doi:10.1016/j.jmatprotec.2008.05.042

journal of materials processing technology 2 0 9 (2009) 24122420 journal homepage: www.elsevier.com/locate/jmatprotec Wear, cutting forces and chip characteristics when dry turning ASTM Grade 2 austempered ductile ironwith PcBN cutting tools under nishing conditionsK. Katuku , A. Koursaris, I. SigalasSchool of Chemical and Metallurgical Engineering, University of the Witwatersrand, PO Box 3, Wits 2050, Johannesburg, South Africaa r t i c l e i n f o

a b s t r a c t Article history:Received 20 August 2007Received in revised form19 May 2008Accepted 23 May 2008Keywords:ADI PcBN Wear rateCutting forceShear localization

Experimental studies of wear, cutting forces and chip characteristics when dry turning ASTM Grade 2 austempered ductile iron (ADI) with polycrystalline cubic boron nitride (PcBN) cut- ting tools under nishing conditions were carried out. A depth of cut of 0.2 mm, a feed of0.05 mm/rev and cutting speeds ranging from 50 to 800 m/min were used. Flank wear and crater wear were the main wear modes within this range of cutting speeds. Abrasion wear and thermally activated wear were the main wear mechanisms. At cutting speeds greater than 150 m/min, shear localization within the primary and secondary shear zones of chips appeared to be the key-phenomenon that controlled the wear rate, the static cutting forces as well as the dynamic cutting forces. Cutting speeds between 150 and 500 m/min were found to be optimum for the production of workpieces with acceptable cutting tool life, ank wear rate and lower dynamic cutting forces. 2008 Elsevier B.V. All rights reserved.1. IntroductionThe automotive industry, which is extremely competitive, is interested in austempered ductile iron (ADI) because it offers properties similar to those of heat-treated alloy steels. These include high strength, high hardness, excellent toughness, high ductility, good fatigue properties and useful wear char- acteristics at lower cost and reduced weight.Because of these properties, ADI is difcult to machine in the austempered condition. With regards to other engineering ductile cast irons, the relatively high strength and hardness of ADI as well as the inclination of its retained austenite to strain hardening lead to short contact length and higher mechani- cal loads on the cutting tools edge (Yamamoto et al., 1995). The relatively high ductility of ADI favours its adhesion on the cutting tool and brings also about higher temperature on Corresponding author. Tel.: +27 82 262 4035; fax: +27 11 435 3838. E-mail address: tkatku@yahoo.fr (K. Katuku).0924-0136/$ see front matter 2008 Elsevier B.V. All rights reserved. doi:10.1016/j.jmatprotec.2008.05.042

