vibration-based tool condition monitoring systems 2005 scheffer
DESCRIPTION
Despite the high level of technology built into every aspect of modern metal cutting operations, the phenomenon oftool wear still hampers the reliability and complete automation of machining processes. Tool wear is the loss ofmaterial on the edge of the cutting tool. This chapter concerns sensor-based tool condition monitoring (TCM), andspecifically those methods that are based on vibration related properties such as force, acceleration, and acousticemission (AE). References are made to systems proposed in the literature and also to commercially availablehardware. The chapter focuses on turning operations. The mechanics of turning are briefly discussed. Variousmethods of obtaining vibration signals from turning operations are described. The vibration signal has to beprocessed in order to estimate the level of wear in the cutting edge of the tool, and several state-of-the-art approachesare discussed. Effective methods of constructing a model relating sensor data and the tool wear, using processedvibration signals, are described. The chapter concludes by indicating some important points that should beconsidered when using vibration-based systems for TCM, and some interesting topics for future research in this fieldof study. Chapters 25 and Chapter 27 present further information of the present subject.TRANSCRIPT
26Vibration-Based Tool
Condition MonitoringSystems
C. SchefferUniversity of Pretoria
P.S. HeynsUniversity of Pretoria
26.1 Introduction .................................................................... 26-1149
26.2 Mechanics of Turning .................................................... 26-1150General Terms † Chatter Vibrations † Tool Wear
26.3 Vibration Signal Recording ........................................... 26-1155Direct and Indirect Systems † Sensor Requirements
for Tool Wear Monitoring † Force Measurement †
Acceleration Measurement † Acoustic Emission
Measurement † Sensor Comparisons
26.4 Signal Processing for Sensor-Based ToolCondition Monitoring ................................................... 26-1159Feature Extraction † Feature Selection
26.5 Wear Model/Decision-Making for Sensor-BasedTool Condition Monitoring .......................................... 26-1163Trending, Threshold † Neural Networks † Fuzzy
Logic † Other Methods
26.6 Conclusion ...................................................................... 26-1168
Summary
Despite the high level of technology built into every aspect of modern metal cutting operations, the phenomenon oftool wear still hampers the reliability and complete automation of machining processes. Tool wear is the loss ofmaterial on the edge of the cutting tool. This chapter concerns sensor-based tool condition monitoring (TCM), andspecifically those methods that are based on vibration related properties such as force, acceleration, and acousticemission (AE). References are made to systems proposed in the literature and also to commercially availablehardware. The chapter focuses on turning operations. The mechanics of turning are briefly discussed. Variousmethods of obtaining vibration signals from turning operations are described. The vibration signal has to beprocessed in order to estimate the level of wear in the cutting edge of the tool, and several state-of-the-art approachesare discussed. Effective methods of constructing a model relating sensor data and the tool wear, using processedvibration signals, are described. The chapter concludes by indicating some important points that should beconsidered when using vibration-based systems for TCM, and some interesting topics for future research in this fieldof study. Chapters 25 and Chapter 27 present further information of the present subject.
26.1 Introduction
Millions of products are manufactured daily by a variety of processes. A basic method to form bulk metal
into a desired final shape is through the process of metal cutting, also referred to as machining. Metal
cutting is essentially the removal of excess material from a workpiece by moving a working tool over the
0-8493-1580-8/05/$0.00+$.50q 2005 by CRC Press LLC 26-1149
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
surface of the workpiece at a certain depth, speed, and feed rate. Conventional machining operations are
turning, milling, and drilling.
Despite the high level of technology built into every aspect of modern metal cutting operations, the
phenomenon of tool wear still hampers the reliability and complete automation of machining processes.
Tool wear is the loss of material on the edge of the cutting tool. Although tool wear can be minimized, it
cannot be eliminated. Unfortunately, excessive or even a small quantity of tool wear may cause a defect in
a machined component, and therefore it is always necessary to be aware of the extent of the current tool
wear before machining can commence. Economic losses due to tool wear occur as a result of the
scrapping of expensive parts and the nonoptimal use of tool inserts. A conservative approach is often
taken, and the insert is recycled long before it should have been. Furthermore, secondary damage due to
tool wear can be extreme and even catastrophic. For this reason, many approaches to tool condition
monitoring (TCM) have been proposed through the years. There exist sensorless and sensor-based TCM
approaches. Sensorless approaches are generally tool-life equations and not monitoring methods. Thus,
sensorless approaches attempt to determine the optimal tool life under certain machining conditions.
These are often extended versions of the famous Taylor equation, which is described by
vTn ¼ C ð26:1Þ
where v is the cutting speed, T is the tool life, and n and C are constants that must be determined
experimentally for a given tool and workpiece combination.
This chapter is focused on sensor-based TCM, and specifically those methods that are based on
vibration related properties such as force, acceleration, and acoustic emission (AE). These sensor types
are known to be most effective for TCM. Furthermore, discussions will be focused on the application of
TCM in turning operations, though reference will be made to other machining operations as well. Besides
vibration-based approaches, other sensor based TCM methods are:
* Use of noncontact capacitive sensors* Vision systems* Measurement of the motor current* Surface roughness monitoring* Ultrasonic monitoring* Temperature monitoring* Laser scatter methods* Audible emission monitoring
The reader is also referred to other excellent overviews of sensor-assisted TCM, published by Dan and
Mathew (1990), Byrne et al. (1995), Scheffer and Heyns (2001a), and Dimla (2001). A TCM database wasQ1
also published by the CIRP, supervised by Teti (1995), which includes more than 500 research papers
focusing only on TCM.
26.2 Mechanics of Turning
26.2.1 General Terms
A typical turning operation is schematically shown in Figure 26.1. The cutting tool moves parallel to the
workpiece and spindle, and hence reduces the diameter of the shaft. The most important machining
parameters are:
* Cutting speed (usually expressed in m/min)* Feed rate (usually expressed in mm/rev)* Depth of cut (usually expressed in mm)
The force response on the tool tip due to the turning operation consists of three components: Fx, Fy,
and Fz. These forces consist of a static and a dynamic part, as shown in Figure 26.2. The static forces are
governed by the static pressure between the tool and workpiece, and are a function of the machining
Vibration and Shock Handbook26-1150
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
parameters. The dynamic forces are governed by forced and free vibrations due to excitation from the
cutting operation. Analytical models exist that can describe the static forces for basic machining
operations (Merchant, 1945). The dynamic behavior is more difficult to model theoretically, although
there is also continuous research in this area (Kapoor et al., 1998).
One of the main difficulties of monitoring tool wear with vibration is to identify the frequency range
that is influenced by tool wear, since machining processes entail various mechanisms that produce
vibrations that are not related to tool wear. The frequency range of vibrations produced during ordinary
machining operations usually falls between 0 and 10 kHz. From the literature, it can be concluded that
the frequency range sensitive to tool wear depends entirely on the type of machining operation, and must
be determined experimentally for each individual case. There are two important vibration frequencies
present during cutting:
* The natural frequencies of the tool holder and its components* The frequency of chip formation
Dynamic tests should be conducted to identify the dynamic properties of tool holders (Scheffer and
Heyns, 2002a). However, the interaction of the
working tool engaged into the rotating workpiece
complicates the situation, and as a result the
dynamic behavior during cutting could be differ-
ent from the expected behavior obtained from off-
line tests. Scheffer and Heyns (2004) compared
continuous cantilever models with modal hammer
tests for different tool holder overhang lengths.
The natural frequency of the first mode as a func-
tion of overhang length is plotted in Figure 26.3
(for a specific tool holder). It can be seen that a
continuous fixed-free cantilever beam model
corresponds well with the results obtained with
hammer tests.
FIGURE 26.1 Turning operation.
dynamic cuttingforce
0times [S]
forc
e [N
]
static cutting force
FIGURE 26.2 Static and dynamic forces.
Vibration-Based Tool Condition Monitoring Systems 26-1151
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
The chip formation frequency can be calculated
with simple equations that take the machining
conditions into account (Lee et al., 1989). The tool
holder natural frequencies and chip formation
frequency are independent. Generally, tool wear
has a larger effect on the vibration amplitudes at
the tool holder natural frequencies but can
influence chip formation frequencies as well.
