edm machinability and dry sliding friction of wc-co cemented … · 2016-01-13 · edm...

Int. J. Manufacturing Research, Vol. 4, No. 4, 2009 375

Copyright © 2009 Inderscience Enterprises Ltd.

EDM machinability and dry sliding friction of WC-Co cemented carbides

K. Bonny* and P. De Baets Department of Mechanical Construction and Production, Ghent University, IR04, Sint-Pietersnieuwstraat 41, B-9000 Gent, Belgium E-mail: [email protected] E-mail: [email protected] *Corresponding author

J. Vleugels and O. Van der Biest Department of Metallurgy and Materials Engineering, Catholic University Leuven, MTM, Kasteelpark Arenberg 44 (bus 2450), B-3001 Leuven, Belgium

B. Lauwers and W. Liu Department of Mechanical Engineering, Catholic University Leuven, PMA, Celestijnenlaan 300 B (bus 2420), B-3001 Leuven, Belgium

Abstract: A number of WC-Co cemented carbides with 6–12 wt. % Co were machined and surface finished by grinding or Electrical Discharge Machining (EDM). EDM was executed in deionised water through several consecutive gradually finer steps. Correlations between Material Removal Rate (MRR), surface finish and EDM parameters were derived. The quality and integrity of surface finishes were analysed by Scanning Electron Microscopy (SEM). Dry reciprocating sliding experiments on EDM’ed and ground samples against WC-Co pins using a Plint TE77 tribometer revealed a considerable impact of the EDM process as well as distinctive EDM finishing steps on the friction coefficient.

[Received 4 September 2008; Revised 5 January 2009; Accepted 3 April 2009]

Keywords: WC-Co cemented carbide; wire EDM; surface integrity; dry friction; reciprocating sliding.

Reference to this paper should be made as follows: Bonny, K., De Baets, P., Vleugels, J., Van der Biest, O., Lauwers, B. and Liu, W. (2009) ‘EDM machinability and dry sliding friction of WC-Co cemented carbides’, Int. J. Manufacturing Research, Vol. 4, No. 4, pp.375–394.

Biographical notes: Koenraad Bonny graduated in 1999 in Materials Engineering at Ghent University, Belgium. From 2000 to 2002 he worked as Scientific Researcher at the Solid State Sciences Department of the same

376 K. Bonny et al.

university, investigating the efficiency improvement possibility of halogen bulbs by deposition of low-emissivity coatings. In 2002 he joined the Mechanical Construction and Production Department at Ghent University, where he started his PhD research in the framework of an IWT project on electrically conductive wear resistant composite materials. A number of research results have been presented on international conferences and have been published by refereed journals.

Patrick De Baets graduated in 1989 in Mechanical Engineering at Ghent University, Belgium. That year he joined the same university, working as Assistant for Professor C. Deconinck at the Department of Mechanical Construction and Production. In 1995, he received a PhD Degree in Mechanical Engineering from Ghent University. Today he is appointed as full Professor, teaching courses in the domain of Mechanical Engineering and Materials Science at Ghent University. He has more than 18 years of research and teaching experience. He has published approximately 100 papers on these topics in refereed journals and conferences.

Jozef Vleugels graduated in 1988 in Chemistry and Agriculture Engineering at Catholic University of Leuven (K.U. Leuven, Belgium). In 1989, he has joint the Metallurgy and Materials Engineering Department (MTM) at K.U. Leuven. In 1995, he received a PhD in Bio-Engineering Sciences from the K.U. Leuven. In 2001, he was appointed as part-time Professor for the course ‘Ceramic Processing’. His research topics comprise powder metallurgy, electrophoretic deposition, functionally graded materials, coatings, ceramic matrix composites, hardmetals and cermets, electrically conductive ceramic composites, cutting tools and wear parts. He has published more than 150 papers in refereed journals and conferences.

Omer Van der Biest graduated in 1970 as Master in Metallurgical Engineering at Catholic University of Leuven (K.U. Leuven, Belgium). From 1970 to 1975, he worked as scientific researcher at the Lawrence Berkeley Laboratory. In 1974, he received a PhD in Materials Science and Engineering from the University of California. Today he is full Professor at the Metallurgy and Materials Engineering Department (MTM) at K.U. Leuven. Keywords describing his research interests include: high temperature materials, powder metallurgical processes, composites, mechanical properties, surface analysis and analytical electron microscopy. He is author on more than 150 publications in refereed journals.

Bert Lauwers graduated in 1987 as Master in Mechanical Engineering at Catholic University of Leuven (K.U. Leuven, Belgium). In 1993, he obtained his PhD at K.U. Leuven with a dissertation on computer-aided process planning and manufacturing for electrical discharge machining. He is active expert of the ISO technical committee for standardisation of programming languages for numerically controlled equipment. Today he works as Professor at the Mechanical Engineering Department at K.U. Leuven. His research focuses on Computer Aided Design (CAD), Computer Aided Manufacturing (CAM), Computer Integrated Manufacturing (CIM), Computer Aided Process Planning (CAPP), advanced NC-programming and modelling and manufacturing of complex shape products.

Weidong Liu works as scientific researcher at the Mechanical Engineering Department of the Catholic University of Leuven, Belgium, more specifically at the divisions Production techniques, Engineering and Automation (PMA) and Production Processes (PP). His research efforts focus on non-conventional manufacturing processes, including laser cutting, rapid prototyping, electrical

EDM machinability and dry sliding friction of WC-Co cemented carbides 377

discharge machining (EDM), five axis milling and high speed milling. A number of his research results have been presented on international conferences and have been published by refereed journals.

