Abrasive micro-blasting to improve surface integrity of electrical discharge machined WC Co composite

Transcription

1 Journal of Materials Processing Technology 166 (2005) Abrasive micro-blasting to improve surface integrity of electrical discharge machined WC Co composite Jun Qu a,, Albert J. Shih b, Ronald O. Scattergood c, Jie Luo b a Metals and Ceramics Division, Oak Ridge National Laboratory, P.O. Box 2008, MS 6063, Oak Ridge, TN , USA b Department of Mechanical Engineering, University of Michigan, Ann Arbor, MI 48109, USA c Department of Materials Science and Engineering, North Carolina State University, Raleigh, NC 27695, USA Received 20 November 2002; accepted 15 September 2004 Abstract This study investigates the improvement of surface integrity of wire electrical discharge machined (EDM) WC Co composite by abrasive micro-blasting. The thermally damaged recast layer generated by EDM has craters, cracks, and bubbles, which deteriorate the surface mechanical properties. The micro-blasting, using 6 12 and 4 20 m size SiC abrasive, enables the removal of the recast layer and is suitable for micro mechanical components. The surface roughness of EDM rough cut WC Co parts was improved significantly, with the average surface roughness (R a ) dropping down from 1.3 to 0.7 m. Scanning electron microscope (SEM) was used to examine the evolution change of surface texture and subsurface cross-section of EDM WC Co workpiece. The SEM micrographs showed that the recast layer was removed efficiently. After 5 s of micro-blasting, surface textures with ridge and cavity patterns were observed on fine and rough cut EDM surfaces, respectively. These surface textures could be correlated to the surface roughness measurement and crater formation in EDM spark erosion. A series of erosion wear experiment was conducted to quantify the weight reduction, calculate the erosion wear rate, and identify the wear mechanism Elsevier B.V. All rights reserved. Keywords: Micro-blasting; EDM; Surface integrity 1. Introduction Electrical discharge machining (EDM) is a thermoelectric process that erodes workpiece material by a series of discrete electrical sparks between the workpiece and electrode. Unlike traditional cutting and grinding processes which rely on a much harder tool or abrasive material to remove the softer work-material, the EDM process utilizes electrical sparks or thermal energy to erode the unwanted workmaterial and generate the desired shape. The hardness and strength of the work-materials are no longer the dominating factors that affect the tool wear and hinder the machining efficiency. This makes EDM particularly suitable for machining hard, difficult-to-machine materials, such as the metal matrix Corresponding author. Tel.: address: qujn@ornl.gov (J. Qu). composites (MMC). As shown in Fig. 1, the wire EDM process uses a traveling wire as the electrode to erode a groove in the workpiece. The close-up view of the gap and electrical sparks between the wire and workpiece is illustrated in Fig. 1(b). Mechanical properties of surface layers generated in this process are important to the part performance, particularly the fatigue and wear life. Although EDM can machine precise, complex, and intricate MMC components [1 4], the undesired surface recast layer and sub-surface heat-affected zone (HA) are generated [5]. Fig. 2 shows SEM micrographs of the surfaces and subsurface cross-sections of a WC Co sample machined under 14 and 2 s pulse on-time and 28 s spark cycle. Pulse ontime is defined as the duration of high voltage in the electrical spark cycle, which is an important EDM process parameter that affects the thicknesses of the recast layer and heataffected zone [2]. Electrical sparks melt the work-material /$ see front matter 2004 Elsevier B.V. All rights reserved. doi: /j.jmatprotec

2 J. Qu et al. / Journal of Materials Processing Technology 166 (2005) Fig. 1. Illustration of the wire EDM process and the gap and continuous electrical sparks. and generate craters, bubbles, and cracks in the recast layer, marked as RL in Fig. 2. A sub-surface heat-affected zone has essentially no porosity and exists between the recast layer and the bulk material. As shown in Fig. 2(a), the longer (14 s) pulse on-time creates thicker recast layer and heat-affected zone. The surface integrity, including the roughness, size of craters, and thicknesses of the recast layer and heat-affected zone, of a wire EDMed WC Co surface has been investigated [2]. Nanoindentation, EDS X-ray, and X-ray diffraction have been applied to study the mechanical properties (hardness and modulus of elasticity) and material compositions of the EDM WC Co surface layers [6]. Nanoindentation and SEM results show the heat-affected zone has more compact microstructure and higher hardness and modulus of elasticity than that of the bulk material [6]. The recast layer with craters, cracks, and bubbles deteriorates the surface properties and reduced the wear and fatigue life of EDM mechanical parts [2 6]. In this study, the abrasive microblasting process was applied to remove the damaged surface layers and improve the surface integrity of WC Co parts machined by wire EDM. Although the erosion properties of WC Co bulk material have been researched extensively [7 15], there is a lack of study on the EDMed WC Co surfaces. Fig. 2. SEM micrographs of the surface and cross-section of a WC Co part machined by wire EDM (CR: crater, RL: recast layer, and HA: heat-affected zone).

