single grains of WC. The events that lead up to this process are:
52
M.G. Gee et al. / Wear 255 (2003) 44–54
Fig. 7. High speed steel relocated area. Note area was subject to some damage from an adjacent erosion exposure before experiment started: (a) 0.051 g exposure; (b) 0.153 g exposure; (c) 0.204 g exposure; (e) 0.607 g exposure.
• Removal of binder phase from the surface layer of the sample. • Plastic grooving of binder phase. • Accumulation of plastic strain in WC grains. • Fracture and fragmentation of individual WC grains. • Cracking between WC grains. • Breakaway of unsupported WC grains. The removal of binder phase from the surface is very immediate. Much of the cobalt binder phase is removed even after only 0.05 g of erodent has been used. To put this figure in perspective, for the 200 m sand used in this study, the area that would be covered by the approximately 500 sand
particles in this first aliquot would only cover just over 20% of the area of the erosion scar (and the particles are not flat so only contact initially over a small area). This suggests that when a single sand particle contacts the surface it breaks up into many fragments that move across the surface gouging out binder phase as they proceed. The build up of plastic strain in individual WC grains often reaches the point at which fracture occurs, weakening and fragmenting the individual WC grain. In addition, the growth of cracks between one WC grain and the next breaks down the strong WC inter-grain network or skeleton in the material leading to a general weakening of the structure and increasing the likelihood of breakaway of WC grains.
M.G. Gee et al. / Wear 255 (2003) 44–54
No evidence for binder phase extrusion [9] was observed in this study, but it was not expected as the wear processes that are taking place would remove any binder phase that was forced to the surface. There was also no evidence at all for long multi-grain cracks on the exposed wear surface; this contradicts some theories of the wear of these materials that expect that wear occurs by the lateral crack mechanism that occurs in indentation [12,29,30]. If these cracks existed they would extend across many grains. In fact all the cracks that were observed were only a single grain or less in size. This confirms similar observations by other studies of the erosive wear of coarse grained hardmetals [1,8,9]. Erosion in the high speed steel was observed to occur by the expected mechanisms of plastic grooving and deformation, with eventual material loss through the fracture of heavily deformed layers of material on the exposed surface. The microstructure of the material seemed to have little effect on the erosion, with the possible exception that some grooves were caused by the erodent debris were clearly impeded by carbide particles. It should be noted that although the material removal mechanisms demonstrated to occur in this paper take place by a locally intermittent process, a macroscopic linear wear rate is observed because these local events are averaged over many hundreds of sites across the wear zone on the sample. 4.2. Possible extensions to this work There are several areas where more work could usefully be carried out. Thus when the exposure used in this study is compared with the typical exposures used in an erosion test (Fig. 1), it can be seen that the exposure in this study is very small and that if there was an incubation period for erosion, as has been reported [31], the experiments carried out here have all occurred in this initial period. Thus even at the end of the stepwise exposure experiments reported here, the original polished grains have only just been removed from the surface. An obvious extension of the experiments reported here is to perform stepwise erosion on a sample that has already been exposed to substantial erosion. This would confirm if the same set of mechanisms takes place when the surface has already been damaged compared to a pristine surface. It is also important to carry out the same type of study on a range of different hardmetals where the composition and structure is varied to include a range of grain sizes and cobalt binder fractions. The hardmetal used in this study was chosen specifically to give a large scale microstructure where the evolution of damage would be easier to follow. This certainly made the experiments relatively simple, but other experiments on finer grained hardmetals, particularly ultra-fine grained hardmetals would confirm the transition in wear mechanism that has been reported to occur from a process dominated by fracture in specific WC grains in the large grain sized material to a process where gross plastic
53
deformation is the key mechanism for fine grained materials [23,24,31,32]. Another useful extension of the work would be to carry out measurements of the surface form of the relocated area as exposure takes place. This can be carried out most conveniently using stereo images combined with subsequent image analysis. This would enable measurements of the volume loss due to wear to be monitored for the exposed area, and would enable models to be formulated relating the wear rate to the individual wear events.
5. Conclusions The stepwise erosion technique has proved to be an effective tool to investigate the mechanisms of wear that occur in WC/Co hardmetals and high speed steel. The preliminary results for the coarse grained hardmetal that was examined show that wear occurs by progressive damage process that involves: • Surface binder phase removal almost instantly. • Build up of considerable plastic deformation in WC grains. • When further plastic deformation cannot be accommodated, or when local impact forces are too high, individual grains fracture and fragment. • Growth of cracking between individual grains in the WC skeleton of the material. • Final breakaway of single WC grains. No evidence for large scale lateral cracking was found. In the high speed steel, there was some evidence for cracking and local grooving, but the main mechanism for removal of material was fracture of unsupported and grossly deformed layers of material on the surface.
Acknowledgements The authors would like to acknowledge the support of the materials measurement programme of DTI for the work reported here. References [1] J. Larsen-Basse, Wear of hard metals in rock drilling: a survey of the literature, Powder Metall. 16 (1973) 1–32. [2] J. Larsen-Basse, C.M. Shishido, P.A. Tanouye, Some features of abraded WC–Co surfaces, J. Aust. Inst. Met. 19 (1974) 271–275. [3] J. Larsen-Basse, C.M. Perrott, P.M. Robinson, Abrasive wear of WC–Co composites. I. Rotary drilling tests, Mater. Sci. Eng. 13 (1974) 83–91. [4] R.I. Blomberg, C.M. Perrot, The wear of WC–Co composites during abrasion on silicon carbide, Wear 19 (1974) 266–269. [5] R.I. Blomberry, C.M. Perrot, P.M. Robinson, Abrasive wear of tungsten carbide–cobalt composites. II. Wear mechanisms, Mater. Sci. Eng. 13 (1974) 93–100.