1 Introduction

Based on economic reasons as well as on ecological considerations such as the conservation of material and energy resources and waste reduction, there is a rising need for adequate limitation of wear and corrosion damage of machines and construction tools. In this way there is an obvious industrial demand for wear resistant materials to be applied under heavy tribological circumstances and preferably without lubrication as for instance for tools (chisels, cutting tools, metal forming dies, punches, etc.) and various machine accessories. Furthermore, in the fields of aerospace and automobile, advanced materials, such as engineering ceramics, with the properties of ultra-hard, erosion/friction-resistant and high-temperature-resistance are more and more claimed and applied. However, a significant disadvantage of these materials is their relatively high coefficient of friction in dry contact conditions, involving heat development and energy loss. Their high hardness renders them intrinsically also difficult to shape and finish by conventional methods. Profile grinding with super hard grinding grains (CBN, diamond) is nearly the only possibility to shape those hard materials, but the shapes that can be generated by grinding cannot be intricate. Electrical Discharge Machining (EDM) is one of the non-conventional manufacturing processes that allow intricate shapes in materials to be produced irrespective of their strength or hardness, i.e., the strength and hardness is no limitation to the machinability, provided the material is electrically conductive, which is generally not the case for ceramic materials. Electro-erosion has successfully proven to be feasible for low electrical conductive materials (Kim and Kruth, 2001; Kozak et al., 2004) and for cemented carbides (Pandit and Rajurkar, 1981; Pandey and Jilani, 1987; Yu et al., 2004), and more specifically, for WC-Co based cemented carbides (Gadalla and Tsai, 1989; Lauwers et al., 2004; Mahdavinejad and Mahdavinejad, 2005). However, difficulties also arise with respect to the surface finish conditions (Puertas et al., 2004; Khan et al., 2006), the corrosion of these materials during machining (Obara et al., 2004), and the influence the machining parameters may have on final properties such as strength and hardness (Tsutsui and Tamura, 1987; Qu et al., 2003; Juhr et al., 2004; Llanes et al., 2004; Jiang et al., 2005; Casas et al., 2006; Lin et al., 2008), or friction and wear characteristics (Casas et al., 2001; Llanes et al., 2001; Bonny et al., 2007, 2009). EDM surface damage is given by residual stresses and cracks produced within the thermally affected zone (recast layer and adjacent regions) beneath the shaped surface (Lenz et al., 1975; Yakou and Hasegawa, 1995; Stössel-Sittig et al., 2004; Jiang et al., 2005).

Flat specimens of a number of WC-Co cemented carbide grades were machined and surface finished under distinctive EDM finish conditions. The MRRs were evaluated under rough cutting EDM. EDM surfaces were examined by SEM, energy dispersive X-ray analysis (EDX) and surface topography scanning in order to identify the occurring material removal mechanisms and to correlate surface integrity and surface roughness with the corresponding EDM parameters under rough cutting and four EDM finishing-steps. Dry reciprocating sliding experiments were executed

378 K. Bonny et al.

on wire-EDM’ed and ground WC-Co flats against WC-Co pins using a pin on plate testing rig, with the goal to elucidate the influence of the EDM process as well as different surface finishing steps within the EDM process on the friction characteristics.

2 Experiment

2.1 WC-Co grades

The chemical, physical, mechanical and microstructural properties of the distinctive cemented carbide grades envisaged for this research are listed in Table 1. The WC10Co, WC12Co(V), WC12Co(Cr), WC10Co(Cr/V) and WC6Co(Cr/V) cemented carbide grades are CERATIZIT grades GC32, GC20, GC20CR, MG12 and MG18, respectively.

Table 1 Chemical, physical, mechanical and microstructural properties of WC-Co cemented carbides

Hardmetal grade WC10Co WC12Co (V) WC12Co (Cr) WC10Co (Cr/V) WC6Co (Cr/V)

Co binder content (wt%)

10 12 12 10 6

WC grain growth inhibitor

None VC Cr3C2 Cr3C2/VC Cr3C2/VC

Density (g/cm³) 14.33 14.08 14.01 14.23 14.62

Thermal conductivity (W.m–1.K–1)

105 95 95 85 90

Vickers hardness HV10(kg/mm2)

1149 10 1286 8 1306 5 1685 38 1913 13

Fracture toughness KIC(30 kg) (MPa m)

15.5 15.4 0.5 15.5 0.6 9.7 0.2 8.8 0.2

E-modulus (Gpa) 578 6 563 2 546 2 541 4 609 4

Mean grain size, dav ( m)

2.2 0.9 0.9 0.3 0.6

WC grain size, d50 ( m)

1.8 0.7 0.8 0.3 0.5


4.2 1.5 1.7 0.6 1.0


6.0 1.8 2.1 0.7 1.2

The HV10 Vickers hardness was measured with an indentation load of 10 kg (Model FV-700, Future-Tech Corp., Tokyo, Japan). The fracture toughness KIC(30 kg) was obtained by the Vickers indentation technique, based on crack length measurements of the radial crack pattern produced by Vickers HV30 indentations. The KIC values were calculated according to the Shetty formula, which is given by equation (1):


IC 0.08894VPK Hl

(1)

with HV, the Vickers hardness, P, the indentation load (N) and l, the total crack length (m), which is defined as the radial crack length (c) minus half the indentation diagonal length (a), equation (2):

.2al c (2)

The Young’s modulus E was measured by the resonance frequency method. The resonance frequency was obtained on a Grindo-Sonic (J.W. Lemmens, Elektronika N.V. Leuven, Belgium), by means of the impulse excitation method (ASTM E 1876-99). The grain size distribution of the cemented carbide grades was acquired using computer image analysis software according to the linear intercept method. At least 1000 grains were measured for each grade. The WC10Co grade exhibits the coarsest WC grain structure, with 50% of the grains being smaller than 1.8 m and 95% being smaller than 6.0 m. The WC10Co(Cr/V) grade has the finest microstructure, with 95% of the grains smaller than 0.7 m.