3 442 J. Qu et al. / Journal of Materials Processing Technology 166 (2005) Fig. 3. SiC micro-blasting of a cylindrical wire EDM WC Co micro shaft and the surface before and after blasting. Abrasive blasting has been used extensively for surface treatment. This process has been applied to increase the surface roughness for higher coating adhesion strength [16]. With proper process setup, the abrasive blasting can remove damaged EDM surface layers and introduce a new surface layer with desirable compressive residual stresses [11,17 22]. Applications of blasting process to modify EDM surface layers have been investigated by Lee and Zhang [20] for Sylon 501, Deng and Lee [21] for Al 2 O 3 /TiC and Si 3 N 4 /TiC, and Lee and Deng [22] for Al 2 O 3 /TiC/Mo/Ni using and m size Al 2 O 3 abrasive. In this study, SiC abrasive with smaller particle size was used. Micro-blasting using fine abrasive media is particularly suitable for microfabrication and surface treatment of micro mechanical components [23]. In contrast to conventional erosion by large-scale particles, no strength degradation occurs on the micro-blasted surfaces of ceramic materials [24]. The EDM process, which generates small cutting force, is ideal to machine micro features [25,26]. The fine abrasive carried by adequately high air pressure enables the micro-blasting process to modify the EDM surface on micro mechanical components. An example of the application of SiC micro-blasting of a micro-shaft is illustrated in Fig. 3. Fig. 3(a) shows the SEM micrograph of a 100 m diameter WC Co micro-shaft machined by the cylindrical wire EDM process [1,2]. The EDM surface and its closeup view, which exhibit cracks, bubbles, and other thermal damages, are shown in Fig. 3(b). This surface is blasted by the 6 12 m SiC particle at MPa air pressure (about 100 m/s particle speed) for 5 s. As shown in Fig. 3(c), the surface is modified and exhibits more uniform texture after 5 s of blasting. This paper presents a systematic study on micro-blasting of EDMed WC Co surfaces using fine SiC abrasive. Results of the surface roughness, surface texture generated during blasting, and erosion wear rate are discussed. 2. Sample preparation and experimental setup The work-material used in this study is the WC Co composite, which consists of 1 m size WC particle in 10 wt% Co matrix. This material has 92 Rc hardness, 3.4 MPa transverse rupture stress, and 14.5 specific density. The WC Co workmaterial was cut in a Brother HS-5100 wire EDM machine using 0.25 mm diameter brass wire. Two types of EDM WC Co surfaces, designated as rough cut and fine cut, were prepared for blasting experiments. Fig. 2(a and b) shows SEM micrographs of the rough cut and fine cut surface, arithmetic average surface roughness (R a ), and subsurface cross-section. The process parameters and material removal rate of the rough and fine cuts are summarized in Table 1. Under the same spark cycle (28 s), the rough cut has longer pulse on-time (14 s), larger crater, rougher surface, higher material removal rate, and thicker recast layer and heat-affected zone. Blasting and erosion wear experiments were conducted on a Mangum blasting machine, manufactured by Hess & Associates Inc. The machine used a venturi-type blasting gun with a 6.35 mm inner diameter Al 2 O 3 nozzle. Since the duration of blasting was short and limited number of experiments was conducted, the nozzle wear was not a prob-

4 J. Qu et al. / Journal of Materials Processing Technology 166 (2005) Fig. 4. Two types of SiC abrasive used in this study. lem. A time-control device was build to enable the blasting with a fixed period of time. The SiC abrasive was selected due to its high hardness. Two types of SiC abrasive, designated as Abrasives A and B, were used. Abrasive A, as shown in Fig. 4(a), was a mixture of particle sizes of 4 8, 6 12, and m, which corresponded to 1200, 1000, and 800 ANSI mesh, respectively. As shown in Fig. 4(b), Abrasive B was only 1000 ANSI mesh with particle sizes in the range 6 12 m. All tests were conducted at 90 incident angle (normal incidence). The air pressure of and MPa was selected. Using the two-disk method [27], the particle speed for Abrasives A at MPa was about 100 m/s. Particle speed at MPa was expected to be higher; however, the apparatus used in this study was unable to accurately measure the high particle speed at this pressure. Two sets of experiments were performed. One was focused on measurements and observations of surface integrity and roughness. Three blasting tests with a 5 s increment in blasting time were conducted. The other set of experiment was aimed at investigating the weight reduction and erosion wear rate. Three tests with 10 s blasting time increment were carried out. Table 2 lists process parameters used in the two sets of experiments. In summary, 24 ( ) tests were carried out in the surface integrity experiment and 6 (2 3) Table 1 Machine setup and process parameters and results of wire EDM of WC Co Process parameters Rough cut Fine cut Spark cycle ( s) Pulse on-time ( s) 14 2 Axial wire speed (mm/s) Wire tension (N) Gap voltage (V) Material removal rate (mm 3 /min) 5.5 a 0.16 b a Cutting a 6.35 mm thick plate. b Removing a 50 m surface layer. tests were conducted in the erosion wear rate experiment. Results of these experiments are presented in the following two sections. 3. Results and discussion 3.1. Surface roughness Fig. 5 shows results of arithmetic average surface roughness R a for the 24 tests conducted in the surface integrity and roughness experiment. Before the blasting test (0 s), as shown in Figs. 2 and 5, R a equals 1.34 and 0.83 m on the rough cut and fine cut surface, respectively. The surface finish was improved over time and eventually reached a stationary level, when the recast layer and heat-affected zone had been removed. The trend to reach steady-state surface roughness can be seen in Fig. 5. For Abrasives A and B, the R a is approaching 0.7 m at longer blasting times. The Abrasive A with larger size particles and/or high air pressure (0.276 MPa) removed the recast layer faster and reduced the surface roughness of rough cut parts more efficiently, as shown in roughness charts (Fig. 5) and SEM micrographs (Figs. 6 9). The EDM fine cut surface may be deteriorated by very short blasting time (5 s) and reached the stationary level at longer blasting time (15 s) using Abrasive A and high air pressure, as shown in Fig. 6. Table 2 Process parameters used in two blasting experiments Experiments Surface integrity test Erosion wear rate test Abrasive A and B A EDM surface Rough and fine cut Rough cut Air pressure (MPa) 0.138, , Blasting time (s) 5, 10, 15 10, 20, 30

5 444 J. Qu et al. / Journal of Materials Processing Technology 166 (2005) Fig. 5. Surface roughness on EDM WC Co using SiC micro-blasting Surface layers The surfaces of the wire EDMed WC Co parts had the surface recast layer and sub-surface heat-affected zone, generated by the high local temperature during machining. The recast layer, also called thermally damaged layer, had craters and thermal cracks and bubbles, as shown in Fig. 2, which deteriorates the surface finish and reduced the wear resistance and fatigue life of the EDMed parts. SEM was used to examine the surfaces and sub-surfaces of the WC Co parts to investigate the effect of the microblasting. The selected SEM micrographs, shown in Figs. 6 9, indicated that the recast layer was eventually removed after 5 and 10 s for EDM fine cut and rough cut parts, respectively. Longer blasting duration also removed the heataffected zone and reached a steady-state material removal in the bulk WC Co material Surface texture SEM micrographs of the fine cut surfaces blasted by Abrasive A at MPa pressure are shown in Fig. 6. As shown in Fig. 6(a), after blasting by 5 s, ridges that manifest crater edges observed in Fig. 2(b) can be seen. During the spark erosion, the molten material was splashed and accumulated to form the crater edges. Micro-blasting erodes the damaged recast layer and exposes the crater edges as the ridge pattern after the first 5 s of blasting. Evidently, these ridges contributed Fig. 6. SEM micrographs of surfaces and cross-sections of the fine cut EDM WC Co blasted by Abrasive A at MPa air pressure after first three blasting of 5 s increment.