the cutting tools edge (Klocke and Klpper, 2002). Because of higher specic loads and higher temperatures that develop on the cutting tools edge when machining ADI in its austem- pered condition, cutting tools often suffer relatively high ank and crater wears compared to hardened steels and other engi- neering grey cast irons. The severe crater scar that develops very close to the cutting tools edge exposes the latter to frac- ture damage (Pashby et al., 1993). Of course, the higher cutting temperatures as well as the relatively low thermal diffusivity and short contact length of ADI could also expose the cutting tools edge to thermal softening (Gekonde and Subramanian,1995).In these conditions, cutting tools for machining ADI should fundamentally yield at the same time: high hot hardness and strength, excellent hot chemical inertia as well as high toughness. Such cutting tools are ideal and do not rigorouslyspeaking exist in the present cutting technology. However, they are the purpose of the continuous research undertaken in the eld. Coating technology appears to be an alternative. Nowadays, coatings on cutting tools are being used to improve the tribological properties of cutting tools in this ideal way (Knotek et al., 2001). Results show signicant improvements in some cases, especially at relatively low cutting speeds. How- ever, the issue of the mechanical stability (aking) of these coatings sometimes reduces the expectations. Thus, it is a matter of nding a compromise cutting tool material and or coating as well as optimum machining parameters.Machining of ADI in its austempered condition is highly desirable because it can yield the tight tolerances and surface nishes generally required (Klocke et al., 2007), save machin- ing time and thus reduce costs (Klocke and Klpper, 2002). In depth fundamental understanding of interactions involved in this particular machining of ADI in its austempered condition should show the way to the optimum cutting tool material and or coating as well as optimum machining parameters (productive cutting speeds and feed rates, etc.). These out- comes would be among the last obstacles to be overcome before intensifying the use of this material in the automotive industry.This very complex issue attracts great interest from the cutting tool industry since nearly two decades. The compre- hensive research conducted hitherto on the machinability of various grades of ADI is almost very little. It has so far addressed few fundamental questions concerning the cutting performance and wear mechanisms of various types of cutting tools under various machining parameters and conditions.Pashby et al. (1993) investigated on the wear of Al2 O3 , Al2 O3 TiC, Al2 O3 SiCw , and Si3 N4 Al2 O3 ceramic cutting tools when dry turning ADI close to ASTM Grade 2 under condi- tions close to light roughing (depth of cut: 2 mm; feed rate:0.18 mm/rev; cutting speed: 100450 m/min). They reported that ank wear was the main wear mode although tool fracture occurred at the highest speed. Si3 N4 Al2 O3 ceramic cutting tools suffered accelerated wear whereas Al2 O3 SiCw ceramic cutting tools signicantly underperformed Al2 O3 and Al2 O3 TiC ceramic cutting tools under most conditions. Fracture damage on the tools cutting edge and chemical interaction between tool and workpiece were identied as important wear mechanisms in controlling tool life.Masuda et al. (1994) investigated on the cutting per- formance and wear mechanism of P20 cemented carbide cutting tools, Al2 O3 ZrO2 (5 wt.%), Al2 O3 ZrO2 (20 wt.%), Al2 O3 TiC (30 wt.%), Al2 O3 SiCw ZrO2 , Al2 O3 SiCw TiC and Si3 N4 ceramic cutting tools when dry turning ADI close to ASTM Grade 1 under conditions close to light roughing (depth of cut: 1 mm; feed rate: 0.1 mm/rev; cutting speed:50400 m/min). They reported that Al2 O3 TiC (30 wt.%) inserts had the longest life at a low cutting speeds of about 100 m/min and less, and ZrO2 -toughened Al2 O3 inserts had a longer tool life at cutting speeds of about 250 m/min or more. Al2 O3 SiCw ZrO2 and Al2 O3 SiCw TiC ceramic cutting tools exhibited aking fracture at 250 m/min whereas Si3 N4 ceramic cutting tools had no wear resistance at all. Cemented carbide inserts had longer life at very low cutting speeds. As cutting speed rose, the ank wear rate increased slightly for Al2 O3 TiC (30 wt%) inserts. In contrast, it decreased for ZrO2 -toughened