26.2.2 Chatter Vibrations
Another phenomenon important to machining
operations is tool chatter vibrations. These are
self-excited vibrations resulting from the gener-
ation of different chip thicknesses during machin-
ing. Initially, cutting forces excite a structural
mode of the machine–workpiece system. This
leaves a wavy surface finish on the workpiece.
During the next revolution, another wavy surface
is produced in the same way. Depending on the
phase shift between these two waves, the maximum chip thickness can grow and oscillate at a particular
frequency that is close to that of a structural mode. This is called the regenerative chatter frequency.
Chatter cause a poor surface finish and can also lead to tool breakage.
The analysis and prediction of chatter has been the subject of research for many years. Morimoto et al.
(2000) developed a piezoelectric shaker/actuator to regenerate the vibrations of the cutting process. In
this way, unwanted vibrations such as chatter can be attenuated. The system is also helpful to determine
the dynamic properties of the machine tool. Koizumi et al. (2000) used a very interesting approach called
the correlation integral in the time domain to identify chatter onset. Lago et al. (2002) designed a sensor
and actuator integrated tool for turning and boring to control chatter. The tool holder shank vibrations
are sent to the actuator via a digital controller. An adaptive feedback control system is used to perform
broadband vibration attenuation up to 40 dB at different frequencies simultaneously.
26.2.3 Tool Wear
26.2.3.1 Tool Failure Mechanisms
Tool wear is caused by mechanical loads, thermal loads, chemical reactions, and abrasive loads. The load
conditions are in turn influenced by the cutting conditions and materials. The different loads can cause
certain wear mechanisms that may occur in combination. These mechanisms have either a physical or
chemical characteristic that causes loss or deformation of tool material. Tool wear mechanisms can be
classified into several types, summarized as follows (Du, 1999):
* Abrasive wear resulting from hard particles cutting action* Adhesive wear associated with shear plane deformation* Diffusion wear occurring at high temperatures* Fracture wear due to fatigue
Other wear mechanisms are plastic deformation and oxidation, which are not very common in
industry. It is estimated that 50% of all tool wear is caused by abrasion, 20% by adhesion, and the other
10% by the other mechanisms (Kopac, 1998). Abrasion is basically the grinding of the cutting toolQ2
material. The volume of abrasive wear increases linearly with the cutting forces. Higher hardness of the
tool material can reduce the amount of abrasive wear. During adhesion, the high pressures and
temperatures on the roughness peaks on the tool and the workpiece cause welding. These welding points
7
6
5
4
3
2
1
040 45 50 55 60
1st m
ode
freq
uenc
y [k
Hz]
overhang length [mm]
fixed-free cantilever beamhammer tests
FIGURE 26.3 Frequency of first tool holder mode.
Source: Scheffer, C. and Heyns, P.S., Mech. Syst. Signal
Process, Elsevier, 2004. With permission.
Vibration and Shock Handbook26-1152
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
are broken many times every second due to the workpiece movement and as a result cause removal of the
tool material (Kopac, 1998). Diffusion wear occurs at even higher cutting speeds, where very high
temperatures are present (especially when using hard metal tools).
26.2.3.2 Tool Failure Modes
Tool wear will generally occur as a combination of a number of wear modes, with one mode predominant.
The dominant mode will depend on the dominant wear mechanism, which in its turn is influenced by the
machining conditions and the choice of tool and workpiece material. For a given tool and workpiece
combination, the dominant wear mode can be determined at different cutting speeds using the product of
the cutting speed and the undeformed chip thickness (Dimla, 2000). The common wear modes are:
* Nose wear* Flank wear* Crater wear* Notch wear* Chipping* Cracking* Breakage* Plastic deformation
Figure 26.4 is a graphical representation of the different tool failure modes. The consequences of tool
wear are deviations in shape and roughness of the machined part that cause the part to be discarded because
it is out of tolerance. Most wear modes cause an increase in cutting forces, although this is not always the
case for all tool and workpiece combinations. The most widely researched tool failure modes for turning
with single point tools are flank wear, breakage (fracture), and crater wear. Flank and crater wear are
accepted as normal tool failure modes, because the other failure modes can be avoided by selecting the
proper machining parameters. The growth of flank and crater wear is directly related to the total cutting
time, unlike some of the other failure modes, which can occur unexpectedly even with a new tool.
FIGURE 26.4 Tool failure modes. Source: Scheffer, C. and Heyns, P.S. 2001. COMA 01, University of Stellenbosch.
Q7
With permission.
Vibration-Based Tool Condition Monitoring Systems 26-1153
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
26.2.3.3 Tool Wear Measurement
Wear measurements of tool inserts are done
through the implementation of an appropriate
international standard, ISO 3685. Flank wear is
quantified in terms of VB, which is the mean of the
wear height on the tool flank. The length of flank
wear is also measured in terms of b. Crater wear is
quantified in terms of the crater depth, K. The
parameters are depicted in Figure 26.5, which is a
scanning electron microscope (SEM) picture of a
worn turning insert.
26.2.3.4 Tool Wear Stages
It is assumed by most authors that tool wear
consists of an initial, a regular, and a fast wear stage
(Zhou et al., 1995). Some authors divide tool wear
into five distinct stages (Bonifacio and Diniz, 1994):
1. Initial stage of wear
2. Regular stage of wear
3. Microbreakage stage of wear
4. Fast wear stage
5. Tool breakage
It has been established by various researchers
that the initial and fast (before tool breakage)
stages occur more rapidly than the regular stage.
Bonifacio and Diniz (1994) explain that, during
the fast wear stage with coated carbide tools, the
tool loses its coating and the tool substrate (which
has less resistance) begins to perform the cut and
wears faster. During the initial stage, the tool edge
loses its sharp edge rapidly, after this the process
stabilizes for a given time. Flank wear in relation to
total cutting time will typically appear as depicted
in Figure 26.6.
The geometrical growth and rate of wear is
unique for every tool insert, even those used with
the same machining parameters. Wear measure-
ments conducted on the shop floor of a piston
manufacturer by Scheffer and Heyns (2004) are
shown in Figure 26.7. It was found that the tools
last between 1000 and 6000 components, which
makes the optimal use of the tool extremely
problematic if the wear is not monitored on-line.
The reason for this behavior is mainly attributed to
fluctuating conditions on the shop floor, for
example, the rate at which components are
manufactured. If the time allowed for the tool to
cool down between workpieces is not constant,
large variations in the tool life can be expected.
FIGURE 26.5 Tool wear parameters.
1 2 34
5initial regular fast
cutting time
flank wear
FIGURE 26.6 Flank wear in relation to cutting time.
0.25
0.2
0.15
0.1
0.05
0 40002000 6000
flan
k w
ear
VB
[m
m]
number of workpieces
FIGURE 26.7 Typical variations in tool life. Source:
Scheffer, C. and Heyns, P.S., Mech. Syst. Signal Process,
Elsevier, 2004. With permission.
Vibration and Shock Handbook26-1154
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
26.3 Vibration Signal Recording
The information from vibration sensors can be treated in numerous ways. The overall aim of a tool
condition monitoring system (TCMS) is to utilize the best processing techniques to extract the relevant
information from sensor signals. Generally, a TCMS consists of the steps depicted in Figure 26.8. Various
methods that could be used in each step will be discussed in more detail.
26.3.1 Direct and Indirect Systems
TCMSs can be divided in two categories, namely, direct and indirect. Direct methods are concerned with
a measurement of volumetric loss at the tool tip, while indirect methods use a pattern in sensor data to
detect a failure mode (Byrne et al., 1995). Direct methods do not utilize vibration and will not be
discussed here. In general, direct methods are sensitive to dirt and cutting chips, and consequently they
are not commonly accepted in industry. Indirect methods have found more acceptance in industry due to
the fact that most indirect methods are easily interpreted, cost-effective, and often more reliable than
direct methods. Also, for some applications, it might not be possible to use a direct monitoring method
due to the nature of the process.
