Wear 247 (2001) 152–160

Cutting tool wear in the machining of hardened steels

Part II: cubic boron nitride cutting tool wear
J. Barry∗ , G. Byrne
Department of Mechanical Engineering, University College Dublin, Room 324 Engineering, Belfield, Dublin 4, Ireland
Received 27 July 2000; accepted 31 August 2000

Abstract
A study was undertaken to investigate the wear mechanisms of CBN/TiC cutting tools in the finish machining of BS 817M40 (AISI
4340) steel of 52 HRC. A fourfold variation in tool wear rate was observed in the machining of three different heats of this steel. One of
the primary characteristics of the tool wear surfaces is the manner in which the TiC phase stands proud of the CBN phase. The relative
abundance of different elements on the wear surfaces of the tools, which are present in the work material in small (Mn, Si) or very small
(Al, S, O) quantities, suggests that the dominant wear mechanism of CBN/TiC cutting tools is chemical in nature. In considering the
relative wear rates of the tools used to machine the different heats of steel, a reasonable correlation is noted between the work material
Al and S content and the corresponding tool wear rate. Examination of built up layers at the trailing edge of the tool, however, suggests
that work material Al content is rate-determining with regards to tool wear. Following these observations, a new mechanism is proposed
to account for the (widely acknowledged) superior wear resistance of CBN/TiC composites in comparison to high-content CBN tools.
© 2001 Elsevier Science B.V. All rights reserved.
Keywords: Tool wear; Machining; Cubic boron nitride

1. Introduction process of significant industrial importance. This, the second

of two papers on tool wear in hard machining, is concerned
Cutting tool wear is a key issue in all metal cutting pro- with the wear mechanisms of CBN/TiC tools. The wear
cesses, primarily because of its detrimental effect on the mechanisms of alumina/TiC cutting tools is discussed in the
surface integrity of the machined component. In the ma- first part of this paper [5].
chining of hardened steels, microstructural alterations to the Commercially available CBN tool materials may be
machined surface are embodied in the ‘white layer’ [1,2], broadly classified as either high content CBN or low con-
a partially transformed nanocrystalline layer of material up tent CBN. The latter, which appear to have been initially
to several microns in thickness [3]. The primary influence developed because of difficulties involved in sintering pure
on the depth of the white layer is the width of the wear CBN [6], usually contain a ceramic binder such as TiC or
land which forms on the tool flank face, its width usually TiN. Generally, in finish hard turning of structural steels,
denoted ‘VBC ’. It follows, therefore, that minimising the ball bearing steels and case hardened steels, the low content
depth of surface white layer requires either a more frequent CBN grades are found to offer superior wear resistance in
replacement of the cutting tool or a reduction in the rate of comparison to the high content CBN grades [6–10]. This
cutting tool wear. The latter, which is the more favourable phenomenon has received several different explanations,
approach, is clearly facilitated by a thorough understanding including the degree of work material adhesion to the tool
of the mechanisms of tool wear. surfaces [7,9] and the lower dislocation density in the CBN
King and Wheildon [4] amply demonstrated the potential grains constituting the CBN/TiC composites; the latter
of alumina cutting tools in machining steels of 61 HRC, resulting in a lesser degree of ‘etching’ wear [8].
however, it was not until the development of alumina/TiC, Regarding the chemical stability of CBN, Narutaki and
cubic boron nitride (CBN) and in particular CBN/TiC and Yamane [7] noted that no reaction occurs between CBN
CBN/TiN composite tools that ‘hard machining’ became a and Iron in static high temperature diffusion tests. Konig
and Neises [11] have presented similar findings. In contrast,
∗ Corresponding author. Tel.: +353-1-706-1983; fax: +353-1-283-0534. Zimmermann et al. [12], using Auger electron spectroscopy,
E-mail address: [email protected] (J. Barry). concluded that the CBN phase of CBN/TiC tools is subject