Al2 O3 inserts due to the monoclinic-to-tetragonal transforma- tion of ZrO2 at high cutting temperatures.In order to elucidate the mechanism of poor machinability of ADI, Yamamoto et al. (1995) investigated on the turn- ing of ADI close to ASTM Grade 1 with Al2 O3 SiCw cutting tools under conditions close to light roughing (depth of cut:1.5 mm; feed rate: 0.2 mm/rev; cutting speed: 6300 m/min). Their results showed at cutting speed lower than 36 m/min, the strain-induced residual austenite to martensite transfor- mation occurred in the chips as well as the damaged layer of the machined surface. This strain-induced transformation was responsible of the poor machinability of ADI. At higher cutting speeds this strain-induced transformation occurred only in the damaged layer of the machined surface and not in the chips.Wada et al. (1998) investigated on the wear of coated cemented carbide cutting tools, coated Al2 O3 ceramic cutting tools and coated Si3 N4 ceramic cutting tools in dry turn- ing of ADI close to ASTM Grade 2 under conditions close to light roughing (depth of cut: 1 mm; feed rate: 0.2 and0.4 mm/rev; cutting speed: 30400 m/min). They found that Ti(C,N)Al2 O3 TiN coated P10 carbide inserts had the slowest ank wear progress with regard to TiCAl2 O3 TiN coated P20 and TiN coated K10 carbide inserts. TiN coated Al2 O3 ceramic inserts had tool wear progress similar to Ti(C,N)Al2 O3 TiNcoated P10 carbide inserts. Abrasive wear was observed on theank face of TiN coated Al2 O3 and TiNAl2 O3 TiN coated Si3 N4 ceramic inserts at relatively low cutting speeds. The ank wear of TiNAl2 O3 TiN coated Si3 N4 ceramic inserts increased rather slowly at the high feed rate of 0.4 mm/rev. On the other hand, the TiN coated Al2 O3 ceramic inserts had a tendency to fracture easily at this high feed rate of 0.4 mm/rev.Klocke and Klpper (2002) investigated on the turning of ADI close to ASTM Grade 1 with coated cemented carbide cut- ting tools (Al2 O3 coated K10, Ti(C,N) coated K10, TiNAl2 O3 coated P15), Al2 O3 and Si3 N4 ceramic cutting tools under con- ditions close to light roughing (depth of cut: 1 mm; feed rate:0.2 mm/rev; cutting speed: 120400 m/min) with and without cutting lubricants. They pointed out that coated cemented car- bides cutting tools could be successfully used in the range of low cutting speeds. In the range of high cutting speeds, the use of Al2 O3 ceramic cutting tools was attractive. The perfor-mance of Si3 N4 ceramic cutting tools was very poor. Cuttinglubricants were very effective in the reduction of the ank and crater wear scars of cemented carbide tools, particularly at relatively high cutting speeds.Goldberg et al. (2002) studied the dry interrupted facing of an ASTM Grade 3 ADI with Al2 O3 TiC and Al2 O3 SiCw ceramic cutting tools under conditions close to light roughing (depth of cut: 2 mm; feed rate: 0.10.4 mm/rev; cutting speed:425 m/min) and nishing (depth of cut: 0.5 mm, feed rate:0.10.4 mm/rev; cutting speed: 700 m/min). Their results indi- cated that Al2 O3 SiCw ceramic inserts performed better than Al2 O3 TiC ceramic inserts both for rough interrupted facing and nish interrupted facing at high cutting speeds. The lack of overwhelming performance for Al2 O3 TiC ceramic inserts in this very situation would be linked to their poor thermal shock resistance. They reported that the tool wear characteris- tic was exclusively ank wear which was a direct consequence of adhesiveabrasive wear mechanism.There is strong interest in extending the application of polycrystalline cubic boron nitride (PcBN) cutting tools beyond the traditional machining of hardened steels, ake graphite cast irons and steels produced by powder metallurgy meth- ods (Chou et al., 2003). The machining of ADI, at least under nishing conditions, is among the recent prospects. Indeed, PcBN cutting tools appear to be an alternative for the machin- ing of ADI at high cutting speeds and temperatures. At these high cutting speeds and temperatures, cemented carbide cut- ting tools do not maintain hardness, and Al2 O3 based cutting tools lack to offer adequate toughness (Heath, 1989).In an earlier investigation, Shintani et al. (1990) reported that form the standpoints of surface roughness and tool life, cBN-TiC cutting tools performed better than Al2 O3 TiC ceramic cutting tools for the machining of ADI close to ASTM Grade 3 under nish conditions.Kato et al. (1991) investigated on the wear performance of PcBN cutting tool in the turning of ADI close to ASTM Grade 3 ADI under nishing conditions (depth of cut: 0.2 mm; feed rate: 0.05 mm/rev; cutting speed: 40300 m/min). They observed that from the standpoints of surface roughness and ank wear rate, the optimum cutting speed was 100 m/minaround which the cutting temperature was 827 C. They sug-gested that the cutting performance of PcBN cutting tools was controlled by the size and volume fraction of cBN grains as well as the thickness of the binder phase.The dependence of cutting performance upon size and vol- ume fraction of cBN grains was also corroborated by Goldberg et al. (2002) upon their study of the dry interrupted facing of an ASTM Grade 3 ADI with PcBN cutting tools.Klocke and Klpper (2002) investigated on the dry turn- ing of ADI close to ASTM Grade 1 with PcBN cutting tools under conditions close to light roughing (depth of cut: 1 mm;

These workpieces were characterized in terms of their hardness and microstructure. The microstructure was rst studied with an optical microscope followed by a more detailed examination in a scanning electron microscope (SEM).

Dry nish turning experiments were designed so as to investigate the wear (wear mode, wear mechanism, tool life, cutting length to end of tool life, ratio of volume of metal removed per unit ank wear, ank wear rate) of uncoated Seco CBN 100 PcBN cutting tools when machining ADI workpieces at different cutting speeds, according to the ISO Standard3685-1977(E) for single point turning (International Standard,1977). In conformity with this standard, the wear criterion used for all the machining experiments was 300 m of maxi- mum ank wear.After a certain cutting distance (cutting time), turning was stopped, the cutting tool insert removed from the toolholder and the ank wear and crater wear scar morphologies were assessed by means of microscopic examination on the optical microscope. The maximum width of the ank wear scar was then measured.The insert was then carefully replaced in the toolholder, and the procedure repeated until the tool wear exceeded the criterion. Experiments were repeated in order to measure the cutting forces.Plots of maximum ank wear VBC ( m) against cutting time t (s) as well as plots of volume of ank wear per unit of engagement length V ( m2 ) against cutting length lc (m) were produced.The volume of ank wear per unit of engagement length V corresponding to the geometry of the cutting tools used in this study has been shown to be (Barry and Byrne, 2001):2feed rate: 0.2 mm/rev; cutting speed: 160400 m/min). Theirresults pointed out that the application of PcBN cutting tool was relatively acceptable for cutting speeds in the range of

V = 0.05 (VBC )where VBC is the maximum width of the ank wear scar.

