Davim J. Paulo (editor). Machining. Fundamentals and Recent Advances
Подождите немного. Документ загружается.
110 W. Grzesik
4.3.4 Wear of Ceramic and PCBN Tools
Figure 4.15 illustrates typical tool wear features observed in finish turning on
CBN tools with the predominant wear effect: VB
max
and VB
C
localized on the
clearance face and tool corner, respectively. In many hard-finishing turning opera-
tions with mixed ceramic tools the notch wear is also concentrated on the active
secondary cutting (trailing) edge, as shown in Figure 4.16(b). This is responsible
for the profile sharpening effect [24], causing the blunt irregular initial peaks that
are transformed into final sharp individual peaks from the side of the trailing edge
by copying the notch grooves.
The contact area between the worn PCBN tool, the workpiece and the chip is
divided into five characteristic sections, as shown in Figure 4.16(a). In the outer
zones 1 and 5 no tribological effects occur. Similarly, in zone 3, where a stable
flow layer prevents tool wear they were not observed. Probably, a protective layer
consisting of boron oxide films (details Y and X in Figure 4.13(b)) is formed at
high contact temperatures. The flank wear concentrated in zone 4 with characteris-
tic scouring areas is affected by severe abrasive wear. In zone 2 the predominant
wear mechanisms are material deformation and continuous sliding friction caused
by the chip. The bottom side of the saw-tooth chips shows a developed white layer
whose scale is almost independent of tool wear progress.
4.3.5 Modelling of Hard Cutting Processes
Recently, more numerical investigations dealing with FE modelling of fundamen-
tal characteristics (deformations, forces, temperatures) under orthogonal hard
cutting conditions have appeared [25–27], as well as much narrower considera-
tions focussed on the prediction of surface quality, residual stresses and rolling
contact for hard-machined components [28,
29]. In addition, it is proposed to
Figure 4.15. Typical wear types in CBN finish hard turning [25]
Machining of Hard Materials 111
model surface finish and subsurface residual stresses using multiple regression
models [30] and an artificial neural network (ANN) approach [31], respectively.
Figures 4.17(a,b) show the simulated results of chip formation and temperature
distribution in the cutting zone when orthogonally machining AISI H13 tool/die
steel treated to about 49 HRC with PCBN tools using the ABAQUS/Explicit
TM
FEM package. Figure 4.17(a) shows the deformed finite element mesh for a shear-
localized segmental chip which agrees well with the photomicrograh of the chip
section obtained from actual machining experiment (inserted alongside the mesh
plot). In particular, the shear angle for the segmental chip was 50°. The predicted
temperature map generated during segmental chip formation, in which the shear
zone temperature ranges between 600 and 700°C, is shown in Figure 4.17(b). In
contrast, for continuous chips this temperature was only 194−243°C [25].
It is noteworthy that FEM with the J–C equation predicts higher tool–chip in-
terface temperatures when using a tool with lower CBN content [27]. Numerical
simulations with the commercial FE code MARC have been performed [26] to
investigate the variables of tool edge radius and cutting speed and to consider
various material constitutive models for the prediction of stress and temperature
fields for precision hard cutting of the ball bearing steel 100Cr6. It was found that
a material model with isotropic work hardening, temperature and strain-rate ef-
fects predicts the maximum tool–chip temperatures close to those experimentally
measured (about 550°C and 700°C for cutting speeds of 60 and 120 m/min, re-
spectively). The 3D FE cutting models satisfactorily predict the cutting forces but
significantly underestimate the thrust force.
The simulation of bearing rolling contact with the input of process-induced
compresive residual stress profile was performed for AISI 52100 steel with 62
HRC hardness using the ABAQUS FEM package [29]. Based on the simulation of
evolutions of equivalent plastic strain for the hard-turned and ground surfaces the
Figure 4.16. (a) Typical wear pattern of a PCBN tool corresponding to saw-tooth chip
formation [11] and (b) the developed notch wear on the secondary flank face of mixed
ceramic tool [24]. Cutting conditions in case (a) were: 16MnCr5 of 60–62 HRC,
α
0
=
7°,
γ
0
=
–7°,
κ
r
=
95°,
ε
r
=
80°, r
β
=
130 μm
112 W. Grzesik
effects of the residual stress profile on the contact stress and strain were found to
be of secondary importance. On the other hand, it was shown that a negative slope
of a compressive residual stress profile reduced fatigue damage and increased
fatigue life. Moreover, rolling contact tends to reduce initial compressive residual
stress due to friction. By using special adaptive-meshing techniques in the FEM
method, which improves the stability of saw-tooth chip formation, it is possible to
predict the stresses, strains and temperatures in the subsurface generated by hard
machining more accurately [28].