26.3.2 Sensor Requirements for Tool Wear Monitoring
Machine tools represent very hostile environments for sensors. Sensors used for TCM (also see Chapter
15) must meet certain requirements, such as (Byrne et al., 1995) the following:
* Must measure as close as possible to the point of metal removal* Must not cause a reduction in the stiffness of the machine tool* Must not cause a restriction of the working space of the machine* Should be wear and maintenance free, easy to replace, and of low cost* Must have resistance to dirt, chips, and electromagnetic and thermal influences* Should function independent of tool and workpiece* Must provide reliable signal transmission, e.g., from rotating to fixed machine components
26.3.3 Force Measurement
Worn tools cause an increase in the cutting force components. It is also known that both the dynamic and
static components generally increase with tool wear due to frictional effects. The three components of the
cutting force each responds uniquely to varying machining parameters and the different wear modes.
Depending on the type of process that is investigated and the specific experimental setup, results among
researchers vary. This can be attributed to dynamic effects of the machine tool and the measurement
equipment. There are a number of different sensor configurations to collect forces from machining
operations and these are described below.
26.3.3.1 Direct Measurement Dynamometers
Tool holder dynamometers are by far the most popular method for collecting cutting forces. These
sensors utilize the piezoelectric effect and can measure quasistatic and dynamic cutting forces very
accurately. However, dynamometers are very expensive and bulky instruments and are not practical for a
sensor selectionand deployment
signal recordingand conditioning
generate signalfeatures
select wearsensitive features
model featuresand wear
relationship
FIGURE 26.8 TCMS steps.
Vibration-Based Tool Condition Monitoring Systems 26-1155
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
typical shop floor. Furthermore, their usable
frequency range is limited to approximately
1 kHz. An example of a tool holder dynamometer
is shown in Figure 26.9.
Tarmal and Opavsky (2000) investigated the
dynamics of a conventional force dynamometer
for machining operations. It was found that the
dynamometer has significant amplitude distortion
in the frequency range that is quoted as the
operating range by the manufacturer. The authors
suggest that the dynamic characteristics of the
dynamometer (while clamped as it would be
during measurements) be identified with a
modal test and the effect of dynamometer
dynamics be compensated for after measurements
are made to obtain the true cutting force.
26.3.3.2 Indirect Force Sensors
There are numerous small force sensors available
for the purpose of force measurement on machine
tools. These measure forces in load-carrying
components of the machine tool and are thus
not direct force measurement devices. The advan-
tages of these sensors are their size, low cost, and
significantly higher operational frequency range.
A disadvantage is that a suitable position for the
sensor can only be determined experimentally.
These sensors are suitable for tool breakage
monitoring in rough machining or detection of
other catastrophic events such as collisions. An
example of a three-component force sensor is
shown in Figure 26.10.
26.3.3.3 Piezoelectric Strain Sensors
The use of piezoelectric strain sensors for wear
monitoring of synthetic diamond tool inserts was
reported by Scheffer and Heyns (2000a). These
sensors are ultrasensitive to changes in cutting
forces if they are installed in an appropriate
location. The best location for the sensor must
once again be determined experimentally, but
generally it should be installed on a load-carrying
component of the machine as close as possible to
the tool tip, for example, on the tool holder itself
(Scheffer and Heyns, 2001b). An example of a
piezoelectric strain sensor that can be used on machine tools is shown in Figure 26.11.
26.3.3.4 Resistance Strain Gauges
A quite simple method to estimate both the static and dynamic components of cutting forces without any
distortion is to use resistance strain gauges (see Chapter 15). These comply with most of the requirements
for TCM sensors, and they can accurately follow the static and dynamic response of a system up to
FIGURE 26.9 KISTLER force dynamometer type 9121.
Source: KISTLER Brochure 2002. Courtesy of Kistler
Instrumente AG.
FIGURE 26.10 KISTLER three-component force sen-
sor type 9251A. Source: KISTLER Brochure 2003.
Courtesy of Kistler Instrumente AG.
FIGURE 26.11 KISTLER strain sensor type 9232A.
Source: KISTLER Brochure 2004. Courtesy of Kistler
Instrumente AG.
Vibration and Shock Handbook26-1156
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
50 kHz. Scheffer and Heyns (2002a) developed a sensor-integrated tool holder using strain gauges. It was
shown that the system is robust, cost-effective, and fit for an industrial TCMS. The physical layout of the
strain gauges on a boring bar is shown in Figure 26.12. The system was calibrated with a special device to
directly obtain the three cutting forces from the strain gauge signals.
26.3.3.5 Customized Force Sensors
There are a number of customized force sensors available that can be used with specific machining
operations. These are:
* Force measuring plates, pins, and bearings* Special force measuring bolts* Force and torque measuring rings that fit on spindles
26.3.4 Acceleration Measurement
Piezoelectric accelerometers can measure the machine vibration caused by oscillations of cutting forces. It
is well known that high-frequency vibrations (higher than 1 kHz) yield large acceleration levels, giving
accelerometers an advantage over force-based monitoring. Accelerometers fulfill the environmental
requirements for tool wear monitoring because they are resistant to the aggressive media present during
machining. Accelerometers are also less expensive than force dynamometers and can measure vibration
levels within a very wide frequency range, typically 5 Hz to 10 kHz.
Various authors have shown that acceleration levels change with tool wear. Li et al. (1997) found that
the coherence function of two crossed accelerations can be used as an easy and effective way to identify
tool wear and chatter. They found that with progressive tool wear, the autospectra of the two
accelerations and their coherence function increase gradually in magnitude around the first natural
frequencies of the cross-bending vibration of the tool shank. As the tool approaches a severe wear stage,
the peaks of the coherence function increase to values close to unity. Scheffer et al. (2003) reported on the
use of an accelerometer for wear monitoring during hard turning. It was found that certain frequencies
show repeatable amplitude increase with increasing tool wear. These frequencies corresponded to the tool
holder natural frequencies. Some authors, for example, Bonifacio and Diniz (1994), also found that aQ3
wear sensitive frequency will increase with increasing tool wear and then suddenly decrease near the end
of tool life. This can be attributed to an increased damping effect due to plastic deformation and
microbreakage of the cutting edge.
26.3.5 Acoustic Emission Measurement
Cutting processes produce elastic stress waves that propagate through the machine structure. Different
sources in the cutting process generate these stress waves known as acoustic emission (AE). Sources of AE
FIGURE 26.12 Application of resistance strain gauges. Source: Scheffer, C. and Heyns, P.S., Mech. Syst. Signal. With
permission.
Vibration-Based Tool Condition Monitoring Systems 26-1157
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
in metal cutting are:
* Friction on the tool face and flank* Plastic deformation in the shear zone* Crack formation and propagation* Impact of the chip at the workpiece* Chip breakage
A typical AE sensor for use on machine tools is
shown in Figure 26.13.
The fact that crack formation generates AE
makes AE ideal for tool breakage detection.
Collection of the AE requires special hardware that can bandpass filter the signals to the AE
range (between approximately 50 and 250 kHz). Furthermore, amplification is required and an
analogue root-mean-square (RMS) circuit with a short time constant is generally also included to collect
the AERMS. The different steps required to collect AE are depicted in Figure 26.14 (adapted fromQ4
Jemielniak, 2000).
Araujo et al. (2000) investigated sliding friction as a possible source of AE during metal cutting.
The AERMS values in different frequency ranges were collected for different widths of cut and also
with the tool rubbing against the workpiece without cutting. It was found that the level of AE
remains almost constant for all width of cut conditions, and hence it was concluded that the main
mechanism for AE during metal cutting is the sliding friction between the tool and workpiece.
Consequently, an increase or decrease of AE can be expected with tool wear depending on the effect
on the sliding friction due to that tool wear. Furthermore, it is believed that the cutting
temperatures will affect the AE due to thermal expansion effects. Chiou and Liang (2000)
investigated AE with tool wear and chatter effects in turning. A model is presented that can predict
the chatter AERMS amplitude with certain levels of flank wear. Good correlation was found between
the model and the experimental results. Kim et al. (1999) reported on the use of AE to monitor the
tool life during a gear shaping process. The AERMS is collected and used in a software program to
predict the remaining tool life.