0043-1648/01/$ – see front matter © 2001 Elsevier Science B.V. All rights reserved.
PII: S 0 0 4 3 - 1 6 4 8 ( 0 0 ) 0 0 5 2 8 - 7
 J. Barry, G. Byrne / Wear 247 (2001) 152–160 153

to diffusion wear while the softer TiC phase is subject to

abrasion. Klimenko et al. [13,14] noted the presence of re-
action products, in the form of micron-sized dendritic spher-
ical particles, on the wear surfaces of CBN composite tools
and in the atmosphere surrounding the tool tip during cut-
ting. These new compounds, reported to contain elements
such as ‘B, C, N, O, Fe, Al, Si, Cr, etc.’, were thought to
possess a lower melting temperature than that of the orig-
inal materials in contact. Their formation was attributed to
either of two mechanisms: (a) direct tool-workpiece interac-
tion due to tribo-chemical activation of their surface layers,
or (b) the formation of low melting point boric anhydride Fig. 1. The geometry of the workpiece used for facing and turning tests.
(B2 O3 ) and its subsequent interaction with the work material
to form metal oxyborides. Chou [10] observed the presence
of a SiO2 rich layer on the wear surfaces of CBN tools after radius of all tools was, rε = 1.2 mm, the clearance angle,
machining AISI 52100 steel; its formation was attributed to α = 6◦ , the average edge radius (or sharpness), rβ = 8 ␮m
a ‘tribo-chemical reaction’. Another study of interest is that and the (effective) rake angle, γ = −26◦ (cutting took place
of Tsuji et al. [15] who investigated the feasibility of using exclusively on the edge chamfer, bγ n = 200 ␮m).
monocrystalline CBN tools for the ultra-precision machin- The tool wear tests were performed using parameters rep-
ing of hardened steels. It was noted that the wear rate of resentative of those used for finish turning; cutting speed,
such tools was extremely rapid and was largely independent ν c = 2.5 and 4.17 m/s, feed, f = 100 ␮m and depth of cut,
of the crystal orientation. It was concluded that CBN was a p = 200 ␮m. The majority of tests were undertaken on
subject to a predominantly chemical wear mechanism. a Daewoo PUMA 4-3A CNC lathe under constant surface
The aim of this investigation was to assess the validity speed control and were replicated three times. The tests with
of the wear mechanisms discussed above in relation to the heat 3 of BS 817M40 steel were undertaken on a Colch-
machining of a quenched and tempered structural steel, BS ester Mastiff 1400 lathe. (The validity of a comparison of
817M40 of 52 HRC (equivalent to AISI 4340 steel). Partic- the tool wear rates on different lathes is discussed in Section
ular cognisance was given to the work material inclusionary 3.1.) The tool forces were measured with a three component
element content and the influence this may have on the wear piezoelectric platform dynamometer. Flank wear and crater
rate of CBN/TiC cutting tools. Tests were undertaken us- wear parameters were measured using a toolmakers micro-
ing finish machining conditions and the resulting tool wear scope and a stylus profilometer, respectively. The wear sur-
surfaces examined under SEM equipped with EDS. faces were examined under a scanning electron microscope
(SEM) equipped with an energy dispersive X-ray spectrom-
eter (EDS) with a resolution of 20 eV.
2. Experimental The work material inclusion content and composition is
detailed in the first part of this paper [5]. In brief, all three
Finish turning and facing tests were undertaken using heats of the BS 817M40 steels contained MnS inclusions.
three heats of BS 817M40 steel, the compositions of which In addition, heat 1 contained duplex inclusions, typically
are given in Table 1. The geometry of the workpieces used comprised of a mixed oxide ((Ca, Al)O) phase enveloped in
for both cutting arrangements is shown in Fig. 1. The work- a sulphide phase ((Mn, Fe, Ca)S). A relatively small number
pieces were austenitised at 850◦ C for 1 h, oil quenched and of such inclusions were observed in heat 3. Of the relatively
tempered for 2 h at 250◦ C, resulting in a hardness of 52 few duplex inclusions observed in heat 2, most contained
HRC. traces of aluminium and oxygen (note the relatively high Al
The CBN/TiC tools employed in the study contained 50% and O content of this steel, Table 1).
CBN; 40% TiC; 6% WC; 4% (AlN, AlB2 ) [8] and had
a grain size of 1–2 ␮m. All cutting inserts were of ISO
(1832:1991) designation; TNGA 160412 T02020. The nose 3. Results

3.1. Tool wear rates

Table 1
Composition of the three heats of BS 817M40 steel used for machining
tests, in wt.%; O and Ca contents are in ppm Table 2 lists the flank volume wear rates, (1Vα /1lc ), of
the CBN/TiC cutting tools in machining the three heats of
Heat C Si S Cr Ni Mo Al Ti O Ca
BS 817M40 steel of 52 HRC. The wear rate parameter is
1 0.40 0.163 0.034 1.176 1.371 0.283 0.029 0.003 30 25 calculated according to the relationship illustrated in Fig. 2.
2 0.38 0.193 0.021 1.104 1.350 0.254 0.035 0.003 50 3
As noted above, the tests with heat 3 were undertaken on a
3 0.44 0.204 0.012 1.163 1.653 0.189 0.004 0.006 20 6
different lathe as those tests with heats 1 and 2. As such, cau-
154 J. Barry, G. Byrne / Wear 247 (2001) 152–160