(1)160200 m/min.Data accumulated so far on the machinability of ADI in its austempered condition with PcBN cutting tools allowed obtaining, although partially, fundamental understanding of interactions involved. However, these database need to be extended and consolidated in terms of tool life, wear rate, cutting forces, surface nish and geometric accuracy of the machined components, chip formation mechanisms, surface integrity of the machined components, wear mechanisms of cutting tools, etc. with regards to the recent improvements in the PcBN cutting tool processing technology.This paper focuses on experimental studies of wear (wear mode, wear mechanism, tool life, cutting length to end of tool life, ratio of volume of metal removed per unit ank wear, ank wear rate), cutting forces and chip characteristics when dry turning ASTM Grade 2 ADI with PcBN cutting tools under nishing conditions.2. Experimental proceduresThe ASTM Grade 2 ADI workpieces used had the following chemical composition: 3.51% C, 2.61% Si, 0.19% Mn, 0.016% P,0.009% S, 0.002% Ni, 0.62% Cu, and 0.044% Mg.

Power regression was used to t the plots of VBC against t, to estimate the cutting tool life corresponding to the wear criterion and to determine the Taylor cutting tool life equation. The ank wear rate used in this study was the slope of the plot of V against lc .A tri-axial dynamometer mounted on the turrets lathe andcoupled to a multi-channel amplier was used for the mea- surement of cutting forces.The force signals acquired were analysed so as to evalu- ate the static and dynamic cutting forces corresponding to a time. The static cutting forces were estimated as the average of the signals. The dynamic cutting forces were estimated as the variation from the static cutting forces (Li and Low, 1994).Chips were collected after each cut and examined visu- ally. Their morphology and microstructure were investigated with an optical and a scanning electron microscopes. The chip hardness was measured to assess the interplay between strain hardening and thermal softening.The morphology, microstructure, hardness and average thickness of chips were investigated after mounting in cold resin and metallographic preparation. Bright eld optical microscopy and differential interference contrast (DIC) tech- niques were used. The DIC technique was used to improve the contrast between the primary and secondary shear zones.Transmission electron microscopy was used to examine the nanostructure of the secondary shear zone.Hardness tests of ADI workpieces were done on a ground surface using a Leco V-100-A2 Vickers hardness-testing machine. Hardness tests were done using a load of 30 kgf and a dwell time of 15 s. The etching reagents used for the microstructural investigation of workpieces were 3% Nital and4% Picral.Dry nish turning experiments were carried out on a LA 200L Liouy-Hsing CNC Lathe rated at 14.72 kW. The uncoated Seco CBN 100 PcBN cutting tool inserts used contained about 50% by volume of cBN, 2 m in grain size and TiC binder (Seco, 2006). Their index specication was SNGN 090312 S (nose radius1.2 mm, honed, chamfer 0.1 mm 20). They were mountedon a toolholder described as CSDNN 2525M12C. The combina- tion of insert and toolholder resulted in a rake angle of 26, a clearance angle of 6 and an approach angle of 45.A Quartz 3-Component Kistler Dynamometer Type 9257B was used for the measurement of cutting forces. A multi- channel Kistler amplier Type 5070 A was coupled to the dynamometer. Dynoware acquisition software Type 2825A-02Version 2.4.1.3 was used for data logging.The dry nish turning experiments were carried out using a depth of cut of 0.2 mm which is within the dimensions of the tool chamfer. The feed was constant at 0.05 mm/rev and cutting speeds ranged from 50 to 800 m/min. The sampling rate of the analogical input force signals was 500 Hz. The force signals acquired were analysed for a cutting time of 1 s.Chip hardness measurements were made with a Leco M-400A microindentation hardness-testing machine. A load of50 gf and a dwell time of 10 s were used for chip hardness mea- surements. The average chip thickness was measured using the image analysis software AnalySIS 5. Chip etching was car- ried out in 3% Nital.An Olympus BX41M optical microscope coupled to an Olympus Camedia Camera was used to measure the width of the ank wear scar and to obtain optical micrographs. A Philips XL 30 ESEM-FEG XL series scanning electron micro- scope was used to obtain SEM micrographs.Chip samples for transmission electron microscopy were prepared with focused ion beam (FIB) using 30 kV Ga ions in a dual beam Nova FIB with in situ lift-out. A Philips 420 scan- ning transmission electron microscope (STEM) using a LaB6 electron source at 120 kV was used for this investigation. STEMimages were captured in bright eld mode.3. Results3.1. Workpiece characterizationThe hardness of the ADI workpieces was 312 HV30 . The principal micro-constituents were ausferrite (ferrite needles and stringer-like retained austenite) and islands of residual austenite and graphite (Fig. 1).Image analysis of optical micrographs of unetched sam- ples gave a volume fraction of graphite of 15%, a distribution of graphite particles of 190 nodules/mm2 , a mean graphite nodule size of 30 m and a graphite nodule spacing of72 m.