6 J. Qu et al. / Journal of Materials Processing Technology 166 (2005) Fig. 7. Close-up view of surface of the fine cut EDM WC CO blasted by Abrasive A at MPa air pressure after 10 s blasting time. to the temporary increase in R a at the beginning of blasting, i.e., the maximum in the curve in Fig. 5(a) (open triangles). As shown in Fig. 6(b and c), these ridges gradually disappeared after 10 and 15 s of blasting. The cross-section surface showed the recast layer was removed after 5 s of blasting, only a shallow residual of heat-affected zone remained. After 10 s of blasting, the heat-affected zone was also removed. A similar ridge pattern could be seen on all of the blasted fine cut surfaces, though not necessarily as pronounced as that seen in Fig. 6. Close-up views of Fig. 6(b) are shown in Fig. 7. The boundaries between the 1 m size WC grains are not easily identified. Furthermore, there are no distinct cleavage facets indicative of individual brittle fracture events. Since the erodent particles have sizes in the range of m for Abrasive A, one could expect that erosion impact events at the high velocities used would overlap a large number of WC grains. The erosion mechanism for the blasting conditions used is therefore a composite-like removal of multiple WC grains per impact facilitated by plastic flow in the surrounding Co matrix. This is in contrast to the case reported by Anand and Conrad [13] where small erodent particles at low velocity produce individual impacts in single WC grain and a clear brittle-like erosion response. SEM micrographs of the surface and subsurface crosssection under the same blasting condition as in Figs. 6 and 7, Fig. 8. SEM micrographs of surfaces and cross-sections of the rough cut EDM WC CO blasted by Abrasive A at MPa air pressure after first three blasting of 5 s increment.

7 446 J. Qu et al. / Journal of Materials Processing Technology 166 (2005) Fig. 9. Close-up view of surface of the rough cut EDM WC CO blasted by Abrasive A at MPa air pressure after 10 s blasting time. Fig. 10. Erosion wear rate of SiC micro-blasting on wire EDM WC Co. but on the rough cut surfaces, are shown in Figs. 8 and 9. As shown in Fig. 8(b and c), a surface texture with many small, less than 1 m size cavities that follow the pattern of crater edges, as seen in Fig. 2(a), can be observed. The 2 m tip radius diamond stylus for the profilometer surface roughness measurement would not detect these cavities and the R a results in Fig. 5 would not be affected. Fig. 9 shows close-up views of Fig. 8(b). From the overall pattern of distribution of these features, it can be assumed that the source of the cavities is also related to the crater formation during EDM spark erosion. For the rough cut condition, the energy density in each spark erosion was higher and more molten material was generated in each spark erosion. Bubbles and cavities on the rough cut surfaces (Fig. 2(a)) are therefore larger and more prominent than those of the fine cut surfaces (Fig. 2(b)). As shown in Fig. 8, these cavities were exposed in the first 5 s of blasting and accentuated in the follow-up blasting at 10 and 15 s. Most cavities disappeared when the process reached steady-state after 30 s blasting, as shown in Fig. 10. The remaining few cavities on the blasted surface Fig. 11. SEM micrographs of surfaces of rough cut EDM WC CO blasted by Abrasive A at MPa air pressure for 30 s (steady-state).

