Experimental investigation on CBN turning of hardened AISI 52100 steel

Experimental investigation on CBN turning of hardened AISI 52100 steel

Journal of Materials Processing Technology 124 (2002) 274±283 Experimental investigation on CBN turning of hardened AISI 52100 steel Y. Kevin Choua,*...

665KB Sizes 2 Downloads 37 Views

Journal of Materials Processing Technology 124 (2002) 274±283

Experimental investigation on CBN turning of hardened AISI 52100 steel Y. Kevin Choua,*,1, Chris J. Evansa, Moshe M. Barashb a

Manufacturing Engineering Laboratory, National Institute of Standards and Technology, Gaithersburg, MD, USA b School of Industrial Engineering, Purdue University, West Lafayette, IN, USA Accepted 2 April 2002

Abstract This study investigated the performance and wear behavior of different cubic boron nitride (CBN) tools in ®nish turning of hardened AISI 52100 steel. Tool performance was evaluated based on the part surface ®nish and the tool ¯ank wear. Wear conditions of CBN cutting tools were primarily characterized by scanning electron microscopy (SEM). Machining results showed that low CBN content tools (CBN-L) consistently perform better than high CBN content counterparts (CBN-H), despite the CBN-L has inferior mechanical properties. The ¯ank wear rates were proportional to cutting speed and CBN-H showed accelerated thermal wear associated with high cutting temperatures. Reducing depth of cut would only improve surface ®nish to CBN-L, but not to CBN-H, despite similar wear rates. The transferred layer on the ¯ank wear land may result in adhesion of the binder compound and signi®cantly affect the tool wear process. The metallic binder in CBN-H has stronger af®nity to the transferred layer and may result in plucking out of CBN particles and consequent severe abrasive wear. # 2002 Elsevier Science B.V. All rights reserved. Keywords: Cubic boron nitride; Hard turning; Tool wear

1. Introduction The commercialization of cubic boron nitride (CBN) tools since 1970s has resulted in a rapid advancement of hard machining technology [1]. Among the applications of CBN cutting tools precision hard turning attracts great interests since it potentially provides an alternative to conventional grinding in machining high precision, high hardness components in small volume production. Cutting mechanics in hard turning with CBN tools is different from conventional machining and has been frequently studied. It has been consistently reported that saw-tooth chips, due to catastrophic localized shear, are produced in hard turning [2]. Furthermore, because of high strength of hardened steel and brittleness of CBN tools, a chamfered cutting edge, which leads to a large negative rake angle, is employed to increase the wedge angle and thus, tool strength. The large negative rake angle will, however, lead to more severe plastic strain and consequently high cutting temperatures, * Corresponding author. Present address: Department of Mechanical Engineering, The University of Alabama, 290 Hardaway Hall Box 870276, 35487-0276 Tuscaloosa, AL, USA. Tel./fax: ‡1-205-348-0044/6419. E-mail address: [email protected] (Y.K. Chou). 1 Currently an Assistant Professor in Mechanical Engineering Department at the University of Alabama.

giving a negative impact to the tool performance. Several advantages associated with precision hard turning have been reported [3] including low production cost, part longevity, environmental consciousness, etc. However, one major challenge is to maintain surface ®nish and dimension/form accuracy with a reasonably long tool life. Many researchers have suggested that tool materials and the control of tool wear are of decisive importance [4]. CBN tool wear has been frequently studied. Due to great hardness and abrasive resistance, CBN tools generally have greater wear resistance than conventional tool materials such as carbide, cermet, and ceramic. CBN tools can be roughly divided into two groups: high CBN content (90 vol.%) with a metallic binder (noted as CBN-H), and low CBN content (50±70 vol.%) with ceramic as a major binder (noted as CBN-L). One puzzling phenomenon in CBN tool wear has been that in ®nish hard turning, CBN-L gives longer tool life and produces better surface ®nish (typically 0.25 mm Ra or better) than CBN-H. This is surprising because CBN-H has greater hardness and fracture toughness than CBN-L [5]. There have been many different explanations toward this puzzling phenomenon. Some authors suggested that the longer tool life of CBN-L is due to greater bonding strength [6±8], some proposed that the welded layers on the tool ¯ank wear land of CBN-L create a protection effect [9], and others

0924-0136/02/$ ± see front matter # 2002 Elsevier Science B.V. All rights reserved. PII: S 0 9 2 4 - 0 1 3 6 ( 0 2 ) 0 0 1 8 0 - 2