The above mentioned cemented carbide grades were manufactured and surface finished by wire-EDM (ROBOFIL 2030SI, Charmilles Technologies, Switzerland) or grinding (JF415DS, Jung, Göppingen, Germany) with a diamond grinding wheel (type MD4075B55, Wendt Boart, Brussels, Belgium). Surface conditions of ground and wire-EDM’ed WC-Co grades were analysed by Scanning Electron Microscopy (SEM, XL30-FEG, FEI, Philips, The Netherlands) and Energy Dispersive X-ray (EDX) spectroscopy.

2.2 Friction testing

The frictional behavior of wire-EDM’ed and ground cemented carbides was evaluated using a high frequency tribometer, in which a WC6Co(Cr/V) cemented carbide pin (CERATIZIT grade MG12) was slid against plate specimens (width 38 mm, length 58 mm, thickness 4 mm) of the above mentioned cemented carbide grades, in air-conditioned atmosphere of 23°C and relative humidity of 60%, conform the ASTM G133 linearly reciprocating sliding wear test standard and similar to the test procedure described in previous work (Bonny et al., 2008). The average rounding radius and roughness parameters Ra and Rt of the pins were determined to be 4.08 mm, 0.35 mand 2.68 m, respectively, by means of surface scanning equipment (Somicronic EMS Surfascan 3D, needle type ST305). The imposed contact load was 15 N. The stroke length of the oscillating motion was 15 mm. A sliding velocity of 0.3 m/s was applied. Before each test, the specimens were cleaned ultrasonically with acetone.

3 Results and discussion

3.1 Wire EDM parameters

The wire-EDM process was performed in deionised water (dielectric conductivity 5 S/cm), using a brass (CuZn37) wire electrode with a diameter of 0.25 mm and

380 K. Bonny et al.

a tensile strength of 500 MPa. Initial rough cutting was carried out with high spark thermal energy to get a higher MRR. In order to improve the surface quality, several consecutive finish cuts with globally decreasing energy input and pulse duration were performed on the WC-Co cemented carbides, with following sequential EDM finishing steps: E3, E17, E8, E10, E11, E12, E13, E21, E22 and E23. One rough (E3) and 4 finish (E8, E13, E21, E23) surface variants were selected for this investigation. The corresponding generator settings for these EDM finishing steps are summarised in Table 2.

Table 2 Wire-EDM parameters and surface characteristics for rough EDM step (E3) and 4 EDM finishing steps (E8, E13, E21 and E23)

EDM setting E3 E8 E13 E21 E23 Open voltage (V) 80 80 140 140 140

Pulse duration te ( s) 1.2 1 3 1 1

Pulse interval t0 ( s) 8.3 10 6.6 4 4 Maximum speed (mm/min) 14.5 14.5 6.1 6.1 8 Reference servo voltage Aj (V) 50 13.2 7 6 0 Pulse ignition height IAL (A) 8 16 5 4.5 2.5 Flushing pressure (bar) 6.5 0 0 0 0 Wire tension (N) 11 16 12 10 10 Wire winding speed (m/min) 8 6.8 6.8 6.8 4.8

It is clear that the rough cut combines a high energy input with a high flushing pressure to remove the molten material. After this rough cut, a first finishing cut (E8) with lower energy (shorter pulse duration, longer pulse interval) is made. In this pass, only a limited amount of material is removed, and thus, there is no flushing to avoid wire vibrations, making it possible to attain higher accuracy. In each following pass, the wire EDM surface layer from the former pass is cut off, each time with lower energy than the foregoing pass. The last two finishing cuts do not remove material, but just re-melt the roughness peaks, lowering the surface roughness.

The most important parameter to limit the energy in the finishing cuts is the pulse ignition height, i.e., the current necessary to de-ionise the sparking gap to initiate the sparks. If this current is smaller, sparks do not initiate as easily as with higher ignition currents. With very low pulse ignition current (e.g., finishing-step E23) the magnetic field, which ionises the sparking gap, appears only strong enough on the roughness peaks of the surface to initiate sparking. In this way, only the protruding material will be hit by sparks and will re-melt to lower the surface roughness.

3.2 Material Removal Rate

During the rough manufacturing EDM step, the open voltage was fixed at 80 V, with a pulse interval time of 8.3 s, a reference servo voltage of 50 V and a pulse ignition height of 8 A. The obtained MRR for the distinctive cemented carbide grades are presented in Table 3. Comparing the results for the WC12Co(V) and the WC12Co(Cr) alloy,


both displaying similar grain size distribution (see Table 1) and equal binder content, indicates that the grain growth inhibitor, i.e., V and Cr respectively, has no influence on the MRR.

Table 3 MRR during rough EDM cutting (E3)

Grade MRR (mm2/min)WC10Co 40.1 WC12Co(V) 44.2 WC12Co(Cr) 44.3 WC10Co(Cr/V) 45.3 WC6Co(Cr/V) 46.1

Plotting the MRR, measured during rough wire EDM of the cemented carbides, vs. their d50, d90 and d95 WC grain size, as shown in Figure 1(a), reveals a linear correlation between these parameters. The MRR is found to increase with finer grain size distribution. The R2-values in Figure 1(a) indicate that the trend lines match the experimental results quite well.

The decreasing MRR with increasing grain size is probably linked to the thermal conductivity of the cemented carbides, Figure 1(b). Two grades with equal Co content, i.e., WC10Co and WC10Co(Cr/V), but coarse and fine grain size distribution, respectively, exhibit a considerably different thermal conductivity. The higher thermal conductivity for the WC10Co alloy can be explained by the larger amount of cobalt occurring between the coarse WC grains, compared to the finer grains of the WC10Co(Cr/V) grade.

Figure 1 Relationship between grain size distribution and (a) MRR, (b) thermal conductivity

(a) (b)

Furthermore, MRR is noticed to decrease with rising thermal conductivity, as illustrated in Figure 2(a). This correlation can be explained regarding the material removal mechanism, which has been identified as full melting, as will be demonstrated further in this paper. A lower thermal conductivity therefore reduces the heat

382 K. Bonny et al.

losses due to heat conduction in the work piece, leading to a more focused energy distribution to melt the material and, thus, yielding a higher MRR.