Table 2
Flank volume wear rates (1Vα /1lc , in ␮m2 /m) of CBN/TiC cutting tools
in machining different heats of BS 817M40 steel of 52 HRCa
Al2 O3 /TiC CBN/TiC
Lathe/heat 3 1 2 3
Daewoo 0.25 0.32 0.26 –
Colchester 0.27 – – 0.085
a The wear rate parameter is illustrated in Fig. 2.

tion must be exercised in making a quantitative comparison

of the tool wear rates in machining heat 3 and heats 1 and
2. The validity of a semi-quantitative comparison, however,
is confirmed by the results of wear tests undertaken with
alumina/TiC tools in which heat 3 was machined on both
lathes; near identical wear rates were observed, Table 2.
It is evident from Table 2 that there is an almost fourfold
variation in the flank (volume) wear rates of the CBN/TiC
cutting tools in machining the different heats of BS 817M40 Fig. 2. Flank and crater wear parameters.
steel. The relative tool crater wear rates, (1KT/1lc Fig. 2), in
machining the different heats was similar to the relative flank

Fig. 3. Scanning electron micrographs and EDS spectra of the wear surfaces on a CBN/TiC tool after machining heat 1 of BS 817M40 steel of 52 HRC
with cutting speed, ν c = 2.5 m/s.
 J. Barry, G. Byrne / Wear 247 (2001) 152–160 155

volume wear rates; for example in machining heats 1 and 3,

the crater wear rates were 6.7 and 1.6 ␮m/km, respectively,
a difference of 418%. The difference in the flank volume
wear rates in machining heats 1 and 3 is 376%.

3.2. Morphology of the tool wear surfaces

Examination of the wear surfaces on the tools used to

machine the different heats of BS 817M40 steel revealed
the presence of patchy built up layers (BUL), somewhat
more complex in composition than the layers observed on
alumina/TiC tools used to machine the same heats [5]. Fig. 3
shows a number of micrographs and EDS spectra of the wear
surfaces of a CBN/TiC tool used to machine heat 1. Fig. 3b
Fig. 4. Scanning electron micrograph and EDS spectrum of a layer
shows a patchy BUL in the centre of the crater. The layer is
formed at the trailing edge of a CBN/TiC tool used to machine heat 3
composed primarily of Mn and S with lesser amounts of Al, (ν c = 2.5 m/s).
Si and Ca, Fig. 3d. Al is present in the tool material as AlN
and AlB2 , though accounts for only 4% in comparison to
40% TiC. Noting the relative heights of the Al and Ti peaks layer appears bright due to charging. EDS analysis of the
in the spectra in Fig. 3d, it is clear that Al is a constituent of darker regions of the layer reveal the composition to be iden-
the layer. The size of the BUL’s observed on the tool wear tical to the lower bright region, note the EDS spectrum in
surfaces suggest that they are in fact inclusionary deposits Fig. 4. The charging effect is due solely to the thickness of
rather than what may be considered stable BUL’s, such as the layer being sufficient to prevent electrons from reaching
those observed on P-type carbide tooling after machining the underlying tool material which is conductive.
Ca-treated steels [16]. The morphology of the wear surfaces on the tools used to
Between the BUL’s in Fig. 3b, the underlying structure of machine heat 2 of BS 817M40 steel were almost identical
the crater wear surface may be seen to be almost porous-like to those on the tools used to machine heat 1, Fig. 5a. The
in appearance. This is particularly evident in the micrograph spectra from the crater wear surfaces reveal the BUL’s to
in Fig. 3c, in which a different region of the crater surface contain the same elements as were noted on the tools used
is viewed at a small oblique angle. The particles which pro- to machine heat 1, with the notable exception of Ca, Fig. 5b.
trude or stand proud are TiC grains, Fig. 3e. The morphology (The Ca content of heats 1 and 2 is 25 and 6 ppm, respec-
of the flank wear surface is similar to that of the crater wear
surface, Fig. 3f. The accompanying EDS spectrum, Fig. 3g,
reveals the presence of significant amounts of Fe, Al, Si,
S and Mn. (The EDS spectrum of the field in Fig. 3c was
noted to be qualitatively similar to that in Fig. 3e.) Thus,
the primary difference between the layers in Fig. 3b and the
general chemistry of the wear surfaces is the absence of sub-
stantial amounts of Fe in the former. It should be noted that
the layers may also contain B, C and N. B and C could not
be detected due to the absorption of low-energy X-rays by
the window of the EDS detector. Only when N was present
in substantial amounts could it reliably be detected. Also,
it is noted that due to the absorption of low energy X-ray
quanta, the relative height of the O peaks in the EDS spectra
is not representative of the relative O content of the speci-
men. It has been estimated that approximately 92% of the
X-ray quanta from O are absorbed by the detector window
(used in this study), i.e. the apparent O content of the speci-
men/surface is approximately 1/13th of the actual O content.
Examination of the trailing edge of the tool in Fig. 3a (i.e.
the left side of the micrograph) reveals a BUL of consider-
able size. Such layers were observed at the trailing edges of
all CBN tools, irrespective of which heat of steel was tested. Fig. 5. Scanning electron micrograph and EDS spectrum of the crater
Fig. 4 shows a layer formed over the trailing edge of a tool wear surface of a CBN/TiC tool after turning heat 2 of BS 817M40 steel
(different to that shown in Fig. 3a). The lower region of the of 52 HRC with cutting speed, ν c = 2.5 m/s.
156 J. Barry, G. Byrne / Wear 247 (2001) 152–160