Li (2002) presented an overview of using AE for TCM in turning operations. It is stated the AE is
heavily dependent on cutting conditions and, as a result, methods should be employed to handle this
problem effectively. Some methods are proposed that include advanced signal processing, sensor fusion
and modeling techniques. Many other AE-based tool wear and breakage monitoring systems have been
implemented successfully in research. One problem still lies with an appropriate interpretation of the AE
frequency spectrum. In most studies, an explanation for the choice of certain frequencies and their
advantages are not given or not investigated. In fact, Jemielniak (2000) found that using the average value
of AE (or AERMS) is the most suitable. A similar conclusion was made by Scheffer et al. (2003), who
compared different processing methods of the AE signal during hard turning.
FIGURE 26.13 Kistler AE sensor type 8152B. Source:
PCB Website 2002. Courtesy of Kistler Instrumente AG.
FIGURE 26.14 Steps for collecting AE during turning. Source: KISTLER brochure 2004. Adapted from Jemielniak,
K., Ultrasonics, 2000.
Vibration and Shock Handbook26-1158
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
26.3.6 Sensor Comparisons
Future research should be directed towards directly comparing different sensors for tool wear
monitoring. Choi et al. (1999) developed a single sensor for parallel measurement of force and AE. A
finite element analysis was carried out to determine the optimal position for the sensor away from the
tool holder because the sensor obstructed the working space of the machine. The approach was successful
for breakage detection but no wear estimations are reported. Barrios et al. (1993) compared AE,
vibration, and spindle current for TCM during milling. It was found that the spindle current is the most
sensitive sensor for detecting tool wear, with AE the least sensitive. However, contradictory results are
reported in other publications, and hence more research would be required to determine ultimately
which sensor is the best for which machining operation. Govekar et al. (2000) compared force and AE
methods for TCM, and concluded that the best result is achieved when sensory information is combined.
Dimla and Lister (2000a) compared the use of force and vibration signals for TCM and also combined the
information in a single decision-making technique (Dimla and Lister, 2000b). Similar comparative
studies were reported by Scheffer et al. (2003).
26.4 Signal Processing for Sensor-Based Tool ConditionMonitoring
Using the sensor information from the different sensor systems described in the previous section, a
decision must be made with respect to the tool condition. This decision is generally referred to as the data
classification. It is often better to combine sensory information to solve a complex problem such as TCM.
Such a combined approach is referred to as sensor fusion. Sick (2002) proposed a generic sensor fusion
architecture for TCM, which summarizes the various sensor fusion levels of a TCMS. These are:
* Analogue preprocessing* Digital preprocessing* Feature extraction* Wear model* Decision making
Fusion of sensor information can occur at any of these levels. Analogue and digital preprocessing are
activities such as signal amplification, conditioning, filtering, calibration, and temperature compen-
sation. The feature extraction step is probably the most important step, because here the sensor signals
must be condensed and reduced to only a few appropriate wear sensitive values. Many different methods
are available to achieve this. The wear model level establishes a relationship between the chosen features
and the tool condition. In many cases, neural networks (NNs) are used in this step, and sensor fusion
takes place within the NN. A decision level can also be included where a final decision is made with
respect to the tool condition, for instance a “competing experts” formulation if a TCMS is used in
conjunction with a tool-life equation. Discussions on the various techniques follow.
26.4.1 Feature Extraction
Most decision-making techniques for process monitoring are based on signal features. Through
appropriate signal processing, features can be extracted from signals that show consistent trends with
respect to tool wear. Features are mainly derived through processing in the time, frequency, or joint time-
frequency domain or statistical analysis.
26.4.1.1 Time Domain
Features extracted from the time domain are usually fundamental values such as the signal mean or RMS.
Other techniques include the shape of enveloping signals, threshold crossings, ratios between
Vibration-Based Tool Condition Monitoring Systems 26-1159
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
time-domain signals, peak values, and polynomial
approximations of time-domain signals. Examples
of time-domain features from an interrupted
cutting operation are shown in Figure 26.15.
It has been found that some of the time-domain
features show good correlation with tool wear and
are easy to implement (Scheffer, 1999). Bayramo-
glu and Dungel (1998) investigated the use of
several different force ratios (calculated from the
static cutting forces). It was found that certain
force ratios can be used to monitor tool wear
under a wide range of cutting conditions. Most
commercial TCMS rely on time-domain infor-
mation. However, time-domain features are
known to be sensitive to disturbances and should
be complemented with features from another
domain.
26.4.1.2 Frequency Domain
The power or energy of certain frequency bands
in the fast Fourier transform (FFT) is often
suggested as a feature for TCM. It is very
challenging to identify spectral bands that are
sensitive to tool wear. It is even more difficult to
determine exactly why these frequencies are
influenced by tool wear. Power in certain bands
will often increase due to higher excitation forces
because of the increase in friction when the tool
starts to wear. Sometimes a peak in the FFT will
also shift due to changing process dynamics as a
result of tool wear. An early frequency-domain
approach is reported by Jiang et al. (1987), in
which frequency-band energy is determined from
the power spectral density (PSD) function as a
feature for tool wear.
Some authors suggest that two frequency ranges be identified from the original signal (Bonifacio and
Diniz, 1994). The one range must be sensitive to tool wear, the other must be insensitive. For instance, if
the measurement was made from 0 to 8 kHz, it must be split (using appropriate filters) into a 0 to 4 kHz
signal and a 4 to 8 kHz signal. If the lower range is more sensitive to tool wear, a ratio between the two
ranges can be calculated. If this ratio exceeds a certain pre established value, it can be concluded that the
end of the tool life has been reached. This can also apply for a ratio between the signal recorded from a
fresh tool to that compared with a worn tool. Examples of frequency-domain features from cutting forces
are shown in Figure 26.16.
One difficulty with frequency-domain approaches is that the dynamics of the operation and
measurement hardware is not always fully understood. The fact that measurement hardware dynamics
instead of process dynamics are often measured was also recently identified by Warnecke and Siems
(2002). The response of a force dynamometer is influenced by its clamping condition, which may cause it
to experience nonlinearities at relatively low frequencies. There are also some uncertainties when using
these instruments, relating to their calibration and other varying parameters. A model for expressing the
uncertainties when collecting cutting forces with a dynamometer was proposed by Axinte et al. (2001).
These uncertainties might be responsible for the scatter of force components often reported in the
FIGURE 26.15 Simple time-domain features.
FIGURE 26.16 Frequency-domain features.
Vibration and Shock Handbook26-1160
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
literature. An interesting study is also reported by Bahre et al. (1997), concerning determination of the
natural frequencies of the machine tool components using the finite element method (FEM). These are
taken into account for interpretation of the vibration/AE signal.
26.4.1.3 Statistical Processing
In the case of statistical features, signals are assumed to have a probabilistic distribution, and
consequently, useful information can be extracted from the statistics of the distribution. Hence, the signal
is regarded as a random process. Generally, machining processes are nonstationary but are assumed to be
stationary for the short periods during which features are calculated. Several statistical features have been
investigated for TCM and can be applied to several machining operations. The main features are those
that describe the probability distribution of a random process (variance, standard deviation, skewness,
kurtosis, etc.) and coefficients of time-series models. There are also miscellaneous other statistical
features, such as cross-correlations, the coherence function, and the harmonic mean.
One useful approach is the use of autoregressive (AR) and autoregressive moving average (ARMA)
coefficients. AR coefficients computed for a signal represent its characteristic behavior. When the signal
changes during the cutting operation due to tool wear, the model coefficients also change and can then be
utilized to monitor the progressive tool wear. Baek et al. (2000) report on the use of an eighth order AR
model for tool breakage detection during end milling. It was found that the AR approach is somewhat
more accurate than the frequency band energy method. Yao et al. (1990) used the ARMA method to
decompose the dynamic cutting force signals, and wear-sensitive frequencies were identified. This
assisted in identifying the importance of certain vibration modes with respect to TCM.
The use of statistical process control (SPC) methods is also reported by some authors. Jun and Suh
(1999) considered the X-bar and exponentially weighted moving average (EWMA) for tool breakage
detection in milling. Jennings and Drake (1997) used statistical quality control charts for TCM. Different
statistical parameters are calculated and examples of one-, two- and three-variable control charts are given.