The influence of the Co concentration on the MRR for the distinctive cemented carbides is presented in Figure 2(b). MRR is noticed to decrease with increasing Co content, excluding the WC10Co alloy. This trend is in agreement with (Kim and Kruth, 2001). The deviation from this trend for the WC10Co grade can be attributed to its significantly coarser grain size distribution, compared to the other grades, for which the influence of the Co concentration on the MRR is inferred to be dominant over the grain size effect.

Figure 2 Relationship between MRR and (a) thermal conductivity and (b) Co content

(a) (b)

3.3 Surface integrity

The surface and cross-sectioned views of the rough cut WC10Co cemented carbide (EDM condition E3) are shown in Figure 3. Both atomic number (BSE mode) and secondary electron (SE mode) SEM analysis reveals the presence of many droplets in clusters, voids and microcracks on the wire-EDM’ed surface. These phenomena point out that the cemented carbide is initially molten by the sparking thermal energy during rough cutting. Most molten and oxidised material is flushed away by the dielectric, whereas a small amount of molten material is not expelled but rapidly quenched by the dielectric, and resolidifies on the EDM surface to form clustered droplets. At the same time, the material on the surface shrinks in the process of resolidification after sparking due to the dielectric cooling. The solidified microstructure of the recast layer, exhibiting droplets and voids, as well as microcracks, can be clearly seen in Figure 3(a), (b) and (d).

Due to the larger thermal expansion coefficient of the resolidified layer compared to the cemented carbide substrate, microcracks are formed in the recast layer owing to the thermal impact during wire-EDM. Additionally, tensile residual thermal stresses occur in the surface layer of the cemented carbide (Jiang et al., 2005).

The thickness of the recast layer and the heat-affected zone (HAZ) was derived from cross-sectioned SEM views of wire-EDM’ed cemented carbides, as illustrated in Figure 3(c). The plate samples were cut and subsequently polished along a plane perpendicular to the wire-EDM’ed surface. SEM analysis allowed the structure of the


recast layer and the HAZ to be distinguished from the pristine material structure, and thus, the thickness of the recast layer and the HAZ to be measured. For the rough wire-EDM condition E3 and WC10Co alloy, the thickness of the recast layer was determined to be approximately 15 m, whereas the corresponding HAZ was found to reach depths up to 40 m, Figure 3(c). It should be noted that similar values were found for the other examined grades. The HAZ especially modifies the binder due to its lower melting temperature compared to WC, and may locally alter the performance of the WC-Co alloy. The depth of this zone depends on the amount of current applied during the distinctive EDM conditions.

Figure 3 Cross-sectioned views (a, b, c) and surface views (d) of the wire-EDM rough cut WC10Co cemented carbide and corresponding X-ray diffraction pattern (e)

(a) (b)

(c) (d)

(e)

X-ray diffraction analysis revealed the formation of tungsten subcarbide (WC1–x), eta-phase (W3Co3C and W6Co6C) and copper (Cu) on the rough cut EDM surface, Figure 3(e). These phases are formed in the recast layer during solidification of the

384 K. Bonny et al.

molten material, which has a reduced carbon content. The small amount of copper appertains to a thin, easily removable top layer originating from the brass wire-electrode. Eta-phase was found on all EDM surfaces for all grades and EDM surface finish conditions. From the findings about composition and morphology of the EDM surfaces, as well as the influence of material parameters on the MRR, it can be inferred that the mechanism for material removal of these cemented carbides under EDM occurs through melting of both matrix (Co) and carbide grains.

The rough cut wire-EDM part does not meet the requirements for industrial applications because of the poor surface quality. Therefore, several consecutive finish cuttings with globally decreasing sparking energy are carried out to remove the droplets, craters and microcracks on the surface. This is clearly demonstrated by Figure 4, in which the surface and cross-sectioned views after 4 EDM finishing steps (E8, E13, E21 and E23) are compared for a WC12Co(Cr) cemented carbide. Thermal cracking of the larger carbide grains was frequently observed in the finer EDM finishing-steps, e.g., Figure 4(c).

Comparing the wire EDM parameters listed in Table 2 and the cross-sectioned views shown in Figure 4, reveals that the thickness of both recast layer and HAZ decreases with sequential EDM surface finishing, and thus, decreases with reduced power applied during wire EDM. For instance, the first finishing cut removes approximately 40 m of material, corresponding to the thickness of the HAZ induced during the roughing step, and introduces a new but smaller HAZ (thickness ca. 30 m).

Figure 4 Surface (left) and cross-sectioned (right) SEM views of WC12Co(Cr) cemented carbide for distinctive wire-EDM conditions: (a) E8; (b) E13; (c) E21 and (d) E23

(a)

(b)


Figure 4 Surface (left) and cross-sectioned (right) SEM views of WC12Co(Cr) cemented carbide for distinctive wire-EDM conditions: (a) E8, (b) E13, (c) E21 and (d) E23 (continued)

(c)

(d)

The gradually lower energy input during the consecutive finishing steps, compared to the roughing step, leads to lower surface roughness as well, since the craters that are formed by EDM are smaller. The average values of the surface roughness parameters Ra and Rt under rough and finish cutting steps, are compared in Table 4 for the distinctive cemented carbides. In all cases, Ra and Rt are noticed to exhibit a similar trend. The surface roughness is progressively reduced during the consecutively performed EDM finishing-steps, and thus, also depends on the amount of power applied during the distinctive EDM conditions. For the WC12Co(Cr) samples, for instance, the Rt-value decreased from 16.08 m after rough cutting to 1.01 m after the final finish cut E23. This implies that the large droplets, craters and microcracks left by previous cutting steps have been almost completely removed, although a small amount of recast material remains even after the final finishing cut (see Figure 4(c) and (d)). As for the other cemented carbides, surface integrities are comparable to the WC12Co(Cr) alloy, after both rough and finish cutting EDM.