Residual stresses as well as surface roughness and tool flank wear can be pre-
dicted using neural network or/and multiple linear regression models [30,
31].
Figure 4.17(c) shows the hypersurface describing the effect of feed rate and cut-
ting speed on the roughness average Ra when hard turning AISI D2 cold work
steel (60±1 HRC) using wiper ceramic inserts. It can be observed in Figure 4.17(c)
that for this case (and also for cutting force and flank wear) the optimal cutting
speed is 120 m/min. The ANN predictions of both tangential and axial compo-
nents of residual stresses induced into AISI 52100 specimens after PCBN hard
turning reveal good correlations with experimental data and FEM simulations
(with prediction errors that do not exceed 10%) [31].
Figure 4.17. Results of FEM modelling: (a) segmental chip formation, (b) coresponding
temperature distribution and (c) surface finish prediction by multiple linear regression
models [30,
32]
Machining of Hard Materials 113
4.4 Surface Integrity in Hard Machining Processes
The performance of ceramic and CBN cutting tools and the quality of the surfaces
machined are highly dependent on cutting conditions, including cutting speed,
feed rate, depth of cut and tool nose radius, which all significantly influence sur-
face roughness, white layer formation, residual stress profile and tool wear.
4.4.1 Surface Roughness
Surface roughness is the most frequently investigated characteristic of hard ma-
chining processes, mainly due to the continuous competition between hard turning
and grinding, and appropriate data can be found in numerous references, for ex-
ample, [3,
11,
33–36]. These data deal with different grades of hardened steels
(alloy, bearing, tool and cold-work steels) mostly used in the automotive and
die/mould industries. However, the 2D and 3D characterization range of surface
roughness seems to be incomplete, when considering the service properties that
are demanded [37]. Figure 4.18 presents some characteristic surface profiles and
corresponding 3D topographies for hard turning with CBN (left) and mixed ce-
ramic (right) tools at the same cutting parameters (v
c
=
100 m/min, f
=
0.1 mm/rev
and a
p
=
0.2 mm). When comparing the relevant surface profiles they contain
sharper (CBN HT) or blunter (MC HT) peaks with characteristic lateral smaller
flashes (more pronounced in MC HT) resulting from the side flow effect shown in
Figure 4.13(b).
Characteristic surface profiles generated by conventional and wiper ceramic
tools are magnified in Figure 4.19. It was observed that, due to the smoothing
effect, wiper tools produce blunt irregularities with the root-mean-square (RMS)
profile slope ranging from RΔq
=
1.5° to 5.5°. On the other hand, sharper profiles
Figure 4.18. Typical surface profiles and corresponding 3D visualizations produced in (a)
CBN and (b) MC turning operations. 3D images by permission of J. Rech
114 W. Grzesik
with RΔq values of 5–10°
were recorded [33]. As can be expected distinct differ-
ences in surface profile shapes will result in corresponding bearing area curves
(Figure 4.24(a)) and further in their contact capabilities. Figure 4.20 presents the
deterioration of surface roughness Ra during natural wear testing performed at
cutting speed of 100 m/min during 30 min. It is evident that for both fresh tools Ra
values of 0.5–0.6 µm were obtained.
Only minor changes of the Ra parameter, around 10%, were recorded during
the first 15 min of wear testing, with corresponding VB
C
nose wear of about 0.1
mm. For the lowest feed rate of 0.04 mm/rev and conventional ceramic tools the
values of the height parameters Ra, Rz and Rt were 0.25, 1.60 and 1.80 µm re-
spectively. In general, the values of the average peak spacing RSm were approxi-
mately equal to the appropriate feed rates used.
4.4.2 Residual Stresses
In order to substitute grinding process, the profile of residual stress induced in hard
turning process becomes an important factor for assuring the quality and fatigue
life of the machined components, especially roller bearings. Numerous studies
[11,
18,
38,
39] including measurements and predictions, about the process-induced
residual stresses have been conducted to determine the relationship between the
cutting variables (cutting speed, feed rate, depth of cut, tool edge preparation, tool
nose radius and tool wear) and the magnitude and the entire profile of stresses. In
general these variables depend on the thermoplastic deformation of the workpiece.