Adhered fragments of oxidised work material are clearly

evident, (note the high O content), as were observed on the
alumina/TiC tools used to machine heat 3 [5]. In addition to
Fe and O, these surfaces contain Al and Si, Fig. 6d. In con-
trast to the wear surfaces on the tools used to machine heats
1 and 2, significant levels of Mn and S were not present.
Removal of the adhered fragments of oxidised work mate-
rial by leaching with sulphuric acid for 2 h, revealed the un-
derlying wear surface, Fig. 6b. These (leached) surfaces also
contained traces of Al, Si and O, indicating the presence of
compounds which resisted attack by the acid.
Another notable difference between the tools used to ma-
chine heat 3 and those used to machine heats 1 and 2 was
the morphology of the flank wear surfaces, Fig. 6c. In con-
trast to the almost porous-like structure of the flanks on the
tools used to machine heats 1 and 2, the constituent phases
were fairly evenly worn on the tools used to machine heat 3,
however, the (brighter) TiC phase is thought to stand slightly
proud of the (darker) BN phase. Adhered fragments of steel
or built up layers were not noted on the flank after machin-
ing heat 3. EDS analysis of the tool flanks revealed only
traces of Fe, Al, Mn and Si, Fig. 6e. The spectrum in Fig. 6e
is shown on a log ordinate because of the small concentra-
tion of some elements present. Tungsten (W) is present in
the tool material, reported as 6% WC [8]. While the Si-kα .
peak overlaps with the W-mα peak, the presence of Si is
confirmed by the height of the Si-kα peak, which is greater
than that of the W-lα and W-lβ peaks. If only Tungsten were
present, the m peaks would be expected to be lower than the
l (ell) peaks.

4. Discussion

Previous studies on the wear of CBN cutting tools have

generally focused on either predominantly mechanical or
predominantly chemical wear mechanisms. Of the latter, dif-
fusion of BN in the steel chip [12] and the formation of low
melting point reaction products [13,14] have been suggested.
Among the mechanical or thermo-mechanical wear mech-
anisms discussed in the literature are micro-fracture due to
adhesion between the chip and the tool [7], etching wear [8],
soft abrasion [17] and self-attrition by spalled grains [14].
While the operation of these mechanisms can not be ruled
out, the relative abundance of different elements on the wear
surfaces of the tools used to machine heats 1 and 2, which
Fig. 6. Scanning electron micrographs and EDS spectra of the wear are present in the work material in small (Mn, Si) or very
surfaces of a CBN/TiC cutting tool after turning heat 1 of BS 817M40 small (Al, S, O) quantities, suggest that the dominant wear
steel of 52 HRC with cutting speed, ν c = 2.5 m/s.
mechanism is chemical in nature.
The large number of elements present on the tool wear
surfaces, which as previously noted may also include B, C
tively.) The morphology and chemistry of the flank wear and N, make it virtually impossible to determine the phases
surfaces were similar to those of the crater wear surface. present and in particular, the reactions through which they
The crater wear surfaces on the CBN/TiC tools used to were produced. It is clear however, that many of the elements
machine heat 3 were somewhat similar to those on the identified on the wear surfaces are work material inclusion-
CBN/TiC tools used to machine heats 1 and 2, Fig. 6a. ary elements, i.e. Al, Mn, S, Ca and O. While Si is also
 J. Barry, G. Byrne / Wear 247 (2001) 152–160 157