Y.K. Chou et al. / Journal of Materials Processing Technology 124 (2002) 274±283 Table 1 Characteristics of different CBN tools [13] Characteristics

BZN6000

BZN8100

CBN content (vol.%) Major binder phase Grain size (mm) Knoop hardness (GPa)a Transverse rupture strength (GPa)

92 Cobalt 2±4 28 0.9

70 Titanium nitride 2±4 25 1.0

a Knoop hardness: 3 kg load; transverse rupture strength: three-point bending test.

attributed tool wear to greater plastic deformation and higher defect density in CBN-H [10,11]. Bossom [5] noted the lower thermal conductivity of CBN-L and suggested resultant softening of workpieces in the shear zone as another explanation. Current hypotheses based on mechanical and physical properties, however, do not adequately describe the observed phenomenon. Rather, microstructure and chemistry might have a bearing on CBN tool wear behavior. Recently, chemical and microstructural aspects have been widely considered in tool wear studies. Klimenko et al. [12] indicated that in machining with CBN tools, chemical reactions occurred in the contact zone may alter the composition of the materials in contact and affect the tool wear process. By conducting equivalent machining experiments on both CBN-H and CBN-L, this research investigated differences in the performance and wear process of both types of CBN tools. Scanning electron microscopy (SEM) and X-ray photoelectron spectroscopy (XPS) were used to examine the microstructures and to identify the chemical composition at the tool surfaces. 2. Experimental details Two types of CBN tools, GE BZN6000 (CBN-H) and BZN81002 (CBN-L), were mainly used in this study to evaluate the tool performance and wear mechanisms of CBN tools in ®nish turning of hardened steels. Other CBN-L's (Sumitomo BN200 and BN3002) were also tested at one particular cutting condition for comparison purpose. The characteristics of BZN6000 and BZN8100 are listed in Table 1 [13]. BZN6000 has a higher CBN content, greater hardness, and contains cobalt (Co) as the binder, whereas BZN8100 has titanium nitride (TiN) binder phase and a small trace of Co. The microstructures of two types of CBN tools, after polishing and gold coating, are shown in Fig. 1. Dark particles are CBN grains, light areas are Co compounds, and the gray phase in BZN8100 is TiN. Workpieces were through hardened rings (94.0 mm outer diameter and 68.6 mm inner) made of AISI 52100 steel hardened and 2 Specific products are identified to completely specify the experimental conditions. Such identification does not imply their endorsement by the National Institute of Standards and Technology, nor that they are necessarily the best available for the purpose.

275

tempered to 61±63 HRC hardness. The machining process was dry facing conducted on a Hardinge Superslant2 computer numerical control turning center. Fifty-®ve degrees diamond-shaped inserts with a 25  0:1 mm chamfered cutting edge and 0.8 mm tool nose radius were used. Combination of the insert and the tool holder resulted in a 308 rake angle and a 58 clearance angle. Three cutting speeds (60, 120, and 240 m/min) and three depths of cut (10, 50, and 250 mm) with a ®xed feed rate (12.5 mm/rev) were tested. In ®nish hard turning, the cutting action takes place exclusively in the tool nose radius, and therefore, the cutting edge angles are determined by the tool nose radius and the depth of cut, irrelevant to the angles provided by the insert and the tool holder [14]. With 0.8 mm nose radius and 50 mm depth of cut, the major and minor cutting edge angles are 708 and 208, respectively. Arrangements of cutting parameters are listed in Table 2. For each cutting set, 10 cutting passes (12.7 mm cutting length per pass) were conducted on one insert, and ®ve inserts were used in each set to obtain statistical results. The machine used has both functions of constant cutting speed and constant feed per revolution, and thus, even with a facing operation where the tool moves across the radial direction, constant cutting speed and feed rate can be maintained and used to evaluate the machining performance. It is also known that changes of workpiece rotational rate while keeping constant cutting speed during facing will not affect part surface ®nish. The tool performance was evaluated based on the part surface ®nish and the tool ¯ank wear. Surface ®nish is one of the most stringent requirements placed on ®nishing processes, and surface ®nish and form accuracy generally degrade with tool wear. During machining tests, the average surface roughness (Ra) of the machined parts in the direction of feed and the maximum ¯ank wear land width (VBmax) of the CBN tools were measured and recorded after each pass (facing one ring). A surface roughness gage with a piezoelectric probe (0.8 mm cut-off length and 6.47 mm traverse length) was used to measure Ra and a high-power optical microscope to measure VBmax. Topography of worn tool surfaces can often infer which mechanisms have been dominant during the tool wear process. In this investigation, SEM was used for frequent examinations of worn CBN inserts. SEM photographs consistently showed a transferred layer on the ¯ank wear land of CBN tools after some cutting time. The transferred layer seemed to adhere to the tool surface. In order to examine the underneath structures, the transferred layer was leached by Table 2 Process parameters in different machining sets Cutting test