0
0.2
0.4
0.6
0.8
1
0 3.5 7 10.51417.52124.52831.5
Cutting time (min)
Parameter Ra,
μ
m
wiper - 0.2 mm/rev
conventional - 0.1mm/rev
Figure 4.20. Changes of the Ra parameter with time [33]
Figure 4.19. Characteristic shapes of the profiles generated in turning with conventional (a)
and wiper (b) ceramic tools for a constant feed rate of 0.1 mm/rev; Vertical magnification
7000×, Horizontal magnification 200×
Machining of Hard Materials 115
The changes of the distribution of residual tangential stress in the sublayer with
cutting time are shown in Figure 4.21(a). It should be noted that the tensile
stresses occur near the machined surface and their values increase progressively
with tool wear. In addition, a corresponding displacement of the point of maxi-
mum compressive stresses (about –800 MPa) deeper beneath the surface is ob-
served. The effect of the influence of tool flank wear was confirmed for Al
2
O
3
-TiC
ceramic tools and the differences in the shapes of stress distribution curves for
these two cutting tool materials are marginal.
The combination of martensite and extremely fine-grained austenite structure
constituents can cause different residual stresses in white layer. In particular, the
residual stresses of the dominant austenite material components are clearly shifted
towards compressive stresses. In contrast, the martensite tensile residual stresses
result from lower specific volume and consequently higher density of austenite
[11]. Compressive residual stresses induced by hard turning were found to im-
prove rolling contact fatigue (RCF). Hard turned surfaces may have more than
100% longer fatigue life than ground ones, with an equivalent surface finish of
0.07 µm [40] but this positive effect can be significantly reduced by the presence
of a white layer.
4.4.3 Micro/Nanohardness Distribution and White-Layer Effect
The corresponding microhardness distribution (using a Knoop indenter and 100 g
load) in the surface layer is shown in Figure 4.22(a). It should be noted that the
hardness is similar in the white layer of a few microns in thickness, whereas the
dark layer of overtempered martensite is substantially softened. The total machin-
ing thermally affected zone is about 20 µm thickness, which is about one tenth of
the cut depth (a
p
=
200 µm). Typically, the depth of the white layer increases rap-
idly with tool wear and cutting speed but for the bearing steel tested saturation is
observed at v
c
=
3 m/s [41].
Figure 4.21. (a) Modification of residual stress profile in the surface layer within tool life
period [18]; work material 16MnCr6 case-hardened steel of 62 HRC, tool material CBN;
(b) Comparison of residual stress distribution when turning with PCBN and Al
2
O
3
-TiC
ceramic tools [11]
116 W. Grzesik
Figure 4.22. Microhardness (a) and nanohardness (b) distribution in hard turned and
ground surfaces [41,
42]
Distribution of nanohardness measured by means of a diamond Berkovich in-
denter with a maximum load of 8 mN along the subsurface layers for hard turned
and ground surfaces on a 61–62 HRC heat-treated AISI 52100 steel is shown in
Figure 4.22(b). It was found that the surface nanohardness for the ground speci-
men is about 25% higher than of the turned sample (8.8 versus 7.0 GPa). It re-
mains higher to a depth of about 30 μm. The trend in nanohardness data agrees
with the measured microhardness data.
White layers consist of over 60% austenite which is a non-etching white com-
ponent, in contrast to dark martensite scoring [11]. Due to this fact, Klocke et al.
[7] suggested that the term white layer is misleading and proposed to distinguish
between two groups of this structure’s appearance in micrographs. When applying
special etching chemicals or increased etching time, white layers of the first group
remain white. On the other hand, white layers of the second group are affected by
chemicals and a fine-grained martensitic structure becomes visible. Both types of
white layer are distinguished by high brittleness and susceptibility to cracking and
are therefore classified as damage of the machined components.
Figure 4.23. Examples of microstructural changes in a hard turned surface of 52100 steel:
(a) optical micrograph, and (b) SEM micrographs of three characteristic zones [41]
Machining of Hard Materials 117
In the contact zone, the austenite temperature of the workpiece material is
reached after an extremely short time of approximately 0.1 ms and structural
changes must be expected [11]. A metallurgically unetchable structure, called the
white layer, followed by a dark etching layer, results from microstructural changes
observed in hard machining, for example, on AISI 52100 steel surfaces of 60 HRC
hardness (Figure 4.23). The three lower scanning electron microscopy (SEM)
images show the white layer, the underlying dark layer and the bulk material,
consisting of carbides distributed in a martensitic matrix. In general, carbide size
and its distribution in both the white and dark layers indicate no difference from
the bulk.
4.4.4 Modification of Surface Finish in Hybrid Processes
In general, as shown in Figure 4.24(a), the shapes of the bearing (Abbot’s) curves
obtained with conventional CBN and MC cutting tools are not satisfactory and the
surfaces machined by hard turning require further improving by finishing belt
grinding or super-finishing operations. Additionally, it is possible to generate,
using a special wiper tools with varying chamfer geometry, the flatness of the
bearing area curve (BAC) comparable to honing effect [7].