Table 3
CBN/TiC tool wear rates and work material inclusionary element content for three heats of BS 817M40 steel
Heat Crater wear rate 1KT/1lc Flank wear rate, 1Vα /1lc , (␮m2 /m) Work material composition — inclusion elements (wt.%)

Al S Ca O Mn Si

1 6.7 0.32 0.029 0.034 0.0025 0.003 0.608 0.163

2 5.4 0.26 0.035 0.021 0.003a 0.005 0.538 0.193
3 1.6 0.085 0.004 0.012a 0.0006a 0.002 0.637a 0.204
a This denote elements which were either present in very small quantities, or not detected, on the tool wear surfaces.

frequently observed in inclusions in steels [18], it was not

noted in the inclusions in the steels tested in this study. Ta-
ble 3 lists the inclusionary element composition of the three
heats of steel and the corresponding flank and crater wear
rates. Those elements marked with footnote were present
only in very low quantities, or not at all, on the wear sur-
faces of the tools used to machine the respective heats.
Of the six elements listed, a reasonable correlation appears
to exist between a high work material Al and S content and
a relatively rapid tool wear rate. The variation in the Mn
and Si content of the three heats is probably insignificant
(as is the Cr, Ni and Mo content). Rather, the amount of
Mn present in the work material inclusions is determined
largely by the S content. However, the presence of Si on
the wear surfaces suggests that elements in solution (in the
work material) may also react with the tool and need not
be present in the inclusion compounds. The role of Ca in
the assumed chemical reaction is probably also relatively
insignificant; heat 2 contains only 3 ppm Ca, compared to 6
ppm in heat 3, yet results in a tool wear rate three times that
of heat 3.
Whether or not S and Al play an important role in the (as-
sumed) chemical reaction is difficult to ascertain solely from
analysis of the wear surface composition, as it is possible
that such elements may be merely deposited on the surface
as part of inclusion compounds. In this regard, the built up
layers observed at the trailing edge of the tool are potentially
insightful as they are thought to be composed of the reaction
compounds formed on the wear surfaces, i.e. the products of
the chemical wear mechanism. In contrast to the chemical
composition of the wear surfaces, which may contain un-
reacted or partially reacted tool material compounds, work
material and adhered work material inclusions, the BUL’s
at the trailing edge of the tool might provide a permanent
record of the reaction products. This suggestion is based on
the fact that work material inclusions cannot be deposited
outside the tool/workpiece contact zone.
Furthermore, that BUL’s were observed only at the trail-
ing edge of the tools and not at the leading edge, is thought to
be significant. Considering the geometry of the finish turn-
ing process, Fig. 7a–c, it is evident that the leading cutting
edge angle, κ e eff , is 33.5◦ , whereas, the trailing cutting edge
angle, κe0 eff , is only 4.8◦ . Also shown in Fig. 7d and f are
the trailing and leading edges, respectively, of a CBN/TiC Fig. 7. The formation of a built-up-layer at the trailing edge of the tool
tool used to machine heat 3 of BS 817M40 steel. A BUL due to the geometric constraint there.
158 J. Barry, G. Byrne / Wear 247 (2001) 152–160