8 J. Qu et al. / Journal of Materials Processing Technology 166 (2005) Fig. 12. Close-up views of surface of the rough cut EDM WC CO blasted by Abrasive A at MPa air pressure for 30 s (steady-state). are possibly due to the porosity in the original WC Co material. This can be validated by examining the close-up views in Fig. 11. The surface texture, disregarding the cavities and porosity, is similar to that in Fig. 7(b). The boundaries between WC grains are not easily identified, and there is no distinct cleavage facet to indicate the individual brittle fracture of WC. In general, all rough cut EDM surfaces exhibited similar pattern of cavities, as seen in Figs. 8 and 9. All fine cut EDM surface showed a ridge pattern similar to that in Figs. 6 and 7. These observations manifest the formation of ridges, bubbles, and cavities in EDM recast layers. Further investigations are necessary to establish the details of the formation mechanisms Erosion wear rate The erosion wear rate experiments were conducted by blasting a 6 mm 6 mm rough cut EDM WC Co surface using Abrasive A for 10 s time increments and then measuring the weight loss. A blasting time increment of 10 s was used in order to measure the steady-state erosion wear rate. Due to the limitation of the EDM workpiece sizes, this test setup was different from the standard erosion tests specified in ASTM G76-95 [28], in which a larger surface area should be used. As shown in Table 2, six erosion wear rate tests were carried out at and MPa air pressure with 10, 20, and 30 s blasting time. Two parts were blasted and the weight loss was measured in each test. The average weight loss was divided by the weight of SiC abrasive blasted to calculate the erosion wear rate. On average, 31 and 46 mg/s of Abrasive A were blasted at and MPa air pressure, respectively, during a 10 s interval. Fig. 12 shows the experimental results of the cumulative weight reduction and erosion wear rate for Abrasive A. The higher air pressure (0.276 MPa) generated higher erosion rates than the lower air pressure (0.138 MPa). A four-fold increase in erosion rate was obtained by doubling the air pressure. A higher erosion wear rate was observed in the first 10 s than the following two blasting durations. This transient can be attributed to the materials removal associated with thermal cracks and bubbles. These defects result in lower erosion wear resistance of the recast layer. At 30 s, when the erosion rates approach steady-state, the and g/g erosion wear rates are higher but comparable to that reported by Beste et al. [12] and Anand and Conrad [13]. 4. Concluding remarks The abrasive micro-blasting process was used to improve the surface integrity of WC Co samples machined by wire EDM. Two sizes of SiC abrasive, two levels of air pressures, and three time durations were applied to blast EDM rough cut and fine cut WC Co surfaces to explore effects of blasting parameters on the surface integrity. The recast layer and heataffected zone can be removed by SiC micro-blasting in very short times using easily achieved erosion conditions. Larger size abrasives and higher air pressure produce higher erosion wear rate and reduce the EDM surface roughness more effectively. The unique surface texture, related to the crater formation of the recast layer, has been revealed on SEM micrographs and studied to help understand the influence of the micro-blasting process on the surface integrity of the wire EDMed WC Co parts. Acknowledgements The authors gratefully acknowledge the support by National Science Foundation Grant # (Dr. K.P. Rajurkar, Program Director) and the assistant by Z. Yu. Portion of this research was sponsored by the User program in the High Temperature Material Lab, Oak Ridge National Lab and the Heavy Vehicle Propulsion Systems Materials Program, Office of Transportation Technologies, US Department of Energy.