No large differences are encountered in the obtained Ra- and Rt-values amongst the WC-Co alloys. Under equal set-up and identical series of EDM parameters, the WC10Co specimens obtain the smoothest surfaces, whereas the highest roughness values are encountered with the WC6Co(Cr/V) cemented carbides. As can be seen in Table 4, Ra levels below 0.2 m are attainable for the cemented carbides, except for the WC6Co(Cr/V) grade, which has a higher surface roughness. The relatively small differences in roughness can be attributed to the various WC grain size (hardness), cobalt binder concentration and thermal conductivity of the cemented carbide grades (see Table 1), as illustrated in Figure 5.

386 K. Bonny et al.

Table 4 Ra and Rt roughness ( m) for the rough (E3) and smoother EDM conditions (E8–E23) and ground (measured along and perpendicular direction) surface variants of the cemented carbides

Grade WC10Co WC12Co (V) WC12Co (Cr) WC10Co (Cr/V) WC6Co (Cr/V)

Surface finish

Ra Rt Ra Rt Ra Rt Ra Rt Ra Rt

E3 2.08 15.18 2.31 15.84 2.37 16.08 2.34 17.36 2.08 17.02 E8 1.07 6.62 1.3 7.16 1.29 8.67 1.22 6.81 1.00 5.95 E21 0.24 2.16 0.26 2.38 0.24 2.99 0.24 2.73 0.37 3.05 E23 0.15 1.02 0.18 1.04 0.16 1.02 0.17 1.08 0.24 1.29

Grinding 0.25 2.04 0.27 2.08 0.19 1.74 0.26 2.25 0.22 1.83

Grinding 0.08 0.56 0.09 0.68 0.09 0.67 0.22 1.83 0.09 1.95

Figure 5 Correlation between Ra and Rt surface roughness for EDM condition E23 and (a) hardness; (b) grain size; (c) thermal conductivity and (d) binder content

(a) (b)

(c) (d)

Surface roughness is found to increase with increasing hardness, Figure 5(a). An inverse trend is observed when surface roughness is plotted against grain size, Figure 5(b), or thermal conductivity, Figure 5(c). The large variation in surface roughness between the two grades with the lowest grain size, Figure 5(b), and/or thermal conductivity, Figure 5(c), can be explained by the large difference in hardness and/or binder content between these two grades.


One of the major factors affecting the final surface roughness is the Co binder concentration, Figure 5(d). Indeed, a reduced Co content requires a higher amount of energy in order to obtain the same quantity of eutectic liquid, since that is determined by the Co content. Since the final EDM finishing cuts just re-melt the surface layer to decrease the surface roughness, it is obvious that lower binder content involves higher roughness. This trend also confirms that the EDM material removal occurs by melting of the cemented carbide material. Furthermore, the surface roughness does not increase with increasing grain size, which would have been the case if the material removal mechanism had been full grain fall out. Indeed, in that case, the largest grain size would yield the highest surface roughness.

The residual thermal stress occurring in the surface layer of the wire-EDM’ed cemented carbides is noticed to decrease during the consecutively performed finish cuttings, as demonstrated in Figure 6, in which the surface residual stresses in the WC phase of WC12Co(Cr) and WC10Co(Cr/V) alloys after grinding and distinctive EDM conditions are compared. The measurements were executed by X-ray diffraction on a Siemens D500 XRD, using the d-sin2 method. The (300) WC peak, corresponding to a diffraction angle 2 = 133.31° was applied in order to acquire the residual stress. The sin2 range was varied from 0 to 0.6 in steps of 0.1, and the angle 2 was varied between 130° and 136° at 0.02°/steps of 5 s.

Figure 6 Surface residual stress in the WC phase of 2 cemented carbides after grinding and different EDM (E13–E23) conditions: a+ value indicates a tensile stress, whereas a – value represents a compressive stress

The stress state of the bulk material, in which the WC phase is considered to be in compression whereas the Co phase is in tension due to the higher thermal expansion coefficient of the binder phase compared to WC after cooling down from the sintering temperature, is found to be affected considerably by the mechanical impact during grinding, i.e., the compressive residual surface stress is significantly increased. Although it is not possible to measure the residual stresses in the Co-binder phase accurately, due to the uncertainty on the W and C content, it is most likely that the nature of the stress in the binder phase after grinding is also compressive, resulting in a surface layer that is entirely in compression. The presence of high compressive stresses on the surface of ground samples results in better final properties such as strength and wear resistance.

EDM finishing cuts, however, are noticed to induce tensile residual stresses. As a result, relatively small surface cracks arise, as observed on the EDM cross-sections, Figure 4. Beside small thermal cracks in the WC grains, Co depletion is another strength

388 K. Bonny et al.

limiting factor of EDM samples, Figure 4(b) and (c). This phenomenon is evidenced by Table 5, in which the results of flexural strength measurements at room temperature in a three-point bending test are compared for ground and EDM finished WC-Co cemented carbides. Indeed, due to the high intrinsic strength of the WC-Co alloys, the actually measured strength on bending samples is actually controlled by the surface finish, or more particularly, by the surface damage generated during machining. The strength data indicate that EDM under the applied parameters strongly reduces the flexural strength of all grades. It is also clear that the detrimental effect of EDM on the bending strength is more pronounced for the hardest grades with a modest toughness and smallest WC grain size, i.e., WC10Co(Cr/V) and WC6Co(Cr/V), than for the coarser grades. The strength reduction is clearly less pronounced compared to the increase of the fracture toughness of the studied WC-Co grades (see Table 1).