26.4.1.4 Time–Frequency Domain
The most common time–frequency domain processing method in TCM applications is wavelet analysis.
A comprehensive discussion on the advantages and disadvantages of wavelet analysis for TCM is
described by Sick (2002). It is often stated that wavelets are used because they provide information about
the localization of an event in the time as well as in the frequency domain. However, locating discrete
frequency-related events in the time domain is rarely of importance with respect to tool wear (which is a
gradually increasing phenomenon). In contrast, tool breakage will have a large localized effect in the time
domain, but this can be monitored more effectively using time-domain techniques. Furthermore,
wavelets are time variant and the exact contribution of a particular frequency at any given time can never
be determined accurately due to Heisenberg’s uncertainty principle.
Despite the above arguments, the use of wavelet analysis for TCM is reported in several publications.
Lee and Tarng (1999) use the discrete wavelet transform for cutter breakage detection in milling and find
that the technique is reliable even under changing machining conditions. Luo et al. (2002) published
results of a TCMS using wavelet analysis of vibration signals. In this case, the wavelet is used as a filter to
enhance wear-sensitive features in the signals. However, the results are not compared with conventional
digital filtering. A comparative study between wavelets and digital filtering for tool wear monitoring was
carried out by Scheffer (2002). It was found that, although the wavelet packets act as automated filters, a
very similar (if not better) result could be achieved with appropriate digital filtering. The use of wavelets
increase the complexity of the TCMS, which is a disadvantage for shop-floor implementations.
Furthermore, the results from digital filtering can be physically related to the machining operation and
tool wear, whereas the behavior of wavelet packets is more difficult to interpret.
Another method of time–frequency analysis that can be applied for TCM is spectograms (e.g., the
Gabor distribution). Spectograms are very useful to identify stationarity in dynamic signals, and
for detection of disturbances that may be time-localized in signals. The use of the Choi–Williams
Q5 time–frequency distribution for TCM during multimilling is described by James and Tzeng (2000).
Vibration-Based Tool Condition Monitoring Systems 26-1161
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
Wear-sensitive regions on the time–frequency
distribution are calculated and used as inputs to
a NN for wear classification. An example of a
change in the dominant chip curl frequency during
hard turning is shown in Figure 26.17. It is obvious
that, due to some disturbance (perhaps tool wear),
the dominant dynamic force frequency “jumps”
from 27 to 9 Hz.
26.4.2 Feature Selection
Various authors attempt to generate features that
are sensitive to tool wear but insensitive to
changing machining parameters. For most oper-
ations, the machining parameters can be included
in the wear model and hence the sensitivity of the
features is not such an important issue. There are
also other techniques to normalize sensor data
with respect to machining parameters, for instance
the use of a theoretical model (Sick, 1998). This is
very useful if the machining conditions change so often that not enough data can be collected for training
or calibrating a model. Numerous techniques exist to select the most wear-sensitive features or to reduce
the input feature matrix to a lower dimension. The main techniques for feature selection and reduction
are listed below:
* Principal component analysis (PCA)* Statistical overlap factor (SOF)* Genetic algorithm (GA)* Partial least squares (PLS)* Automatic relevance determination (ARD)* Analysis of variance (ANOVA)* Correlation coefficient* Simulation error calculations
Al-Habaibeh et al. (2000) presented a TCMS for a parallel kinematics machine tool for high-speed
milling of titanium. An interesting approach to feature selection is employed, called self-learning
automated sensors and signal processing selection (ASPS). This approach is based on an on-line self-
learning methodology, whereby a certain feature will be selected automatically based on a correlation
with tool wear. A linear regression is performed on each feature in the sensory feature matrix to detect the
sensitivity of each feature with respect to tool wear. A very interesting cost analysis is then preformed to
determine if the installation of a sensor justifies its costs.
Ruiz et al. (1993) proposed the use of a discrimination power for feature selection in a TCM
application. The method is similar to that of the SOF. An automated version is proposed that also checks
for linear correlation between features. It is difficult to assess the success rate of the automated procedure
because the experiments/simulations are not described in enough detail. Lee et al. (1998) describe the use
of ANOVA to determine the best force ratio for TCM statistically. Several ratios between the three main
cutting forces are computed and the influence of controllable parameters (e.g., machining conditions) on
these ratios are investigated by means of ANOVA.
Du (1999) describes the use of a blackboard system, which is a knowledge-based approach for feature
selection and decision-making. An advantage is the fact that a physical interpretation of a feature can be
linked to phenomena in the machining operation. The method is also flexible, but suffers from the
disadvantage of requiring a large quantity of data and expertise to establish the knowledge-based rules.
FIGURE 26.17 Time-frequency distribution of cutting Q7
force signal. Source: Scheffer, et al., Int. J. Mach. Tools
Manuf., Elsevier, 2003. With permission.
Vibration and Shock Handbook26-1162
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
696
697
698
699
700
701
702
703
704
705
706
707
708
709
710
711
712
713
714
Mdlazi et al. (2003) compared the performance of ARD and PCA for feature selection for two damage
detection case studies. It was found that the performance of the methods is similar, but one might
perform better on a particular data set. Generally speaking, the PCA yields better results for damage
detection problems. Scheffer and Heyns (2002b) compared several feature selection methods for TCM,
such as SOF, PCA, GA, ANOVA, and the linear correlation coefficient. It was found that the correlation
coefficient approach and the SOF should be preferred for TCM applications. PCA could also be of
assistance, but the feasibility of PCA for on-line applications is still questionable. The correlation
coefficient and SOF is expressed as percentages in Figure 26.18 (from Scheffer and Heyns, 2004) for 30
different wear monitoring features in a turning tool wear case study. Ideally, a feature with a high level of
correlation and SOF should be selected.
As a last step, engineering judgment is required for proper feature selection because automated
methods will often select features that are dependant on one another, thus not achieving the goals of
sensor fusion. The following rules can be used as a guideline for selecting features for TCM:
* Select features from the static and dynamic parts of force signals.* Select features measured in different directions.* Use time- and frequency-domain features.* Features based on simple signal processing methods are preferred.* There should be a reasonable physical explanation for the behavior of a feature with respect to
tool wear.
26.5 Wear Model/Decision-Making for Sensor-Based ToolCondition Monitoring
26.5.1 Trending, Threshold
A very simple decision-making technique is to
trend features and to establish threshold values.
When a certain feature or set of features crosses a
threshold value, an estimation of the tool con-
dition can be made. Unfortunately, these threshold
values can only be determined experimentally.
The difficulty with this method is to determine
the correct threshold value, especially under
diverse cutting conditions. Furthermore, the
method is extremely sensitive to disturbances.
The trend of the mean feed force with increasing
flank wear is shown in Figure 26.19, and two
thresholds are shown as examples. It is clear that
this technique is not very reliable due to the large
variance in the trend.
FIGURE 26.18 Comparison of correlation coefficient and SOF for feature selection. Source: Scheffer, C. and Heyns,
Q7
P.S., Mech. Syst. Signal Process, Elsevier, 2004. With permission.
threshold-replace
threshold-warning
flank wear VB [mm]
feat
ure
valu
e [n
orm
alis
ed]
0.04 0.06 0.08
−1
−0.5
0
0.5
1
FIGURE 26.19 Example of trend and thresholds.
Source: Scheffer, C. and Heyns, P.S., South African Inst.
Q7
Tribol. 2002. With permission.
Vibration-Based Tool Condition Monitoring Systems 26-1163
715
716
717
718
719
720
721
722
723
724
725
726
727
728
729
730
731
732
733
734
735
736
737
738
739
740
741
742
743
744
745
746
747
748
749
750
751
752
753
754
755
756
757
758
759
760
761
762
763
764
765
26.5.2 Neural Networks
The use of NNs as a secondary, more sophisticated signal processing and decision-making technique is
often found in TCM applications. The simultaneous utilization of many features and the robustness
towards distorted sensor signals are two of the most attractive properties of NNs. Neural networks also
assist in the fusion of sensor information for TCM. In other words, combining features from acceleration,
AE, and force signals in a NN can result in a method that can predict the tool condition with increased
accuracy (Silva et al., 1998). The successful implementation of NNs is dependent on the proper selection
of the network structure, as well as the use of the correct training and testing methods.