The residual stress profiles for the four finished surfaces on a 58–62 HRC car-
burized medium carbon steel are presented in Figure 4.24(b). It can be observed in
Figure 4.24(b) that, although the surface compressive stress of the HT bearing is
less than the HN bearing, its area of compressive stress is much wider, which
leads to higher relative bearing life. In contrast, the GD surface has the lowest
surface stress and unsuitable surface finish (Table 4.1), which decrease the fatigue
life of such bearing surfaces.
In order to compare the four differently machined bearing surfaces some key
2D and 3D roughness parameters are given in Table 4.1. In contrast, IF generates
the lowest average roughness Ra and GD the highest one. This trend is also ob-
served for the maximum roughness height Rt. The values of the plasticity index
Ψ
exceed unity (indicating that the asperities are deformed plastically) in all cases
Figure 4.24. Comparison of hard turned and abrasive finished surfaces. (a) Differences in
BAC curves obtained in CBN (1) and MC (4) hard turning and finish abrasive processes
(preliminary (2) and finish (3) belt grinding and super-finishing (5)) after [33]. (b)
Distributions of residual stresses [37]
118 W. Grzesik
instead of the IF process. The average slope Δq differs substantially and profiles
with blunter peaks (plateau) are generated in both HN and HT processes (Fig-
ure 4.19(b)). The values of the 3D RMS roughness height are close to the Ra val-
ues. The fastest decay autocorrelation length Sal, which identifies the direction of
low areal autocorrelation function (AACF), has a large value for turned surfaces
and a several times smaller for honed surfaces. The density of summits Sds, which
provides an estimate of the average number of asperities per a unit sampling area,
is highest for HN and GD surfaces (which implies a larger contact area and corre-
sponds well with the negative value of kurtosis Rsk) and are significantly lower
when using HT and IF operations.
4.4.5 Cutting Errors and Dimensional Accuracy
The achievable dimensional accuracy in precision hard turning using CNC ma-
chine tools with high static and dynamic stiffness, thermal stability and high preci-
sion motion control is standardised in ISO IT5 [5,
43]. As shown in Figure 4.25
many factors affect the part accuracy and process stability and reliability.
Figure 4.25. Major error drive factor and error sources in precision hard turning [40]
Table 4.1. Comparison of 2D and 3D surface roughness parameters for grinding (GD),
honing (HN), hard turning (HT) and isotropic finishing (IF) [21]
GD HN HT IF
Ra (μm)
0.51 0.24 0.29 0.05
Rt (μm)
3.81 1.72 2.37 0.32
Rsk
–0.66 –1.6 0.22 –1.5
Δq (°)
15.7 10.7 8.7 2.0
Ψ
4.7 3.2 2.7 0.6
Sq (μm)
0.52 0.26 0.32 0.06
Sal (μm)
18.2 8.8 44.8 43.8
Sds (μm
–2
)
636 2130 92 73
Machining of Hard Materials 119
In order to minimize possible errors in precision hard turning magnetic chucks,
a solid carbide tool holder and active error compensation and special monitoring
systems are proposed. Three main factors, namely tool wear, cutting forces and
cutting temperature, are the driving factors providing the error sources in process,
tooling, machine structure and clamping devices. For instance, when flank wear of
a CBN tool reaches VB
=
0.2 mm, the dimensional errors on 100Cr6 (60−62 HRC)
steel parts measured in the radial/axial directions increase to 25 µm.
Moreover, because the cutting temperature is at least 800°C, substantial thermal
expansion of both the tool shaft and the workpiece is observed. It is assumed that
thermal effects can contribute to more than 50% of the overall error of the ma-
chined parts. In particular, for the same machining conditions (v
c
=
160 m/min,
f
=
0.05 mm/rev, a
p
=
0.05 mm) thermal expansion on the CBN tool tip and work-
piece can reach up to 10 and 15 µm, respectively [40].
4.5 Applications of Hard Machining Processes
Hard machining has been found to be very efficient in many branches of industry,
including the automotive, aerospace, bearing, hydraulic and die/moulds sectors.
4.5.1 Hard Turning
A range of machining operations performed on lathes on hardened transmission
parts including longitudinal continuous and interrupted turning, facing and boring,
and grooving and threading are shown in Figure 4.26. Further improvement of the
hard turning process can be achieved by using special single-edged tools which
performed orthogonal cuts, known as hard plunge turning [34,
44].
The cutting speed for HT of 60 HRC steels with CBN inserts varies from 100
to 200 m/min. Conventional values of feed rates vary between 0.05 and
Plunge turning
Figure 4.26. Typical examples of hard turning, boring and threading operations