Table 5 Flexural strength after distinctive surface finishing operations on WC-Co cemented carbides

Flexural strength (MPa) WC10Co WC12Co (V) WC12Co (Cr) WC10Co (Cr/V) WC6Co (Cr/V)

After grinding 3064 91 4279 61 2919 101 3509 168 3078 295 After EDM (finishing-step E23)

2200 53 2151 22 2117 168 1300 15 1298 71

% reduction 28 50 27 63 58

3.4 Coefficient of friction

For most engineering materials, friction usually does not depend on apparent contact area, due to plastic deformation in the micro-scale of the asperities. Friction is therefore usually not affected by surface finishes. However, for very hard materials, plastic deformation is very limited and friction coefficient has indeed been observed to increase with increase in surface roughness (Hayward, 1992; Bhushan et al., 1993). On the other hand, wear of materials is strongly affected by stresses at and beneath the area of real contact. Numerical analysis on the contact problem between a smooth surface and a rough surface showed that the maximum contact pressures are significantly higher between the rougher surfaces than those between ideal smooth surfaces (Bailey and Sayles, 1991).

In this research, the applied normal force (FN) and the concomitant tangential friction force (FT) of pin-on-flat sliding pairs were recorded continuously using respectively a load-cell and a piezoelectric transducer. The FT/FN forces ratio is defined as the coefficient of friction ( ), which can be differentiated in a static ( stat) and a dynamic component ( dyn). For each sliding wear experiment, a new WC6Co(Cr/V) pin was used in order to ensure identical initial surface conditions.

Friction coefficient curves for both wire-EDM’ed (conditions E3, E8, E21 and E23) and ground cemented carbide flat/WC6Co(Cr/V) pin combinations as function of the oscillating sliding distance are shown in Figure 7. Each curve is an average of at least two wear experiments performed under identical conditions, with a deviation of less than 10% between different samples of the same material. The error bars indicating the extent of the variations are excluded to make the figure better readable.


Figure 7 Static and dynamic friction coefficient curves for wire-EDM’ed and ground cemented carbide grades sliding against WC6Co(Cr/V) pin at 0.3 m/s, with a 15 N contact load

(a) (b)

(c) (d)

(e)

The friction coefficient of the investigated tribopairs was measured to be in the range of 0.61–0.76 and 0.39–0.54 for the static and dynamic component, respectively. The dynamic and static component of friction is found to vary similarly as function of the sliding distance, however at a different level. In all cases, the coefficient of friction is noticed to increase abruptly during the first metres of sliding. After a running-in stage, the variations in the friction force curve become marginal. The fluctuations in the

390 K. Bonny et al.

friction curves, both in the initial phase and the equilibrium stage, are due to a continuous breaking and regeneration of micro junctions and indicate a more pronounced adhesion of both contact surfaces. Some numerical data are provided in Table 6, which displays coefficients of friction at various sliding distance for the investigated tribopairs.

Table 6 Static and dynamic friction coefficient as function of surface finish and sliding distance (s) for the cemented carbide flat/WC6Co(Cr/V) pin sliding combinations

stat dyn

GradeSurface finish

s (km) 0.01 1 4 10 0.01 1 4 10 E3 0.75 0.76 0.76 0.75 0.53 0.54 0.53 0.53 E8 0.74 0.75 0.74 0.74 0.52 0.53 0.52 0.52 E21 0.74 0.73 0.73 0.73 0.52 0.51 0.51 0.51 E23 0.72 0.72 0.72 0.72 0.51 0.50 0.50 0.50

WC10Co

Grinding 0.7 0.69 0.69 0.7 0.49 0.48 0.48 0.48 E3 0.74 0.75 0.75 0.75 0.52 0.52 0.52 0.51 E8 0.74 0.74 0.73 0.73 0.51 0.51 0.51 0.50 E21 0.72 0.71 0.70 0.70 0.50 0.50 0.50 0.49 E23 0.72 0.70 0.68 0.68 0.50 0.49 0.48 0.48

WC12Co (V)

Grinding 0.7 0.67 0.67 0.67 0.47 0.46 0.46 0.46 E3 0.74 0.76 0.76 0.74 0.52 0.54 0.54 0.53 E8 0.73 0.74 0.74 0.73 0.50 0.52 0.52 0.52 E21 0.69 0.72 0.71 0.71 0.49 0.50 0.50 0.51 E23 0.67 0.69 0.69 0.70 0.46 0.48 0.48 0.49

WC12Co (Cr)

Grinding 0.68 0.67 0.68 0.68 0.47 0.46 0.47 0.48 E3 0.70 0.72 0.70 0.69 0.46 0.48 0.47 0.46 E8 0.68 0.69 0.68 0.67 0.45 0.46 0.45 0.45 E21 0.66 0.68 0.67 0.66 0.44 0.45 0.44 0.44 E23 0.63 0.64 0.65 0.64 0.42 0.43 0.43 0.42

WC10Co (Cr/V)

Grinding 0.62 0.61 0.62 0.62 0.4 0.39 0.4 0.4 E3 0.76 0.76 0.76 0.75 0.53 0.52 0.53 0.52 E8 0.74 0.75 0.74 0.73 0.52 0.51 0.52 0.51 E21 0.72 0.73 0.72 0.71 0.50 0.51 0.50 0.50 E23 0.70 0.70 0.70 0.69 0.49 0.49 0.49 0.48

WC6Co (Cr/V)

Grinding 0.69 0.67 0.68 0.68 0.47 0.46 0.47 0.47

Generally, the origin of friction is referred to one adhesive and one abrasive (ploughing) component. In these pin-on-flat tests, the ploughing component can be divided into one macro component due to the plastic penetration of the pin into the counter surface, and one micro component due to asperity interaction (Evans and Marshall, 1980). The macro part has its significance during the first revolutions until an equilibrium track is formed in the flat surface. The micro ploughing is strongly related to the surface roughness. Indeed, when one of two rough surfaces in a sliding contact becomes smoother there will be fewer obstacles that have to be overcome. This decreases the tangential force required to


maintain the sliding and therefore the friction coefficient decreases. The lowest friction levels are recorded for the WC10Co(Cr/V)-WC6Co(Cr/V) tribopairs, whereas the WC10Co-WC6Co(Cr/V) combinations display the highest coefficient of friction.