54
M.G. Gee et al. / Wear 255 (2003) 44–54
[6] H. Feld, P. Walter, Powder Metall. Int. 7 (1975) 188–190. [7] H. Conrad, D. McCabe, G.A. Sargent, Effects of microstructure on the erosion of WC–Co alloys, in: R.K. Viswanadham, D.J. Rowcliffe, J. Gurland (Eds.), Science of Hard Materials, Plenum Press, New York, 1981, pp. 775–796. [8] J. Larsen-Basse, Binder extrusion of sliding wear of WC–Co alloys, in: Proceedings of the International Conference on Wear of Materials, ASME, New York, 1985, pp. 39–44. [9] J. Larsen-Basse, Binder extrusion in sliding wear of WC–Co alloys, Wear 105 (1985) 247–256. [10] E.A. Almond, L.A. Lay, M.G. Gee, Comparison of wear mechanisms in ceramics and cemented carbides, in: Proceedings of the Second International Conference on Science of Hard Materials, Conference Series No. 75, Institute of Physics, Adam Hilger, 1986, p. 919. [11] J. Larsen-Basse, N. Devnani, Binder extrusion as the controlling mechanism in abrasion of WC–Co cemented carbides, in: Proceedings of the Second International Conference on Science of Hard Materials, Conference Series No. 75, Institute of Physics, 1986, pp. 883–895. [12] K. Anand, C. Morrison, R.O. Scattergood, H. Conrad, J.L. Routbort, R. Warren, Erosion of multiphase materials, Inst. Phys. Conf. Ser. 75 (1986) 949–961. [13] S.F. Wayne, J.G. Baldoni, S.T. Buljan, Abrasion and erosion of WC–Co with controlled microstructures, Tribol. Trans. 33 (1990) 611–617. [14] D.G.F. Quigley, S. Luyckz, M.N. James, An empirical ranking of a wide range of WC–Co grades in terms of their abrasion resistance measured by the ASTM standard B611-85 test, Int. J. Refract. Hard Mater. 15 (1997) 73–79. [15] M.G. Gee, A.J. Gant, L.P. Byrne, B. Roebuck, Abrasion and reciprocating wear of hardmetals and ceramics, NPL Report CMMT(A) 166, May 1999. [16] M.G. Gee, A.J. Gant, Rotating wheel abrasion tests on hardmetals and ceramics, Measurement Note CMMT (MN)46, National Physical Laboratory, Teddington, Middlesex, UK, 1999. ISSN 1366-4506. [17] M.G. Gee, L. Fang, B. Roebuck, Mechanisms of wear in the high stress abrasion of WC/Co hardmetals, in: Proceedings of the EURO PM 1999, 1999, pp. 319–325.
[18] V.A. Pugsley, C. Allen, Microstructure/property relationships in the slurry erosion of tungsten carbide–cobalt, Wear 225–229 (1999) 1017–1024. [19] V.A. Pugsely, C. Allen, Microstructure/property relationships in the cavitation erosion of tungsten carbide–cobalt, Wear 233–235 (1999) 93–103. [20] Z. Feng, A. Ball, The erosion of four materials using seven erodents— towards an understanding, Wear 233–235 (1999) 674–684. [21] X. Ma, R. Liu, D.Y. Li, Abrasive wear behavior of D2 tool steel with respect to load and sliding speed under dry sand/rubber wheel abrasion condition, Wear 241 (2000) 79–85. [22] A.J. Gant, M.G. Gee, Wear of tungsten carbide–cobalt hardmetals and hot isostatically pressed high speed steels under dry abrasive conditions, Wear 251 (2001) 908–915. [23] C. Allen, M. Sheen, J. Williams, V.A. Pugsley, The wear of ultrafine WC–Co hardmetals, Wear 250 (2001) 604–610. [24] U. Beste, L. Hammersstrom, H. Engqvist, S. Rimlinger, S. Jocabson, Particle erosion of cemented carbides with low Co content, Wear 250 (2001) 809–817. [25] ASTM B611-85, Standard test method for the abrasive wear resistance of cemented carbides, in: Annual Book of ASTM Standards, vol. 02.05, ASTM, 1995, pp. 326–327. [26] ASTM G65-94, Standard test for measuring abrasion using the dry sand/rubber wheel apparatus, in: Annual Book of ASTM Standards, vol. 03.02, ASTM, 1996. [27] ASTM G76-95, Standard test for measuring abrasion using the dry sand/rubber wheel apparatus, in: Annual Book of ASTM Standards, vol. 03.02, ASTM, 1996. [28] N. Axén, S. Jacobsen, A model for the abrasive wear resistance of multiphase materials, Wear 174 (1994) 187–199. [29] E.A. Almond, Indentation phenomena and the wear of surfaces and edges, Mater. Sci. Technol. 2 (1986) 641–646. [30] A.G. Evans, D.B. Marshall, Wear mechanisms in ceramics, in: D.A. Rigney (Ed.), Fundamentals of Friction and Wear of Materials, ASM, Cleveland, OH, 1980, pp. 439–452. [31] I.M. Hutchings, Tribology: Friction and Wear of Engineering Materials, Arnold, Paris, 1992. [32] K. Jia, T.E. Fischer, Wear 200 (1996) 206–214.