Cutting speed (m/min)

Feed rate (mm/rev)

Depth of cut (mm)

Tool material effect Cutting speed effect Depth of cut effect

120 60, 120, 240 120

12.5 12.5 12.5

50 50 10, 50, 250

276

Y.K. Chou et al. / Journal of Materials Processing Technology 124 (2002) 274±283

Fig. 1. Microstructure of CBN tool materials: (a) BZN6000; (b) BZN8100.

placing inserts in a 50±60% sulfuric acid solution at about 60 8C. On the other hand, to identify the chemical composition at tool surfaces, XPS with an X-ray source of a monochromatic aluminum Ka line (energy 1486.6 eV, width 0.86 eV) was used. The elemental composition and chemical shifts at the surface of CBN tools and transferred layers before and after leaching were identi®ed from XPS spectra. 3. Results and discussion 3.1. Tool performance Fig. 2 shows surface roughness and maximum ¯ank wear land width versus cutting length for CBN tools at 120 m/min and 50 mm (feed rate was 12.5 mm/rev, constant through out the experiments). For CBN-L such as BZN8100, Ra remained below 0.4 mm for the entire 127 mm cutting length (21 min of cutting time), and VBmax reached about 0.25 mm. The error bars in the ®gures represent one standard deviation in both directions. For BZN6000, both Ra and VBmax were higher and showed larger variations.

Fig. 3 shows Ra changes with cutting length when machining at different cutting speeds. The results indicated that surface ®nish degrades with increasing cutting speeds. Fig. 4 shows VBmax increase versus cutting length at different cutting speeds. A linear ®t of the data gives average wear rates (DVBmax/Dcutting length) which are shown in Fig. 5 at various cutting speeds. The ¯ank wear rate seems to be linearly proportional to the cutting speed. In addition, the difference in the wear rate between BZN6000 and BZN8100 increases at higher cutting speeds. Moreover, the wear rate of BZN8100 at 240 m/min is close to that of BZN6000 at 60 m/min. By comparing the tempered color of the chips from these two cases, the cutting temperature of BZN8100 at 240 m/min would be higher than BZN6000 at 60 m/min. This indicates that BZN8100 has a greater thermal wear resistance than BZN6000 at high cutting temperatures. Ra and VBmax were also evaluated at different depths of cut. For BZN6000, depth of cut does not have signi®cant effects on the surface roughness. While for BZN8100, the smaller the depth of cut, the ®ner is the surface ®nish. Ra could be kept below 0.2 mm at 10 mm depth of cut. Flank wear of BZN8100 slightly increased with increasing depth

Fig. 2. (a) Surface roughness and (b) flank wear land versus cutting length.

Y.K. Chou et al. / Journal of Materials Processing Technology 124 (2002) 274±283

277

Fig. 3. Surface roughness versus cutting length at different cutting speeds: (a) BZN6000; (b) BZN8100.

Fig. 4. Flank wear land versus cutting length at different cutting speeds: (a) BZN6000; (b) BZN8100.

of cut. For BZN6000, ¯ank wear land at 10 mm depth of cut was notably low. Flank wear values at different depths of cut were also linearized to estimate wear rates. As depth of cut increases from 50 to 250 mm, ¯ank wear rates show very minor changes. On the other hand, at 10 mm depth of cut, the wear rates of both tools drop remarkably and are almost identical. Even though a signi®cant reduction in the ¯ank wear size and the wear rate of BZN6000 at 10 mm depth of cut, no improvement in surface ®nish was observed. 3.2. Tool wear conditions

Fig. 5. Flank wear rate versus cutting speed.