Fig. 1 Microstructure of the workpiece (ASTM Grade 2ADI). Etched, Picral, SEM, secondary electron image.3.2. Tool wear and cutting forcesDamage to the cutting tools over the entire range of cutting speeds was mainly in the form of ank and crater wears. Early formation of crater scar was noticeable at cutting speeds greater than 150 m/min.The Taylor cutting tool life equation was derived from the plot in Fig. 2 ast = 3 107 v1.9781 , with R2 = 0.9958 (2)where t is the tool life (s), and v the cutting speed (m/min). R2is the regression coefcient.The ank wear rate and the ratio of volume of metal removed per unit of ank wear are shown in Fig. 3. The ratio of volume of metal removed per unit ank wear decreased rapidly with increasing cutting speed up to a speed of about300 m/min and more slowly with higher speeds. The ank wear rate showed a rapid increase with increasing cutting speed up to a speed of about 200 m/min. In the range of200550 m/min the ank wear rate increased approximately linearly with cutting speed and more rapidly for higher speeds.

Fig. 2 Effect of cutting speed on tool life of Seco CBN 100PcBN cutting tools.

Fig. 3 Effect of cutting speed on ank wear rate and the ratio of volume of metal removed per unit of ank wear of Seco CBN 100 PcBN cutting tools.

Fig. 4 Effect of cutting speed on static cutting forces forSeco CBN 100 PcBN cutting tools.The static and dynamic cutting forces corresponding to a cutting time of 1 s are respectively shown in Figs. 4 and 5.The static cutting forces were low. They increased rapidly with increasing cutting speed up to a speed of about150 m/min. Between speeds of 150 and 200 m/min, they showed a dramatic decrease. At speeds over 200 m/min, they increased at a slow rate. For speeds of over 150 m/min, the static thrust force was higher than the static tangential force.The dynamic cutting forces decreased with increasing cut- ting speed up to a speed of about 200 m/min and increased at different rates at higher cutting speeds. The dynamic thrust force increased at the highest rate.3.3. Chip characteristicsBluing and chip oxidation started to be evident at cutting speeds greater than 200 m/min. In fact, from 200 m/min, chips

Fig. 5 Effect of cutting speed on dynamic cutting forces forSeco CBN 100 PcBN cutting tools.started glowing in subdued light, indicating a temperature of400 C or higher (Sizes, 2007).At a speed of 50 m/min the surface nish of the workpiece was very poor and the sliding surface of chips was very rough (Fig. 6). Some severely strained material was evident on the sliding surface of chips as was the case with the machined surface of the workpiece. The heavily deformed material indi- cated excessive or erratic built up edge (BUE) that periodically formed on the cutting edge resulting in poor surface nish of the workpiece.At a cutting speed of 100 m/min the amount of strained material on the underside of the chip decreased. However there always were big discontinuities in the deformed material in areas occupied by graphite nodules.At speeds of 150 m/min or more the streaks along the length of the sliding side of chips, were smooth (Figs. 7 and 8). The extent of discontinuities on the sliding side of chips decreased as the cutting speed increased. Streaks and micro- pores were more evident as the cutting speed increased as a result of softening and probably partial melting due to sliding of the chip against the rake face of the cutting tool (Farhat,2003).The chips were continuous with occasional segmentation and highly coiled for cutting speeds of up to 150 m/min. The deformation pattern revealed that deformation occurred quite homogeneously in the entire chip. The ow-pattern in the sec-

Fig. 6 Underside of chip obtained at 50 m/min. SEM, secondary electron image.

Fig. 7 Underside of chip obtained at 150 m/min. SEM, secondary electron image.

Fig. 10 Microstructure of a chip obtained at 700 m/min. SEM, secondary electron image.

Fig. 8 Underside of chip obtained at 700 m/min. SEM, secondary electron image.ondary shear zone, which showed the level of shearing and frictional energy close to the toolchip interface, was also vis- ible (Fig. 9).For cutting speeds between 150 and 200 m/min, the chip characteristics were intermediate between those of continu- ous and segmented chips.At cutting speeds above 200 m/min, the chips became more and more segmented and less coiled (Fig. 10). The segmented

Fig. 9 Microstructure of a chip obtained at 150 m/min. SEM, secondary electron image.