9 448 J. Qu et al. / Journal of Materials Processing Technology 166 (2005) References [1] J. Qu, A.J. Shih, R. Scattergood, Development of the cylindrical wire electrical discharge machining process, Part I: concept, design, and material removal rate, J. Manuf. Sci. Eng. 124 (3) (2002) [2] J. Qu, A.J. Shih, R. Scattergood, Development of the cylindrical wire electrical discharge machining process, Part II: surface integrity and roundness, J. Manuf. Sci. Eng. 124 (3) (2002) [3] L. Llanes, E. Idanez, E. Martinez, B. Casas, J. Esteve, Influence of electrical discharge machining on the sliding contact response of cemented carbides, Int. J. Refract. Met. Hard Mater. 19 (1) (2001) [4] M. Ramulu, M.G. Jenkins, J.A. Daigneanult, Spark-erosion process effects on the properties and performance of a TIB 2 particulatereinforced/sic matrix ceramic composite, Ceram. Eng. Sci. Proc. 18 (3) (1997) [5] R.E. Williams, K.P. Rajurkar, Study of wire electrical discharge machined surface characteristics, J. Mech. Process. Technol. 28 (1991) [6] J. Qu, L. Riester, A.J. Shih, R. Scattergood, E. Lara-Curzio, T. Watkins, Nanoindentation characterization of surface layers of electrical discharge machined WC Co, Mater. Sci. Eng. A 334 (2003) [7] U. Beste, L. Hammerstrom, H. Engqvist, S. Rimlinger, S. Jacobson, Particle erosion of cemented carbides with low Co content, Wear (2001) [8] V.A. Pugsley, C. Allen, Microstructure/property relationships in the cavitation erosion of tungsten carbide cobalt, Wear (1999) [9] K. Anand, C. Morrison, R.O. Scattergood, H. Conrad, J.L. Routbort, R. Warren, Erosion of multiphase materials, science of hard materials, science of hard materials, in: Proceedings of the Second International Conference on the Science of Hard Materials, Inst. Phys. Conf. Ser. No. 75, Rhodes, Greece, Adam Hilger Ltd., 1986, pp [10] A. Ball, A comparative study of the wear resistance of hard materials, science of hard materials, in: Proceedings of the Second International Conference on the Science of Hard Materials, Inst. Phys. Conf. Ser. No. 75, Rhodes, Greece, Adam Hilger Ltd., 1986, pp [11] N.P. Hung, L.H. Teong, C.M. Yoong, Review of unconventional machining of metal matrix composites, in: Proceedings of the 1996 Third Biennial Joint Conference on Engineering Systems Design and Analysis, vol. 75, no. 3, ASME PD, New York, NY, 1996, pp [12] U. Beste, L. Hammerstrom, H. Engqvist, S. Rimlinger, S. Jacobson, Particle erosion of cemented carbides with low Co content, Wear 250 (2001) [13] K. Anand, H. Conrad, Local impact damage and erosion mechanisms in WC-6 wt% Co alloys, Mater. Sci. Eng. A (1988) [14] S.K. Hovis, J.E. Talia, R.O. Scattergood, Erosion in multiphase materials, Wear 108 (1986) [15] H. Engqvist, G.A. Botton, N. Axen, S. Hogmark, Microstructure and abrasive wear of binderless carbides, J. Am. Ceram. Soc. 83 (10) (2000) [16] K.D. Bouzakis, N. Michailidis, S. Hadjiyiannis, K. Efstathiou, E. Pavlidou, G. Erkens, S. Rambadt, I. Wirth, Improvement of PVD coated inserts cutting performance through appropriate mechanical treatments of substrate and coating surface, Surf. Coat. Technol (2001) [17] J.M. Guilemany, I.N. Llorca, P.J. Szabo, Residual stress characterization of grit blasted steel surfaces, Surf. Eng. 12 (1) (1996) [18] A. Ben Rhouma, C. Braham, M.E. Fitzpatrick, J. Ledion, H. Sidhom, Effects of surface preparation on pitting resistance, residual stress, and stress corrosion cracking in austenitic stainless steels, J. Mater. Eng. Perform. 10 (5) (2001) [19] M.O. Lai, Y.H. Siew, Fatigue properties of cold worked holes, J. Mater. Process. Technol. 48 (1995) [20] T.C. Lee, J.H. Zhang, Surface treatment of wire electro-discharge machined engineering ceramics by abrasive blasting, J. Adhes. Sci. Technol. 12 (6) (1998) [21] J. Deng, T.C. Lee, Techniques for improved surface integrity of electrodischarge machined ceramic composites, Surf. Eng. 16 (5) (2000) [22] T. Lee, J. Deng, Mechanical surface treatments of electro-discharge machined (EDMed) ceramic composite for improved strength and reliability, J. Eur. Ceram. Soc. 22 (2002) [23] S. Schlautmann, H. Wensink, R. Schasfoort, M. Elwenspoek, A. Berg, Powder-blasting technology as an alternative tool for microfabrication of capillary electrophoresis chips with integrated conductivity sensors, J. Micromech. Microeng. 11 (2001) [24] M. Wakuda, Y. Yamauchi, S. Kanzaki, Effect of workpiece properties on machinability in abrasive jet machining of ceramic materials, Precision Eng. 26 (2002) [25] T. Masuzawa, H.K. Tonshoff, Three-dimensional micromachining by machine tools, Annal. CIRP 46 (1997) [26] K.P. Rajurkar, Z.Y. Yu, 3D Micro-EDM using CAD/CAM, Annal. CIRP 49 (2000) [27] A.W. Ruff, L.K. Ives, Measurement of solid particle velocity in erosive wear, Wear 35 (1975) [28] ASTM Standard G76-95, Standard Test Method for Conducting Erosion Tests by Solid Particle Impingement Using Gas Jets, American Society for Testing and Materials, 2000.