is present at the trailing edge. A BUL is not evident at the In essence, it appears that the TiC phase is more wear re-
leading edge. It is thought that at the leading edge, the larger sistant than the BN phase, seemingly being left to bear the
angle allows expulsion of the reaction products into the at- mechanical load on the tool surface. Considering the BN
mosphere whereas at the trailing edge, the extruded reaction phase is subject to chemical wear, the lower hardness of the
products build up due to the geometrical constraint there. It TiC phase is of little importance in determining the rela-
is interesting to compare this to the findings of Klimenko tive wear rates of each phase. Such behaviour suggests that
et al. [13,14] who sampled the atmosphere in the vicinity a higher TiC might result in a lower composite wear rate,
of a CBN tool during the machining of a C–Cr ball bearing however, the results of Takatsu et al. [9] clearly indicate that
steel and noted the emission of liquid phase particles. It is the optimum TiC content of CBN/TiC composites for ma-
useful to note that this phenomena, i.e. BUL formation at the chining steels of different compositions and hardness levels
trailing edge, is not related to a temperature variation along is approximately 45%.
the engagement length of the tool. This has been verified by It is also useful to note the many studies which have shown
conducting tests at a higher cutting speed, ν c = 4.17 m/s, that the wear rate of high-content CBN (>90%) and MCBN
with the result that BUL’s are not formed at the leading [15] is significantly greater than the wear rate of low-content
edge of the tool. A higher cutting speed usually results in CBN (∼50%) in finish turning of hardened steels [6–10,21],
higher temperatures at the tool-workpiece and tool-chip in- with the notable exception of machining HSS [6,7]. As noted
terface [19]. Also, it should be noted that BUL formation at above, this behaviour has received a number of explana-
the trailing edge is thought not to be a result of the presence tions, including the superior toughness of low-content CBN
of atmospheric oxygen. If this was the case, layer forma- and a reduced degree of adhesion between the chip and the
tion should also occur at the leading edge. The expulsion of tool [7], a greater degree of welding of the work material on
material perpendicular to the direction of primary motion is low-content CBN, thought to protect the tool [8,9], a lower
commonly encountered in machining, namely, (work) ma- dislocation density in low-content CBN leading to less ‘etch-
terial side-flow [20]. ing wear’ [8], a greater degree of thermal softening in the
If, as suggested, it is valid to consider the BUL’s at the chip formation zone due to the lower thermal conductivity of
trailing edge of the tools as products of a reaction between low-content CBN grades [21] and according to Chou [10], a
the work material, the work material inclusions and the tool reduced degree of adhesive wear associated with BUL for-
material, rather than inclusionary deposits which may not mation. Apart from a degree of inconsistency among these
necessarily react with the tool material, then the elements theories, none suggest that the improved wear resistance of
present in the layer are those which are of importance in the low-content CBN tools is due to the greater wear resistance
reaction. From Fig. 4, these elements are Mn, Si, Al and O of the TiC phase compared to the BN phase. In addition
with smaller amounts of Ca, Cr and S. The composition of to the evidence presented above that this is the case, Zim-
the trailing edge BUL’s on the tools used to machine heats mermann et al. [12] also noted a difference in the wear rate
2 and 3 were very similar, composed primarily of Mn, Si, of the TiC and CBN phases, though attributed this to dis-
Al and O. (Note the above discussion on the absorption of solution/diffusion of BN in the chip. While the findings of
oxygen X-ray quanta and that the apparent oxygen content Klimenko et al. [13,14] suggest that CBN is subject to a
is approximately 1/13th the actual oxygen content.) The rel- predominantly chemical wear mechanism, these authors did
atively small amount of S in the layer suggests that it is not not offer a mechanism which might account for the superior
as important in the reaction as is Al. wear resistance of low-content CBN grades.
Based on the above, it appears that the chemical reaction Considering the findings above which suggest the op-
thought responsible for the preferential wear of the CBN eration of a chemical wear mechanism in which new
phase of the tool material, primarily involves the elements, compounds are formed, another phenomenon may be pro-
or compounds of, Mn, Al, Si and O. The crater wear surfaces posed, which might account for the superior wear resistance
of the tools used to machine heats 1, 2 and 3 of BS 817M40 of CBN/TiC composites compared to high-content CBN
steel, were as noted previously, almost porous-like in mor- grades. Namely, protection of the TiC phase from disso-
phology — the TiC phase standing proud of the BN phase. lution/diffusion wear by the compounds formed between
The flanks of the tools used to machine heats 1 and 2 were the BN phase and the work material/work material inclu-
of a similar morphology. Only on the tools used to machine sions. This phenomenon may be likened to the formation of
heat 3, which showed the lowest wear rate, were the BN protective oxide layers on P-type (mixed) carbide tools in
and TiC phase fairly evenly worn, Fig. 6c. The Al content machining Ca-treated steels [16,22], however, the protect-
of heat 3 is the lowest of the three heats tested (0.004 wt.% ing compounds, in this case, it is suggested, are a product
compared to 0.029 and 0.035 in heat 1 and heat 2, respec- of the reaction involving the BN phase and work material
tively), while the relative differences in the Si and Mn con- elements rather than solely deposits from the work material.
tents are considerably less. An additional feature of such a It is possible, however, that the mixed oxide inclusions in
wear mechanism is likely to be spallation of TiC grains due heat 1 also contribute to the protection of the TiC phase
to preferential wear of the surrounding BN phase, thus re- in accordance with the mechanisms outlined by Helle [22]
sulting in subsequent self-attrition. and Yamane et al. [16].
 J. Barry, G. Byrne / Wear 247 (2001) 152–160 159