Fig. 11 STEM image of the secondary shear zone of chips obtained at 800 m/min, bright eld image.chips consisted of individual segments that were slightly deformed and joined by narrow heavily strained bands. The high strain was localized essentially in the primary and sec- ondary shear zones.Examination in the STEM of the secondary shear zone in chips produced at a cutting speed of 800 m/min revealed nano-sized recrystallized grains (Fig. 11). The massive defor- mation expected in the secondary shear zone was not evident.Plots of average chip thickness and chip hardness against cutting speed are shown in Fig. 12. It can be seen that the average chip thickness decreased rapidly with increasing cut- ting speed up to about 200 m/min and very slowly for higher speeds.Conversely chip hardness increased for speeds of up to200 m/min and dropped precipitously between speeds of 200 and 300 m/min. At higher cutting speeds the chip hardness uctuated signicantly and did not show a clear trend. These uctuations are probably the result of actual variations in hardness within chips, due to variations in the extent of strain hardening.

Fig. 12 Effect of cutting speed on average chip thickness and hardness of ASTM Grade 2 ADI chips.4. Discussion of results4.1. Tool wearThe value of R2 in Eq. (2) indicates good agreement between the Taylor equation and the experimental tool life results over the range of cutting speeds used.The absolute value of the velocity exponent (1.9781) is much higher than 1 which is the absolute value of this expo- nent in cases where abrasive wear is the dominant wear mechanism (Arsecularatne et al., 2006). The higher value of this exponent indicates that abrasion was not the dominant wear mechanism throughout the range of cutting speeds used.It is well known that in the Taylor equation, the cutting speed inuences tool life through its effect on temperature (Arsecularatne et al., 2006). It is then likely that thermally acti- vated wear mechanisms were active over a wide interval of the range of cutting speeds used.The plot of ank wear rate against cutting speed (Fig. 3) showed signicant changes in slope for cutting speeds of 200 and 600 m/min indicative of changes in the dominant tool wear mechanism. The ank wear rate increased rapidly for speeds of up to 200 m/min and slowly for speeds between 200 and 600 m/min. At speeds over 600 m/min there was a sub- stantial increase in ank wear rate.When considering chip hardness and average chip thick- ness (Fig. 12) interesting trends emerge with regard to the interplay between strain hardening and thermal softening. At speeds of up to 100 m/min chip hardness increased due to strain hardening. The practically constant hardness of chips produced at 100200 m/min was probably due to increased chip temperature resulting in a balance between strain hard- ening and recovery processes. At speeds between 200 and300 m/min, the hardness dropped substantially probably due