The effect of the wire-EDM process and the distinctive EDM conditions on the friction coefficient is quite pronounced. The lowest friction level is encountered with the ground cemented carbide samples, whereas the rough EDM cut specimens exhibit the highest coefficient of friction. Moreover, the friction coefficient is noticed to decrease with finer-executed EDM, up to values nearby those recorded for the equivalent ground cemented carbide grades. These trends are in agreement with (Llanes et al., 2001). The differences in friction level between wire-EDM’ed and ground samples appear to diminish with increasing sliding distance.

The dependency of the coefficient of friction on surface roughness for various sliding wear paths is compared in Figure 8(a) and (b) for ground and wire-EDM’ed WC-Co alloys during reciprocative sliding wear experiments using a 15 N contact load. In the range of measured Ra and Rt values, both the static and dynamic coefficient of friction are noticed to decrease with enhanced surface finish refinement. Comparing the coefficient of friction for the finest EDM-related surface finish variant with ground surfaces, both displaying similar roughness levels, reveals that the ground surfaces yield lower friction coefficients. However, the relative scatter in friction as function of roughness is noticed to reduce slightly with increasing wear distance, in full agreement with the results presented in Figure 7.

Figure 8 Influence of surface finishing roughness Ra (a) and Rt (b) on friction coefficient for cemented carbide flat/ WC6Co(Cr/V) pin sliding pairs and various wear paths (v = 0.3 m/s, FN = 15 N)

(a) (b)

The influence of EDM is even more obvious when static and dynamic friction coefficients are plotted against the depth of the wire-EDM induced Heat Affected Zone (HAZ), Figure 9. It is obvious that no recast layer appears with the ground specimens. The higher friction level for wire-EDM’ed cemented carbides compared to their ground equivalents can be explained in terms of different sliding contact conditions. Indeed, the sliding contact on wire-EDM’ed samples is characterised by the occurrence of a thermally induced recast surface layer and a subsurface HAZ, exhibiting both modified tribological compatibilities with the WC6Co(Cr/V) pin, compared to the ground

392 K. Bonny et al.

cemented carbides, where the pin makes direct contact with the original base material. Furthermore, the wire-EDM process was evidenced to cause thermal cracks and residual surface stresses which are tensile in nature, in contrast with the compressive stress state occurring in the bulk material and for ground surfaces.

Figure 9 Influence of the Heat Affected Zone on friction coefficient for cemented carbide flat/WC6Co(Cr/V) pin sliding pairs and various wear paths (v = 0.3 m/s, FN = 15 N)

4 Conclusions

A number of WC-Co based cemented carbides were proven to be feasible for wire-EDM in deionised water. The material removal mechanism was full melting and evaporation. After rough cutting, the heat-affected zone reached depths up to 40 m. EDM was evidenced to initiate a thermally induced recast layer containing tungsten subcarbide, eta-phase, resolidified droplets and voids, involving residual tensile surface stresses and microcracks as well as a considerable reduction of flexural strength. Grain size distribution and cobalt concentration were found to be major factors affecting the MRR and the attainable surface smoothness. The execution of several consecutive EDM surface finishing steps allowed the average surface roughness to be reduced below 1.3 m Rt or 0.3 m Ra. Comparative dry sliding experiments on WC-Co/WC-Co cemented carbide tribo-pairs revealed that wire-EDM increases the friction coefficient, compared to grinding, which should be related to differences in surface roughness, tribological compatibility and surface stress state. The gradually finer execution of EDM surface finishing reduced the friction level.

Acknowledgements

This work was co-financed with a research fellowship of the Flemish Institute for the promotion of Innovation by Science and Technology in industry (IWT) under project contract number GBOU-IWT-010071-SPARK. The authors gratefully recognise all the support, scientific contributions and stimulating collaboration from the partners from the University of Ghent (UGent) and the Catholic University of Leuven (K.U. Leuven). Special acknowledgement goes to CERATIZIT for supplying the hardmetal grades.


References Bailey, D.M. and Sayles, R.S. (1991) ‘Effect of roughness and sliding friction on contact stress’,

ASME Journal of Tribology, Vol. 113, pp.729–738. Bhushan, B., Subramaniam, V.V., Malshe, A., Gupta, B.K. and Ruan, J. (1993) ‘Tribological

properties of polished diamond films’, J. Appl. Phys., Vol. 74, No. 6, pp.4174–4180. Bonny, K., De Baets, P., Ost, W., Vleugels, J., Huang, S., Lauwers, B. and Liu, W. (2009)

‘Influence of electrical discharge machining on the reciprocating sliding wear response of WC-Co cemented carbides’, Wear, Vol. 266, Nos. 1–2, pp.84–95.

Bonny, K., De Baets, P., Van Der Biest, O., Vleugels, J. and Lauwers, B. (2007) ‘Influence of surface finishing and binder phase on friction and wear of WC based hardmetals’, Materials Science Forum, Vols. 561–565, Part 3, pp.2403–2406.

Bonny, K., De Baets, P., Vleugels, J., Salehi, A., Van der Biest, O., Lauwers, B. and Liu, W. (2008) ‘Influence of secondary electro-conductive phases on the electrical discharge machinability and frictional behavior of ZrO2-based ceramic composites’, J. Mater. Process. Tech., Vol. 208, Nos. 1–3, pp.423–430.

Casas, B., Martinez, E., Esteve, J., Anglada, M. and Llanes, L. (2001) ‘Influence of electrical discharge machining on the tribological characteristics of WC-Co alloys’, Revista de Metalurgia (Madrid), Vol. 37, No. 2, pp.140–144.