In ®nish machining, the tool wear zone occurs only in the tool nose radius corner. General wear patterns observed from an SEM at a low magni®cation are crater zone, ¯ank wear land, and a transferred layer at ¯ank wear land. At a higher

278

Y.K. Chou et al. / Journal of Materials Processing Technology 124 (2002) 274±283

Fig. 6. Worn cutting edge of CBN tools (200): (a) BZN6000; (b) BZN8100.

magni®cation, both the ¯ank wear and crater zone show mostly polished-like surfaces. Fig. 6 is scanning electron micrographs showing the worn cutting edges of BZN6000 and BZN8100 machined at 120 m/min and 50 mm for 127 mm cutting length. In addition to a larger wear scale of BZN6000, the transferred layer adhered at the ¯ank wear land has different characteristics. For BZN8100, the transferred layer is smooth and uniform, while the transferred layer on BZN6000 has a rather rough topography with noticeable grooves. It is suggested that the grooves were resulted from abrasion by hard particles such as wear debris or possibly plucked-out CBN particles from the tool itself. The morphology of the transferred layers also changed with the cutting speed. When the cutting speed was reduced to 60 m/min, the transferred layer on the CBN tools became thinner, and fewer grooves were seen on BZN6000. When the cutting speed was increased to 240 m/min, an even rougher surface with more severe grooves was seen

Fig. 7. Worn cutting edge of CBN tools cutting at 240 m/min (200): (a) BZN6000; (b) BZN8100.

on BZN6000 (Fig. 7(a)), while the transferred layer on BZN8100 seemed to only grow thicker (Fig. 7(b)). It is suspected that when the cutting speed is increased (so is the cutting temperature), more hard particles are involved in the wear of BZN6000 which severely gauge through the ¯ank wear land, thus aggravating the tool wear process. Fig. 8 shows the details of the transferred layer of CBN tools after cutting at 240 m/min and 50 mm for 127 mm. On BZN6000, many large-scale grooves are evident. However, on BZN8100, the transferred layer shows a ¯ake-type structure. This suggests that the transferred layer on BZN8100 is not as strongly bonded to the tool as on BZN6000, and this may result in reduced friction between the tool ¯ank and the newly machined surface. With regard to depth of cut effects, no dramatic difference in wear features was observed. However, the larger the depth of cut, the deeper is the crater zone and the longer is the wear land length. The microstructures on the ¯ank wear land (Fig. 9) were clearly exposed when cutting at a low speed (60 m/min).

Y.K. Chou et al. / Journal of Materials Processing Technology 124 (2002) 274±283

Fig. 8. Microstructures on flank wear land (5000): (a) groove structure on BZN6000; (b) flake-type structure on BZN8100.

On BZN6000, clear sliding marks are seen, and the CBN grains and cobalt phase can be identi®ed. On BZN8100, several shallow holes appear on the ¯ank wear land in places previously occupied by CBN grains. This seems to imply

279

that parts of CBN grains were plucked out by adhesion due to loss of bonding. A simple model of adhesive wear with the transferred layer interactions is shown in Fig. 10. The leaching process also led to different results on different CBN tools. The transferred layer of BZN8100 was relatively easy to be cleaned-off within a few hours, depending upon cutting conditions used. For BZN6000, some of the transferred layer could not be dissolved even after a longer soaking. Note that the parent CBN tool surfaces were not affected by leaching. The microstructures of the leached wear land exhibited signi®cant differences. For BZN6000, some of the cobalt compounds was removed and ®ne grooves with a thin ®lm were revealed in some regions (Fig. 11(a)). On the other hand, many shallow holes were observed on BZN8100 (Fig. 11(b)). The XPS spectra of the new CBN tools, after 5 s of ion sputtering, are shown in Fig. 12. The major difference is in the binder phase: cobalt in BZN6000 and titanium nitride in BZN8100. After machining (240 m/min), but before leaching, the ¯ank wear land of both types of CBN tools show strong iron and oxygen peaks, suggesting that the transferred layer is mainly oxidized workpiece materials. The XPS spectra of the ¯ank wear land after leaching are shown in Fig. 13, with iron oxide cleaned by leaching. The striking difference is that for BZN6000, oxygen and carbon peaks split to two and some silicon was identi®ed; the silicon peak also shifted relative to its normal position. Peak shifting was resulted from the charging effect due to non-conducting of the undissolved transferred layer. Fig. 14 shows the detail of silicon and oxygen peaks of overlapped BZN6000 and BZN8100 spectra. Shifting amounts of charging peaks are similar to silicon, carbon, and oxygen. The peaks associated with BZN8100, however, remain at normal positions. Although little silicon had been found on BZN8100 (no charging), further ion sputtering (5 s) showed the removal of the silicon peak. In contrast, the sputtering of BZN6000 resulted in an increased amount of the shifted silicon and oxygen, suggesting the silicon oxide present in the transferred layer. At a lower cutting speed, the charging condition and silicon oxide were also present on BZN6000, but not on BZN8100.