to recrystallization in deformed chips. At speeds in excess of300 m/min the hardness appeared to be constant indicating recrystallization in the chips. This appears to be consistent with the nearly constant chip thickness obtained at speeds of between 300 and 800 m/min (Fig. 12).When considering ank wear rate (Fig. 3), chip hardness and average chip thickness (Fig. 12) strong indications emerge with regard to tool wear mechanisms.At speeds of up to 200 m/min temperatures are low, the chip strain-hardens and abrasion (a mechanical effect) would be expected to be the dominant wear mechanism. At speeds between 200 and 300 m/min temperatures are high and clear evidence of recrystallization emerges. Under these conditions diffusion wear is expected to dominate. This mechanism appears to remain dominant at speeds of up to about 600 m/min. The increased ank wear rate at speeds in excess of 600 m/min can probably be attributed to further temperature increases which facilitate diffusion but also accelerate oxidation which becomes a signicant contributor to tool wear. Liquation at these high cutting speeds can also be expected to enhance wear of the tool. Clearly there are synergistic effects between the three phenomena.4.2. Cutting forcesThe low feed (0.05 mm/rev), low depth of cut (0.2 mm) and low contact length (contact area) explain the low values of cutting force (Trent and Wright, 2000).Kinks on the curves of static cutting force at cutting speeds lower than 150 m/min (Fig. 4) show the effect of the erratic or excessive BUE that probably occurred. In fact, during the cutting operation, the BUE acts as an extension of the cutting tool (Trent and Wright, 2000) and it usually reduces abnormally the static cutting force by restricting the contact of the chip with the cutting tool.The sudden drop in static cutting forces between 150 and200 m/min could be the result of the decrease in the contact area between the chip and the cutting tool and of the thermal softening in the secondary shear zone.Above 200 m/min the balance between the effects of strain hardening and thermal softening resulted in a very slight increase in static cutting force.For speeds of 150 m/min or more, the static thrust force was bigger than the static tangential force because of the largenegative rake angle (Poulachon and Moisan, 2000) (26) andprobably due to a higher crater wear rate.At cutting speeds lower than 150 m/min, the fragmentation of chips and the instability of the BUE (formation and fracture) induced additional dynamic cutting forces.The decrease in dynamic cutting force in the range between50 and 200 m/min could be attributed to the decrease in break- ing frequency of chips.Beyond 150 m/min chips were no longer short but started to become segmented and the wear rate increased signicantly. The segmentation of chips, the ank wear as well as the crater wear imply an increase in dynamic cutting force.Thus, in order to produce workpieces at lower dynamic cut- ting force (better surface nish and dimensional accuracy), cutting speeds in the range of 150500 m/min are indicated.4.3. Chip morphologyThe gradual, continuous and asymptotic decrease in the aver- age chip thickness (Fig. 12) was due to the gradual, continuous and asymptotic increase of the shear angle that occurs when the cutting speed increases. The increase in shear angle with increased cutting speed is linked to the strain hardening in the primary shear zone or according to Oxleys model, to the decrease of ow stress (thermal softening) in the secondary shear zone (Subramanian et al., 2002).The asymptotic decrease in average chip thickness (Fig. 12) accords with the formation of segmented chips (Barry and Byrne, 2002). Since ADI is a relatively ductile material, the formation of segmented chips (Fig. 10) is related to shear local- ization in the primary or/and secondary shear zones. Shear localization was revealed metallographically in the primary and secondary shear zones by the presence of shear bands whose microstructure contrasted clearly with that of the chip segments (Fig. 10).The nano-sized equiaxed grains that appeared in the sec- ondary shear zone of chips at cutting speeds greater than150 m/min (Fig. 11) were the result of dynamic recovery and recrystallization that occurred in this zone. At cutting speeds greater than 150 m/min, shear bands in the secondary shear zone probably transformed partially.The renement of the microstructure of the primary and secondary shear zones may be expected to increase the dif- fusion rates in the chip and enhance the diffusion of cutting tool constituents into the chips.At cutting speeds greater than 150 m/min, friction or/and seizure at the toolchip interface gave rise to high temper- ature that activated the onset of shear localization in the secondary shear zone. The temperature in this zone increased even further with the onset of shear localization. The resul- tant high temperature which favoured transformation and dynamic recrystallization in the secondary shear zone, also favoured thermally activated wear on the crater face of the PcBN cutting tools. This was evidenced by the early appear- ance of the crater wear scar at cutting speeds greater than150 m/min.However, shear localization in the secondary shear zone was nearly reduced by the partial melting that occurred on the chip underside, through its lubrication effect. In fact, the increase in average chip thickness that should be favoured by the onset of shear localization in the secondary shear zone (Subramanian et al., 1999) was negligible above a cut- ting speed of 200 m/min (Fig. 12). This conrms the formation of lubricating transfer layer that maintains more or less unchanged the toolchip interface tribological conditions and consequently a nearly constant average chip thickness (Rech,2006).The partial melting that occurred on the chip underside could be expected to increase the rate of thermally activated wear of the crater face of the PcBN cutting tools.5. Conclusions(1) Flank wear and crater wear were the main wear modes for cutting speeds in the range 50800 m/min.