Casas, B., Torres, Y. and Llanes, L. (2006) ‘Fracture and fatigue behavior of electrical-discharge machined cemented carbides’, International Journal of Refractory Metals and Hard Materials, Vol. 24, Nos. 1–2, pp.162–167.

Evans, A.G. and Marshall, D.B. (1980) ‘Wear mechanisms in ceramics’, in Rigney, D.A. (Ed.): Fundamentals of Friction and Wear of Materials, American Society for Materials, pp.439–452.

Gadalla, A.M. and Tsai, W. (1989) ‘Electrical discharge machining of tungsten carbide-cobalt composites’, J. Am. Ceram. Soc., Vol. 72, pp.1396–1401.

Hayward, I.P. (1992) ‘Effect of roughness on the friction of diamond on CVD diamond coatings’, Wear, Vol. 157, No. 2, pp.215–227.

Jiang, D., Anné, G., Vleugels, J., Vanmeensel, K., Eeraerts, W., Liu, W., Lauwers, B. and Van der Biest, O. (2005) ‘Residual stress in hardmetals caused by grinding and EDM machining and its influence on flexural strength’, Proc. 16th Int. Plansee Seminar – Powder Metallurgical High Performance Materials, Reutte (Austria), Vol. 2, pp.1075–1085.

Juhr, H., Schulze, H.P., Wollenberg, G. and Künanz, K. (2004) ‘Improved cemented carbide properties after wire-EDM by pulse shaping’, Journal of Materials Processing Technology,Vol. 149, Nos. 1–3, pp.178–183.

Khan, A.A., Ali, M. and Shaffiar, M. (2006) ‘Relationship of surface roughness with current and voltage during wire EDM’, Journal of Applied Sciences, Vol. 6, No. 10, pp.2317–2320.

Kim, C.H. and Kruth, J-P. (2001) ‘Influence of electrical conductivity of dielectric on wire-EDM of sintered carbides’, KSME International Journal, Vol. 15 No. 12, pp.1676–1682.

Kozak, J., Rajurkar, K.P. and Chandarana, N. (2004) ‘Machining of low electrical conductive materials by wire electrical discharge machining (WEDM)’, J. Mat. Proc. Tech., Vol. 149, Nos. 1–3, pp.266–271.

Lauwers, B., Liu, W. and Eeraerts, W. (2004) ‘Influence of the composition of WC-based cermets on the manufacturability by wire-EDM’, Transactions of the NAMRI/SME, Vol. 32, pp.407–414.

Lenz, E., Katz, E., König, W. and Wertheim, R. (1975) ‘Cracking behavior of sintered carbides during EDM’, Annals of the CIRP, Vol. 24, pp.109–114.

Lin, Y-C., Hwang, L-R., Cheng, C-H. and Su, P-L. (2008) ‘Effects of electrical discharge energy on machining performance and bending strength of cemented tungsten carbides’, Journal of Materials Processing Technology, Vol. 206, Nos. 1–3, pp.491–499.

394 K. Bonny et al.

Llanes, L., Casas, B., Idanez, E., Marsal, M. and Anglada, M. (2004) ‘Surface integrity effects on the fracture resistance of electrical-discharge-machined WC-Co cemented carbides’, Journal of the American Ceramic Society, Vol. 87, No. 9, pp.1687–1693.

Llanes, L., Idanez, E., Martinez, E., Casas, B. and Esteve, J. (2001) ‘Influence of electrical discharge machining on the sliding contact response of cemented carbides’, Int. J. Refr. Met. and Hard Mat., Vol. 19, pp.35–40.

Mahdavinejad, R.A. and Mahdavinejad, A. (2005) ‘ED machining of WC-Co’, Journal of Materials Processing Technology, Vols. 162–163 (Spec. Iss.), pp.637–643.

Obara, H., Satou, H. and Hatano, M. (2004) ‘Fundamental study on corrosion of cemented carbide during wire EDM’, Journal of Materials Processing Technology, Vol. 149, Nos. 1–3, pp.370–375.

Pandey, P.C. and Jilani, S.T. (1987) ‘Electrical machining characteristics of cemented carbides’, Wear, Vol. 116, No. 1, pp.77–88.

Pandit, S.M. and Rajurkar, K.P. (1981) ‘Analysis of electro discharge machining of cemented carbides’, CIRP Annals – Manufacturing Technology, Vol. 30, No. 1, pp.111–116.

Puertas, I., Luis, C.J. and Álvarez, L. (2004) ‘Analysis of the influence of EDM parameters on surface quality, MRR and EW of WC–Co’, Journal of Materials Processing Technology,Vols. 153–154, Nos. 1–3, pp.1026–1032.

Qu, J., Riester, L., Shih, A.J., Scattergood, R.O., Lara-Curzio, E. and Watkins, T.R. (2003) ‘Nanoindentation characterization of surface layers of electrical discharge machined WC-Co’, Materials Science and Engineering A, Vol. 344, Nos. 1–2, pp.125–131.

Stössel-Sittig, C., Schnyder, B. and Kötz, R. (2004) ‘XPS and AES surface characterization of WC-Co after electro-discharge machining’, Surface and Interface Analysis, Vol. 36, No. 8, pp.777–779.

Tsutsui, T. and Tamura, T. (1987) ‘Effect of the electro-discharge machined surface on the mechanical properties – on the fatigue strength of cemented carbide’, Bulletin of the Japan Society of Precision Engineering, Vol. 21, No. 1, pp.70, 71.

Yakou, T. and Hasegawa, T. (1995) ‘Relations between condition and electrically discharge machining and depth of surface crack in cemented carbides’, Trans. Japan Soc. Mec., Vol. 61, pp.1192–1197.

Yu, Z-B., Jun, T. and Kunieda, M. (2004) ‘Dry electrical discharge machining of cemented carbide’, Journal of Materials Processing Technology, Vol. 149, Nos. 1–3, pp.353–357.

edm machinability and dry sliding friction of wc-co cemented … · 2016-01-13 · edm...

Documents