Fig. 9. Microstructures of tool wear surfaces at low cutting speed (5000): (a) BZN6000; (b) BZN8100.

280

Y.K. Chou et al. / Journal of Materials Processing Technology 124 (2002) 274±283

Fig. 10. A simple model of adhesive wear process interacted with a transferred layer.

Fig. 11. Microstructures on flank wear of leached CBN tools (5000): (a) BZN6000; (b) BZN8100.

Fig. 12. XPS spectra of new CBN tools after iron sputtering.

Y.K. Chou et al. / Journal of Materials Processing Technology 124 (2002) 274±283

281

Fig. 13. XPS spectra of leached CBN tools (240 m/min, 10.5 min cutting time).

Fig. 14. Details of overlapped spectra showing shifting peaks: (a) silicon; (b) oxygen.

3.3. Discussion The machining tests consistently showed that BZN8100 performs better, both on the part surface ®nish and tool ¯ank wear, than BZN6000 at a wide range of ®nishing conditions.

For BZN8100, the tool life of 0.2 mm VBmax was about 21 min, Ra in the range 0.2±0.4 mm. Increasing the cutting speed resulted in a greater difference in the tool performance between BZN6000 and BZN8100. This implies that BZN8100 has greater wear resistance than BZN6000 at

282

Y.K. Chou et al. / Journal of Materials Processing Technology 124 (2002) 274±283

high cutting temperatures. The linearized ¯ank wear rates seemed to be proportional to the cutting speed as well. Depth of cut in the range 50±250 mm showed only minor differences. For BZN6000, decreasing the depth of cut to 10 mm reduced the wear and wear rate, however, gave no improvement on the surface ®nish. On the other hand, for BZN8100, decreasing the depth of cut to 10 mm improved maintainable surface ®nish (<0.2 mm Ra for 21 min), though the ¯ank wear only slightly decreased. Tool wear conditions characterized by the SEM showed a crater zone, ¯ank wear land, and a distinct pattern, namely the transferred layer on the ¯ank wear land. The transferred layer on BZN6000 showed many grooves, while that on BZN8100 had ¯ake-type structures. The transferred layers on BZN6000 seem to be strongly bonded to the tool surface, and thus, result in adhesion of the cobalt phase, and subsequent plucking out of CBN particles. This may further cause severe three-body abrasion, giving abraded grooves, evident from the morphology of the worn surfaces (Fig. 8(a)) and on the leached tools (Fig. 11(a)). On the other hand, the transferred layer on BZN8100 seems to be not as strongly bonded as in BZN6000, resulting in ¯ake-type structures. However, both the worn tools and the leached tool surfaces showed shallow holes previously occupied by CBN grains. This indicates that parts of CBN grains were removed by adhesion of the transferred layer due to loss of bonding. Thus, adhesive wear interacted with the transferred layers seems to be a dominant wear mechanism, though microscale attrition being signi®cant as well. BZN6000 shows more severe wear than BZN8100 due to accelerated abrasion by plucked-out CBN grains. Different binder materials may result in different interactions with the transferred layers, and thus, distinct wear behaviors. The leaching results indicate that the transferred layer on BZN6000 remained some residue even after long-time leaching, but that on BZN8100 could be mostly cleanedoff. Interestingly, the XPS spectra identify that the residue of the transferred layer on BZN6000 had a noticeable amount of silicon oxide. The silicon oxide is unlikely an artifact from leaching. It may be caused due to material characteristics, possibly high af®nity of cobalt to the silicon compounds in the steel. This unique phenomenon needs to be further studied. 4. Summary Machining experiments showed that low CBN content tools (CBN-L) generate better surface ®nish and have lower ¯ank wear rate than high CBN counterparts (CBN-H) in ®nish turning of 62 HRC AISI 52100 steel. CBN-L, though with inferior mechanical properties, has greater wear resistance than CBN-H and the discrepancy increases with cutting speed. Depth of cut has minor effects on tool wear; however, lighter cuts lead to improvement on the surface ®nish to CBN-L, but not to CBN-H. Tool wear examined by