In this regard, it is interesting to note the findings of Ya- BUL’ s at the trailing edge of the tool are thought to provide
mane and Narutaki [23] who investigated the feasibility of a permanent record of the reaction products.
using hexagonal BN as a machinability-enhancing additive Following the observation that the CBN phase of
in (soft) steels. Improvements in tool life were noted, though CBN/TiC tools is subject to a greater wear rate than the
only when machining with P-type carbide tools (which con- TiC phase, a new mechanism is offered to account for the
tain TiC). In addition, the tool wear surfaces were charac- (widely acknowledged) superior wear resistance of low
terised by the presence of AlN, a compound normally found content CBN grades, (e.g. CBN/TiC), in comparison to the
only in nitrided steels and which was presumed to have high-content CBN grades. It is suggested that the products
formed during machining. The formation of Al-containing of the reaction between the BN phase and certain work
compounds is in agreement with the findings above and with material inclusionary deposits, may afford a degree of pro-
the calculations of Klimenko et al. [14] which show the re- tection to the TiC phase against dissolution/diffusion wear.
action between Al and CBN to form either AlN and AlB2 to Such a mechanism may be similar to that which accounts for
be thermodynamically favourable. However, considering the the superior machinability of hexagonal BN treated steels
presence of Mn, Si and O in the trailing edge BUL, the reac- machined with P-type carbide tooling, as recently discussed
tion is likely to be more complex than one between Al and by Yamane and Narutaki [23]. It is suggested that the chem-
BN and may also involve a transformation of the BN phase, ical wear of CBN and CBN/TiC composites may be further
from cubic to hexagonal. Considering the relative heights of elucidated by investigating the chemical reactions between
the peaks in Fig. 4, it is possible that the chemical wear of these materials and spessartite (3MnO·Al2 O3 ·3SiO2 ), a
CBN and CBN/TiC composites may be further elucidated compound which may be present in the trailing edge BUL’s.
by considering reactions between these materials and spes-
sartite (3MnO·Al2 O3 ·3SiO2 ); a compound composed of 3,
2, 3 and 12 parts Mn, Al, Si and O, respectively. Acknowledgements

This work was funded by Enterprise Ireland, the Irish

5. Conclusions State-Agency for science and technology, under their basic
research grants scheme. Also acknowledged is the assistance
A fourfold variation in the rate of CBN/TiC cutting tool of the (Irish) National Heat Treatment Centre, Department
wear has been observed in the finish machining of three of Mechanical Engineering, UCD, the Electron Microscopy
different heats of BS 817M40 steel of 52 HRC. Examination Laboratory, UCD and Materials Ireland, UCD. De Beers
of the crater wear surfaces revealed the TiC phase to protrude Industrial Diamond Division donated the cutting tools.
above the CBN phase. On those tools which suffered the
most rapid wear rates, patchy built up layers (BUL) were
observed, composed of Al, Si, Mn, S, Fe and Cr (and for References
a high-Ca heat, Ca). The flank wear surfaces of these tools
were of a similar morphology and chemistry to the crater [1] Y.K. Chou, C.J. Evans, White layers and thermal modelling of hard
wear surfaces. turned surfaces, Int. J. Mach. Tools Mf. 39 (1999) 1863–1881.
[2] H.K. Tönshoff, C.R. Arendt, R. Ben Amor, Cutting hardened steels,
For the heat which resulted in the lowest tool wear rate,
Ann. CIRP, 49 (2) (2000), in press.
fragments of oxidised work material were adhered to the [3] J. Barry, Machining hardened steels cutting tool wear, acoustic
crater wear surfaces. Removal of such with H2 SO4 revealed emission, chip formation and surface integrity, Ph.D. Thesis,
the presence of Al, Si, Mn, S and Fe. The flank wear lands University College Dublin, 2000.
of these tools appeared relatively evenly worn, in contrast to [4] A.G. King, W.M. Wheildon, Ceramics in Machining Processes,
the flanks of the tools which suffered the higher wear rates. Academic Press, New York, 1966.
[5] J. Barry, G. Byrne, Cutting tool wear in the machining of hardened
The relative abundance of different elements on the wear steels. Part I. Alumina/TiC cutting tool wear, Wear 247 (2000)
surfaces of the tools, which are present in the work mate- 139–151.
rial in small (Mn, Si) or very small (Al, S, O) quantities, [6] H. Eda, K. Kishi, H. Hashimoto, The newly developed CBN cutting
suggests that the dominant wear mechanism of CBN/TiC tool, in: Proceedings of the 21st MATADOR Conference, Swansea,
cutting tools is chemical in nature. In considering the rel- Wales, 1980, pp. 253–258.
[7] N. Narutaki, Y. Yamane, Tool wear and cutting temperature of CBN
ative wear rates of the tools used to machine the different
tools in machining of hardened steels, Ann. CIRP 28 (1) (1979)
heats of steel, a reasonable correlation is noted between the 23–28.
work material Al and S content and the corresponding tool [8] R.M. Hooper, J.I. Shakib, A. Parry, C.A. Brookes, Mechanical
wear rate. Examination of the BUL at the trailing edge of properties microstructure and wear of DBC50, Ind. Diamond Rev.
the tools, however, suggests that work material Al content is 4 (89) (1989) 170–173.
[9] S. Takatsu, H. Shimoda, K. Otani, Effects of CBN content on the
rate-determining with regards to tool wear. In contrast to the
cutting performance of polycrystalline CBN tools, Int. J. Refract.
chemical composition of the wear surfaces, which may con- Hard Met. 12 (1983) 175–178.
tain unreacted or partially reacted tool material compounds, [10] Y.S. Chou, Wear mechanisms of cubic boron nitride tools in precision
work material and adhered work material inclusions, the turning of hardened steels, Ph.D. Thesis, Purdue University, 1994.
160 J. Barry, G. Byrne / Wear 247 (2001) 152–160