(2) The absolute value of the velocity exponent of the Tay- lor cutting tool life equation suggested that abrasion wear and thermally activated wear were the main wear mech- anisms. These indications on tool wear mechanisms also emerged when considering the ank wear rate, chip hard- ness and average chip thickness curves.(3) At cutting speeds less than 150 m/min, abrasion wear was the main wear mechanism. At these cutting speeds, the fragmentation of chips and the instability of the BUE, con- trolled the dynamic cutting forces.(4) At cutting speeds greater than 150 m/min, shear local- ization within the primary and secondary shear zones of chips appeared to be the key-phenomenon that con- trolled the wear rate, the static cutting forces as well as the dynamic cutting forces.(5) At cutting speeds greater than 150 m/min, the higher tem- peratures subsequent to shear localization brought about nano-sized equiaxed grains within the secondary shear zone, via dynamic recovery, recrystallization and proba- bly partial phase transformation. These nano-sized grains probably increased the crater wear rate of the PcBN cutting tools via diffusion route.(6) At cutting speeds greater than 150 m/min, the higher tem- peratures subsequent to shear localization favoured the partial melting of the chip underside. This partial melting could be expected to increase the crater wear rate of the PcBN cutting tools.(7) At cutting speeds greater than 600 m/min, the higher temperatures subsequent to shear localization probably favoured the oxidation wear of PcBN cutting tools.(8) Cutting speeds between 150 and 500 m/min were found to be optimum for the production of workpieces with accept- able cutting tool life, ank wear rate and lower dynamic cutting forces.AcknowledgementsThe authors would like to express their thanks to the DST/NRF Centre of Excellence in strong materials, University of the Wit- watersrand, for nancial support.references Arsecularatne, J.A., Zhang, L.C., Montross, C., 2006. Wear and tool life of tungsten carbide PcBN and PCD cutting tools. International Journal of Machine Tools & Manufacture 46,482491.Barry, J., Byrne, G., 2001. Cutting tool wear in the machining of hardened steels. Part I. Alumina/TiC cutting tool. Wear 247,139151.Barry, J., Byrne, G., 2002. The mechanism of chip formation in machining hardened steels. Transactions of the ASME 124,528535.Chou, Y.K., Evans, C., Barash, M.M., 2003. Experimental investigation on cubic boron nitride turning of hardened AISI52100 steel. Journal of Materials Processing Technology 134,19.Sizes, Colours of heated metals, 2007. http://www.sizes.com/materls/colors of heated metals.htm.Farhat, Z.N., 2003. Wear mechanism of cBN cutting tool during high-speed machining of mold steel. Materials Science and Engineering A361, 100110.Gekonde, H.O., Subramanian, S.V., 1995. High speed machining of ductile iron. AFS Transactions 120, 309317.Goldberg, M., Berry, J.T., Littlefair, G., Smith, G., 2002. A Study of the machinability of an ASTM Grade 3 austempered ductile iron. In: Proceedings of the 2002 World Conference on ADI for Casting Producers, Suppliers and Design Engineers, Ductile Iron Society and American Foundry Society, Galt House Hotel, Louisville, Kentucky, USA.Heath, P.J., 1989. Ultrahard tool materials. In: Metals Handbook9th Edition Vol. 16-Machining. ASM International, USA, pp.105117.International Standard, 1977. ISO 3685-1977(E)Tool Life Testing with Single Point Turning Tools. Switzerland.Kato, H., Shintani, K., Fujimura, Y., 1991. Wear performance of cBN tool in machining of ADI (effect of tool life on sintered elements). Transactions of the Japan Society of Mechanical Engineers C 57 (541), 30273031 (in Japanese).Klocke, F., Klpper, C., 2002. Machinability characteristics of austempered ductile iron (ADI). In: Proceedings of the 2002World Conference on ADI for Casting Producers, Suppliers and Design Engineers, Ductile Iron Society and American Foundry Society, Galt House Hotel, Louisville, Kentucky, USA.Klocke, F., Klopper, C., Lang, D., Essig, C., 2007. Fundamental wear mechanisms when machining austempered ductile iron(ADI). Annals of the CIRP 56 (1).Knotek, O., Lofer, F., Kramer, G., 2001. Applications to cutting tools. In: Bunshah, R.F. (Ed.), Handbook of Hard Coatings, Deposition Technologies, Properties and Applications. Noyes Publications Park Ridge/William Andrew Publishing/LLC, New Jersey, USA/Norwich/New York, USA, pp. 370410.

Li, X.S., Low, I.M., 1994. Cutting forces of ceramic cutting tools.Key Engineering Materials 96, 81136.Masuda, M., Sato, T., Kori, T., Chujo, Y., 1994. Cutting performance and wear mechanism of alumina-based tools whenmachining austempered ductile iron. Wear 174, 147153. Pashby, I.R., Wallbank, J., Boud, F., 1993. Ceramic tool wear when machining austempered ductile iron. Wear 162164, 2233.Poulachon, G., Moisan, A.L., 2000. Hard turning: chip formation mechanisms and metallurgical aspects. Transactions of the ASME 122, 406412.Rech, J., 2006. Inuence of cutting tool coatings on the tribological phenomena at the toolchip interface in orthogonal dry turning. Surface & Coatings Technology 200, 51325139.Seco, 2006. Secomax PcBN Technical Guide 2006, Seco Tools AB, SE-737 82 Fagersta, Sweden, p. 5.Shintani, K., Kato, H., Fujimura, Y., Yamamoto, A., 1990. Cutting performance of cBN tools in machining of austempered bainitic spheroidal graphite cast iron. Journal of the Japan Society of Precision Engineering 56 (12), 22612266 (in Japanese).Subramanian, S.V., Gekonde, H.O., Zhang, X., Gao, J., 1999. Design of steels for high speed machining. Ironmaking and Steelmaking 5 (26), 333338.Subramanian, S.V., Gekonde, H.O., Zhu, G., Zhang, X., 2002. Role of microstructural softening events in metal cutting. Machining Science and Technology 6 (3), 353364.Trent, E.M., Wright, P.K., 2000. Metal Cutting, 4th ed. ButterworthHeinemann, pp. 4279.Wada, T., Fujiwara, J., Hanasaki, S., Yasutomi, M., 1998. Tool wear in cutting of ADI. Transactions of the Japan Society of Mechanical Engineers C 64 (626), 40654071 (in Japanese).Yamamoto, S., Nakajima, H., Miyaji, H., 1995. Mechanisms of poor machinability of austempered ductile iron. The Iron and Steel Institute of Japan 81 (7), 721726 (in Japanese).