SEM indicated that the transferred layers on CBN-H and CBN-L have distinct morphology and different interactions with the ¯ank wear land, resulting in different wear conditions. Removal of the cobalt compound phase and the shallow holes in CBN-H and CBN-L, respectively, suggest adhesion interacted with the transferred layer as a dominant wear mechanism. The metallic binders of CBN-H have high af®nity to the transferred layer, and thus may experience more severe adhesion. Consequently, CBN particles are plucked out due to loss of binder and further cause accelerated abrasive wear. The ®ndings of this study suggest that CBN tool performance is not solely dependent upon the bulk mechanical properties; instead, interactions of microstructures between the tool and the transferred layer on the wear land can be vital to the CBN tool performance. It has been concluded that CBN-L is more appropriate for ®nish hard turning than CBN-H. Future improvement of CBN tool materials would need to optimize combined effects on mechanical properties and microstructures from CBN contents, grain sizes, and the binder phase altogether. Acknowledgements A. Donmez and R. Polvani at NIST have provided support and discussion. L. Ives, also with NIST, gave critique. GE Superabrasives supplied BZN inserts and The Timken Company supplied steel workpieces. Y. Matsumoto at Timken and D. Cerutti at GE Superabrasives made suggestions. N. Turner at NRL provided assistance with the XPS instrument. References [1] Y. Matsumoto, M.M. Barash, Review on cutting technology of hard materials, in: Proceedings of the Sixth International Conference on Production Engineering, Osaka, Japan, 1987, pp. 116±122. [2] M.A. Davies, Y.K. Chou, C.J. Evans, On chip morphology, tool wear and cutting mechanics in finish hard turning, Ann. CIRP 45 (1) (1996) 77±82. [3] T. Ekstedt, Challenge of hard turning, Carbide Tool J. 19 (1987) 21± 24. [4] W. Koenig, A. Berktold, K.F. Koch, Turning versus grindingÐa comparison of surface integrity aspects and attainable accuracies, Ann. CIRP 42 (1993) 39±43. [5] P.K. Bossom, Finish Machining of Hard Ferrous Workpieces, Ind. Diamond Rev. 50 (540) (1990) 228±232. [6] N. Narutaki, Y. Yamane, Tool wear and cutting temperature of CBN tools in machining of hardened steels, Ann. CIRP 28 (1979) 23± 28. [7] H. Eda, K. Kishi, H. Hashimoto, Wear resistance and cutting ability of a newly developed cutting tools, in: Cutting Tool Materials, Proceedings of an International Conference, American Society of Metals, Ft. Mitchell, Kentucky, September 15±17, 1980, pp. 265± 280. [8] Y. Kono, A. Hara, S. Yazu, T. Uchida, Y. Mori, Cutting performance of sintered CBN tools, in: Cutting Tool Materials, Proceedings of an International Conference, American Society of Metals, Ft. Mitchell, Kentucky, September 15±17, 1980, pp. 281±295.

Y.K. Chou et al. / Journal of Materials Processing Technology 124 (2002) 274±283 [9] S. Takastu, H. Shimoda, K. Otani, Effect of CBN content on the cutting performance of polycrystalline CBN tools, Int. J. Refractory Hard Met. 12 (1983) 175±178. [10] R.M. Hooper, C.A. Brookes, Microstructure and wear of cubic boron nitride aggregate tools, in: Proceedings of the Second International Conference of Science on Hard Material, Rhodes, 1986, pp. 907±917. [11] R.M. Hooper, J. Shakib II, C.A. Brookes, Microstructure and wear of TiC±cubic BN tools, Mater. Sci. Eng. 105/106 (1988) 429±433.

283

[12] S.A. Klimenko, Y.A. Mukovoz, V.A. Lyashko, A.N. Vashchenco, V.V. Ogorodnik, On the wear mechanism of cubic boron nitride base cutting tools, Wear 157 (1992) 1±7. [13] M. Deming, Hardened steel machining with PCBN compacts, GE Superabrasives Internal Report, Worthington, OH, 1993. [14] Y. Chou, Wear mechanisms of cubic boron nitride tools in precision turning of hardened steels, Ph.D. Thesis, Purdue University, 1994.