It is important to make a distinction between supervised and unsupervised NN paradigms.
Unsupervised NNs are trained with input data only, and are usually used for discrete classification of
different stages of tool wear. Supervised NNs are trained with input and output data, and these are used
for continuous estimations of tool wear. Furthermore, a distinction should be made between dynamic
and static NNs. In the case of dynamic NNs, temporal (time) information is included in the network with
the aim to model a time series. This can be done explicitly by using a time-based feature as an input to the
network, or implicitly by using recurrent networks or networks with tapped delay lines (TDLs). Dynamic
networks are preferred for TCM because tool wear is time-dependent (tool wear is a monotonically
increasing parameter that is partly a function of machining time).
26.5.2.1 Unsupervised Networks
There are two basic network paradigms for unsu-
pervised classifications, namely adaptive resonance
theory (ART) and the self-organizing map (SOM).
ART is based on competitive learning, addressing
the stability–plasticity dilemma of NNs. The main
advantage is its ability to adapt to changing
conditions. ART networks also have self-stability
and self-organization capabilities. The SOM is
actually a data-mining method used to cluster
multidimensional data automatically. A high-
dimensional feature matrix can be displayed on a
two-dimensional grid of neurons that are arranged
in clusters with similar feature values. Clusters for new and worn tools can be formed, and these are used for
automatic classification of the tool condition. A SOM is depicted schematically in Figure 26.20.
There are many practical advantages for using unsupervised networks. One is the fact that the
machining operation is not interrupted for tool wear measurements during the training phase. There is
also the advantage of practical implementation if machining conditions change very often and
appropriate training samples for supervised learning cannot be collected. Furthermore, the numerous
different combinations of tool and workpiece materials and geometries can make supervised learning
impossible. Normally, unsupervised NNs are used to identify discrete wear classes and cannot be used for
a continuous estimation of tool wear.
Silva et al. (2000) investigated the adaptability of the SOM and ART for tool wear monitoring during
turning with changing machining conditions. It was found that, with appropriate training, the methods
have enough adaptive capabilities to be employed in industrial applications. Govekar and Grabec (1994)
use the SOM for drill wear classification, where the SOM is used as a kind of empirical modeler. It was
found that the adaptability of the SOM and its ability to handle noisy data makes the technique viable for
on-line TCM. Scheffer and Heyns (2000b, 2001b) showed how a TCMS can be adaptable using SOMs.
Different network sizes were compared with define discrete classes of new and worn tools. Larger networks
yielded more continuous results. The TCMS using SOMs was applied to monitoring synthetic diamond
tools for an industrial turning operation. It was found that the SOM can be used for industrial applications,
especially if tool wear measurements are not available.
FIGURE 26.20 Schematic representation of the SOM.
Vibration and Shock Handbook26-1164
766
767
768
769
770
771
772
773
774
775
776
777
778
779
780
781
782
783
784
785
786
787
788
789
790
791
792
793
794
795
796
797
798
799
800
801
802
803
804
805
806
807
808
809
810
811
812
813
814
815
816
Different NN paradigms were compared on a wear monitoring application for aluminum turning by
Scheffer and Heyns (2002b). It was shown that the SOM is useful to identify discrete wear classes, as
shown in Figure 26.21. If an exact value of the tool wear is required, supervised networks will yield better
results but will require proper training samples.
26.5.2.2 Supervised Networks
Common supervised NNs used for TCM are the multilayer perceptron (MLP), multilayer feedforward
(FF) network, recurrent neural network (RNN), supervised neuro-fuzzy system (NFS-S), time delay
neural network (TDNN), single layer perceptron (SLP) and the radial basis function (RBF) network. The
use of an SLP for TCM is described by Dimla et al. (1996), using the perceptron learning rule for training.
The SLP is useful to identify discrete classes of the tool condition. FF networks are usually trained with
the backpropagation algorithm. However, backpropagation should not always be the preferred choice
because other methods are known that outperform this technique in terms of training time and
generalization. The size of the hidden layers in multilayer networks should be optimized for performance.
Many contradictory statements about the use of MLP networks can be found in the literature. One of the
main problems is the selection of the number of input features, size of the network, and the number of
training examples that should be used.
A multilayer feedforward (FF) network is shown schematically in Figure 26.22. Normally, a nonlinear
activation function should be used in the first layer, and linear neurons in the subsequent layers. In the
case of the FF networks, the backpropagation algorithm is often used for training. Backpropagation is an
optimization algorithm based on steepest gradient descent.
The use of FF networks with the backpropaga-
tion training rule is reported by authors such as
Zhou et al. (1995), Das et al. (1996), and
Zawada-Tomkiewicz (2001). Cutting conditions
can also be included in such networks. Lou and Lin
(1997) describe the use of a FF network using a
Kalman filter to avoid the training problems
encountered with backpropagation for a TCM
application. The proposed method is less sensitive
to the network initializations that often cause
convergence problems with backpropagation.
Monitoring a dynamic system such as a cutting
process should be done with a dynamic modeling
technique such as dynamic NN paradigms, for
example, recurrent networks, TDNNs, or explicit
FIGURE 26.21 Unsupervised approach to wear monitoring with the SOM. Source: Scheffer, C. and Heyns, Q7
P.S., South African Inst. Tribol. 2002. With permission.
FIGURE 26.22 Multilayer FF network.
Vibration-Based Tool Condition Monitoring Systems 26-1165
817
818
819
820
821
822
823
824
825
826
827
828
829
830
831
832
833
834
835
836
837
838
839
840
841
842
843
844
845
846
847
848
849
850
851
852
853
854
855
856
857
858
859
860
861
862
863
864
865
866
867
inclusion of temporal information in static networks. Recurrent NNs have feedback connections
from their output to their input. There are various types of recurrent NNs that are useful for
specific applications. Elman networks are quite interesting. Generally, they are two-layer networks
with feedbacks from the first layer output to the first layer input. This type of network can be used to
learn and model temporal patterns. A recurrent network and an Elman network are shown schematically
in Figure 26.23.
Liu and Altintas (1999) report on the use of a FF network using a combination of TDLs and recurrent
connections. Machining conditions are also included. It is stated that the system was integrated into an
industrial TCMS, but was never put to use due to lack of “… robust, practical cutting force sensors …”
(Liu and Altintas 1999). Scheffer and Heyns (2002b) report on the use of an Elman NN for TCM. It was
found that the Elman network has a very smooth response and yielded better results than static NN
paradigms. It should be mentioned that the Elman network requires more time for training, but because
this is done off-line, training time should not be a criterion for evaluating NNs.
Neuro-fuzzy systems (NFS-S) attempts to combine the learning ability of NNs with the interpretation
ability of fuzzy logic. A TCMS using an NFS-S can be generated almost automatically because the fuzzy
rules can be learned by the NN. A combination of supervised and unsupervised training is used for
NFS-S. An in-process NFS-S system to monitor tool breakage was designed and implemented
successfully by Chen and Black (1997), concentrating on end milling operations. Xiaoli et al. (1997)
as well as Chungchoo and Saini (2002) also propose some of the advantages of using an NFS-S for TCM.
RBF networks are often preferred because of the convergence properties of the training algorithm.
In essence, convergence can be guaranteed and is often achieved much faster than in MLPs. The
accuracy of RBFs depends on the choice of the centers for the basis functions, and should be treated
with care. Pai et al. (2001) reported on the use of a resource allocation network (RAN) for TCM. The
RAN is a RBF network utilizing sequential learning. The RAN is compared with the MLP for wear
estimation during face milling. It was found that the RAN has faster learning ability but the MLP is
more robust.
TDNNs have delay elements in the feedforward connections, called TDLs. One advantage of
TDNNs over RNNs is that stability problems are avoided. An investigation towards the inclusion of
one and two phase delays for a TCM application was reported by Venkatesh et al. (1997). Different
network sizes were also investigated, and it was found that the NNs with temporal memory generally
perform better than those without memory. It is also stated that new algorithms should be
investigated for training. Sick and Sicheneder (1997) also describe the use of TDNNs for TCM in
turning. The TDNN is compared with the MLP and a significant improvement was found when using
FIGURE 26.23 Recurrent networks: feedback connection (left) and Elman network (right). Source: Scheffer, C. and Q7
Heyns, P.S., South African Inst. Tribol. 2002. With permission.