[11] W. König, A. Neises, Wear mechanisms of ultra-hard non-metallic [17] Y.K. Chou, C.J. Evans, Tool wear mechanisms in continuous cutting
cutting tools, Wear 162–164 (1993) 12–21. of hardened tool steels, Wear 212 (1997) 59–65.
[12] M. Zimmermann, M. Lahres, D.V. Viens, B.L. Laube, Investigations [18] F.B. Pickering (Ed.), Inclusion Shape Control. Inclusions. Institute
of the wear of cubic boron nitride cutting tools using auger electron of Metallurgists, 1979, pp. 108–126.
spectroscopy and X-ray analysis by EPMA, Wear 209 (1997) 241– [19] T. Ueda, M. Al Huda, K. Yamada, K. Nakayama, Temperature
246. measurement of CBN tool in turning high hardness steel, Ann. CIRP
[13] S.A. Klimenko, Y.A. Mukovoz, V.A. Lyashko, V.V. Ogorodnik, On 48 (1) (1999) 63–66.
the wear mechanism of cubic boron nitride base cutting tools, Wear [20] T.I. El-Wardany, E. Mohamed, M.A. Elbestawi, Material side flow in
157 (1992) 1–7. finish turning of hardened steel with ceramic tool. Contact problems
[14] S.A. Klimenko, Y.A. Mukovoz, L.G. Polonsky, Cutting tools and surface interaction in manufacturing and tribological systems,
of superhard materials, in: X.S. Li, I.M. Low (Eds.), ASME PED-Vol. 67 (ASME TRIB-Vol. 4) (1993) 159–170.
Advanced Ceramic Tools for Machining Application — II. Key [21] W. König, M. Iding, R. Link, Fine turning and drilling hardened
Engineering Materials, Trans. Tech. Publ., Switzerland 114 (1996) steels, Ind. Diamond Rev. 2 (90) (1990) 79–85.
1–63. [22] A. Helle, On the interaction between inclusions in steel and the
[15] K. Tsuji, J. Degawa, S. Yazu, The cutting performance of cbn single cutting tool during machining, Ph.D. Thesis, Helsinki university of
crystals in precision cutting of hardened steels, in: S. Saito, O. technology, Acta Polytech. Scand., Chem. Technol. Ser. No. 228
Fukunaga, M. Yoshikawa (Eds.), Science and Technology of New (1995).
Diamond, 1990, pp. 421–424. [23] Y. Yamane, N. Narutaki, New suppressive mechanisms of tool
[16] Y. Yamane, H. Usuki, B. Yan, N. Narutaki, The formation of a wear by boron nitride, international conference on the behaviour of
protective oxide layer in machining resulphurized free-cutting steels materials in machining: opportunities and prospects for improved
and cast irons, Wear 139 (1990) 195–208. operations, Stratford on Avon, UK, 1998, pp. 23–32.