Vibration and Shock Handbook26-1166
868
869
870
871
872
873
874
875
876
877
878
879
880
881
882
883
884
885
886
887
888
889
890
891
892
893
894
895
896
897
898
899
900
901
902
903
904
905
906
907
908
909
910
911
912
913
914
915
916
917
918
TDNNs. In another instance, Sick et al. (1998) compare the SOM, NFS-S, and MLP networks for
wear estimation. The following critical questions are used to evaluate the different NN paradigms
(Sick et al., 1998):
* Are the generalization capabilities of the NN sufficient (tested on previously unseen data)?* What rate of correct classification can be achieved for different wear stages?* Are the results repeatable (e.g., with a new initialization)?
In the case study presented by Sick et al. (1998), the best results were found with MLPs. It is
stated, however, that the results can be improved when using TDNNs, and such results are reported in
Sick (1998).
A novel combined approach is suggested by Sick (1998) to handle the effect of machining parameters.
An empirical model is used to normalize the data with respect to machining parameters before the data
are entered into the NN. Thus, machining parameters are not included in the NN itself. This approach
solves the extrapolation limitations encountered when an NN is tested with data recorded with
machining parameters it was not trained with. Although many authors test their NNs’ paradigms in such
a way, NNs cannot be expected to extrapolate. NNs should instead be tested with previously unseen data
recorded with same machining parameters it was trained with (hence an interpolation effect). This is a
problem because training and testing patterns for each condition must be supplied. However, if data can
be normalized with respect to machining parameters, training is only required for the normalized
condition. This was in effect achieved by Sick (1998). A difficulty still lies with establishing an appropriate
model, and in many cases it will also require a large number of experimental tests. A possible solution lies
in the incorporation of numerical models, for example, finite element models.
Scheffer (2002) presented another approach to tool wear monitoring of turning operations, using a
combination of static and dynamic NNs. Static networks are trained off-line to model selected features
from cutting forces. A dynamic NN that uses explicit temporal information is then trained on-line
with the particle swarming optimization algorithm (PSOA). The training goal of the dynamic NN
is to minimize the errors between the outputs of the static NNs and the on-line measurements.
The method was tested on various turning operations and was also tested on an industrial shop floor.
It was found that the method is more accurate and reliable than other NN paradigms and can be used
with cost-effective hardware (Scheffer and Heyns 2002a, 2004). The method is depicted schematically
in Figure 26.24.
FIGURE 26.24 Combined static and dynamic NN approach for turning. Source: Scheffer, C. and Heyns, P.S., South Q7
African Inst. Tribol. 2002. With permission.
Vibration-Based Tool Condition Monitoring Systems 26-1167
919
920
921
922
923
924
925
926
927
928
929
930
931
932
933
934
935
936
937
938
939
940
941
942
943
944
945
946
947
948
949
950
951
952
953
954
955
956
957
958
959
960
961
962
963
964
965
966
967
968
969
26.5.3 Fuzzy Logic
Many authors have investigated the use of fuzzy logic to classify tool wear. It has been shown that fuzzy
logic systems demonstrate great potential for use in intelligent manufacturing applications. While NN
models cannot directly encode structured knowledge, it is often stated that fuzzy systems can directly
encode structured knowledge in a numerical framework. Additionally, fuzzy systems are capable of
estimating functions of a system with only a partial description of the system’s behavior.
Du et al. (2002) propose a very interesting method called transition fuzzy probability, which was
applied to a boring operation. This formulation can deal with the uncertainty of process conditions.
The method performs well because TCM has two uncertainties: that of occurrence and that of
appearance. The transition fuzzy probability solves this issue through the use of temporal information,
similar to dynamic NNs. The method was shown to outperform a backpropagation NN, although
very few details are given. It would be interesting to compare this method with dynamic NNs such
as TDNNs.
Fu et al. (1997) combined force, vibration, and AE in a fuzzy classifier for TCM during milling. Time-
and frequency-domain features were used, and it was found that combining the sensory information
achieved the best result. This is done within the fuzzy classifier. Li and Elbestawi (1996) and Kuo and
Cohen (1998) combine fuzzy modeling steps with NNs at different levels for TCM. The latter combined
force, vibration and AE in a multisensor approach with satisfactory results.
26.5.4 Other Methods
There are also a number of other decision-making and modeling methods that have been applied to
TCM, and these include:
* Knowledge-based expert systems (Du, 1999)* Pattern recognition algorithms (Kumar et al., 1997)* Dempster–Shafer theory of evidence (Beynon et al., 2000)* Hidden Markov models (Ertunc and Loparo, 2001; Ertunc et al., 2001)
Of these four approaches, only hidden Markov models have the potential possibly to outperform NNs
and fuzzy systems. However, not enough comparable research has been conducted in this area, and is it
certainly a worthwhile topic for future research.
26.6 Conclusion
Techniques for achieving TCM with vibration-based properties were presented in this chapter. The
sensing methods that have proved to be effective for TCM are force, acceleration, and AE. The sensors
employed must comply with certain requirements such as robustness and cost-effectiveness. Sensors
must be installed as close as possible to the point of metal removal in order to avoid signal-to-noise ratio
problems. Various techniques exist to condition and process the signals in analogue and digital formats.
The aim of signal processing is to generate wear sensitive features from the vibration signals. This could
be done by time, frequency, joint time–frequency, and statistical analysis. Feature selection can be
automated with a variety of procedures, but care must be taken when using these to avoid selection of
linearly dependent data.
The selected features can be used to establish a model of tool wear. Numerous research papers have
shown that NNs should be used due to the many advantages of NN modeling. The training and testing
procedures of NNs are of utmost importance if the system is considered for industrial implementation.
Care must be taken not to overtrain the networks because they will lose their ability to generalize.
Furthermore, NNs cannot be expected to perform well if they are tested with previously unseen
machining parameters. They should also be trained with the minimum and maximum tool wear that is
Vibration and Shock Handbook26-1168
970
971
972
973
974
975
976
977
978
979
980
981
982
983
984
985
986
987
988
989
990
991
992
993
994
995
996
997
998
999
1000
1001
1002
1003
1004
1005
1006
1007
1008
1009
1010
1011
1012
1013
1014
1015
1016
1017
1018
1019
1020
expected. Future work should be directed towards incorporating numerical machining models into the
wear monitoring system to normalize the data with respect to machining parameters. If this can be
achieved, the amount of training data required for an effective TCMS will be reduced, which in turn will
provide a better solution to TCM for the manufacturing industry.
References
Al-Habaibeh, A., Gindy, N., and Radwan, N. 2000. An automated approach for monitoring gradual tool
wear in high speed milling of titanium, pp. 371–380. In Proceedings of the 13th International
Congress on Condition Monitoring and Diagnostic Engineering Management (COMADEM 2000),
Houston, TX, 3–8 December.
Araujo, A.J.M.M., Wilcox, S.J., and Reuben, R.L. 2000. Sliding friction as a possible source of acoustic
emission in metal cutting, pp. 381–387. In Proceedings of the 13th International Congress on
Condition Monitoring and Diagnostic Engineering Management (COMADEM 2000), Houston, Tx,
3–8 December.
Axinte, D.A., Belluco, W., and De Chiffre, L., Evaluation of cutting force uncertainty components in
turning, Int. J. Machine Tools Manuf., 41, 719–730, 2001.
Baek, D.K., Ko, T.J., and Kim, H.S., Real time monitoring of tool breakage in a milling operation using a
digital signal processor, J. Mater. Process. Technol., 100, 266–272, 2000.
Bahre, D., Muller, M., and Warnecke, G. 1997. Basic characteristics on cutting effects in correlation to
dynamic effects, pp. 21–26. 1997 Technical Papers of the North American Manufacturing Research
Institution of SME.
Barrios, L.J., Ruiz, A., Guinea, D., Ibanez, A., and Bustos, P., Experimental comparison of sensors for
tool-wear monitoring on milling, Sensors Actuators A, 37–38, 589–595, 1993.
Bayramoglu, M. and Dungel, U., A systematic investigation on the force ratios in tool condition
monitoring for turning operations, Trans. Inst. Measur. Control, 20, 92–97, 1998.
Beynon, M., Curry, B., and Morgan, P., The Dempster–Shafer theory of evidence: an alternative
approach to multicriteria decision modelling, Omega, 28, 37–50, 2000.
Bonifacio, M.E.R. and Diniz, A.E., Correlating tool wear, tool life, surface roughness and tool vibration in
finish turning with coated carbide tools, Wear, 173, 137–144, 1994.
Byrne, G., Dornfeld, D., Inasaki, I., Ketteler, G., Konig, W., and Teti, R., Tool Condition Monitoring
(TCM) — The status of research and industrial application, Ann. CIRP, 44, 541–567, 1995.
Chen, J.C. and Black, J.T., A Fuzzy-Nets-In-Process (FNIP) system for tool breakage monitoring in end-
milling operations, Int. J. Machine Tools Manuf., 37, 783–800, 1997.
Chiou, R.Y. and Liang, S.Y., Analysis of acoustic emission in chatter vibration with tool wear effect in
turning, Int. J. Machine Tools Manuf., 40, 927–941, 2000.
Choi, D., Kwon, W.T., and Chu, C.N., Real-time monitoring of tool fracture in turning using sensor
fusion, Int. J. Adv. Manuf. Technol., 15, 305–310, 1999.
Chungchoo, C. and Saini, D., On-line tool wear estimation in CNC turning operations using fuzzy neural
network model, Int. J. Machine Tools Manuf., 42, 29–40, 2002.
Dan, L. and Mathew, J., Tool wear and failure monitoring techniques for turning — a review, Int.
J. Machine Tools Manuf., 30, 579–598, 1990.
Das, S., Chattopadhyay, A.B., and Murthy, A.S.R., Force parameters for on-line tool wear estimation: a
neural network approach, Neural Networks, 9, 1639–1645, 1996.
Dimla, D.E., Sensor signals for tool-wear monitoring in metal cutting operations — a review of methods,
Int. J. Machine Tools Manuf., 40, 1073–1098, 2000.
Dimla, D.E. and Lister, P.M., On-line metal cutting tool condition monitoring I: force and vibration
analyses, Int. J. Machine Tools Manuf., 40, 739–768, 2000a.
Dimla, D.E. and Lister, P.M., On-line metal cutting tool condition monitoring II: tool-state
classification using multi-layer perceptron neural networks, Int. J. Machine Tools Manuf., 40,
769–781, 2000b.
Vibration-Based Tool Condition Monitoring Systems 26-1169
1021
1022
1023
1024
1025
1026
1027
1028
1029
1030
1031
1032
1033
1034
1035
1036
1037
1038
1039
1040
1041
1042
1043
1044
1045
1046
1047
1048
1049
1050
1051
1052
1053
1054
1055
1056
1057
1058
1059
1060
1061
1062
1063
1064
1065
1066
1067
1068
1069
1070
1071
Dimla, D.E., Lister, P.M., and Leighton, N.J. 1996. Investigation of a single-layer perceptron neural
network to tool wear inception in a metal turning process, pp. 3/1–3/4. In Proceedings of the 1997
IEE Colloquium on Modelling and Signal Processing for Fault Diagnosis.
Du, R., Signal understanding and tool condition monitoring, Eng Appl. Artif. Intell., 12, 585–597, 1999.
Du, R., Liu, Y., Xu, Y., Li, X., Wong, Y.S., and Hong, G.S. 2002. Tool condition monitoring using
transition fuzzy probability. In Metal Cutting and High Speed Machining, pp. 375–392, Kluwer
Academic/Plenum Publishers, New York.
Ertunc, H.M. and Loparo, K.A., A decision fusion algorithm for tool wear condition monitoring in
drilling, Int. J. Mach. Tools Manuf., 41, 1347–1362, 2001.
Ertunc, H.M., Loparo, K.A., and Ocak, H., Tool wear condition monitoring in drilling operations using
Hidden Markov Models (HMMs), Int. J. Mach. Tools Manuf., 41, 1363–1384, 2001.
Fu, P., Hope, A.D., and Javed, M.A., Fuzzy classification of milling tool wear, Insight, 39, 553–557,
1997.
Govekar, E. and Grabec, I., Self-organizing neural network application to drill wear classification, Trans.
ASME: J. Eng. Ind., 116, 233–238, 1994.
Govekar, E., Gradisek, J., and Grabec, I., Analysis of acoustic emission signals and monitoring of
machining processes, Ultrasonics, 38, 598–603, 2000.
James, L.C. and Tzeng, T., Multimilling-insert wear assessment using non-linear virtual sensor, time–
frequency distribution and neural networks, Mech. Syst. Signal Process., 14, 945–957, 2000.
Jemielniak, K., Some aspects of AE application in tool condition monitoring, Ultrasonics, 38, 604–608,
2000.
Jennings, A.D. and Drake, P.R., Machine tool condition monitoring using statistical quality control
charts, Int. J. Mach. Tools Manuf., 37, 1243–1249, 1997.
Jiang, C.Y., Zhang, Y.Z., and Xu, H.J., In-process monitoring of tool wear stage by the frequency band
energy method, Ann. CIRP, 36, 45–48, 1987.
Jun, C. and Suh, S., Statistical tool breakage detection schemes based on vibration signals in NC milling,
Int. J. Mach. Tools Manuf., 39, 1733–1746, 1999.
Kapoor, S.G., DeVor, R.E., and Zhu, R. 1998. Development of mechanistic models for the prediction
of machining performance: Model-building methodology, pp. 109–120. In Proceedings of the
International Workshop on Modelling of Machining Operations, Atlanta, GA, May 19.
Kim, J., Kang, M., Ryu, B., and Ji, Y., Development of an on-line tool-life monitoring system using
acoustic emission signals in gear shaping, Int. J. Mach. Tools Manuf., 39, 1761–1777, 1999.
Koizumi, T., Tsujiuchi, N., and Matsumura, Y., Diagnosis with the correlation integral in the time
domain, Mech. Syst. Signal Process., 14, 1003–1010, 2000.
Kopac, J., Influence of cutting material and coating on tool quality and tool life, J. Mater. Process.
Technol., 78, 95–103, 1998.
Kumar, S.A., Ravindra, H.V., and Srinivasa, Y.G., In-process tool wear monitoring through time series
modeling and pattern recognition, Int. J. Prod. Res., 35, 739–751, 1997.
Kuo, R.J. and Cohen, P.H., Intelligent tool wear estimation system through artificial neural networks and
fuzzy modelling, Artif. Intell. Eng., 12, 229–242, 1998.
Lago, L., Olsson, S., Hakansson, L., and Claesson, I. 2002. Design of an efficient chatter control system for
turning and boring applications, pp. 4–7. In Proceedings of the 20th International Modal Analysis
Conference (IMAC XX), Los Angeles, CA, February.
Lee, B.Y. and Tarng, Y.S., Milling cutter breakage detection by the discrete wavelet transform,
Mechatronics, 9, 225–234, 1999.
Lee, L.C., Lee, K.S., and Gan, C.S., On the correlation between dynamic cutting force and tool wear, Int.
J. Mach. Tools Manuf., 29, 295–303, 1989.
Lee, J.H., Kim, D.E., and Lee, S.J., Statistical analysis of cutting force ratios for flank-wear monitoring,
J. Mater. Process. Technol., 74, 104–114, 1998.
Li, X., A brief review: acoustic emission method for tool wear monitoring during turning, Int. J. Mach.
Tools Manuf., 42, 157–165, 2002.
Vibration and Shock Handbook26-1170
1072
1073
1074
1075
1076
1077
1078
1079
1080
1081
1082
1083
1084
1085
1086
1087
1088
1089
1090
1091
1092
1093
1094
1095
1096
1097
1098
1099
1100
1101
1102
1103
1104
1105
1106
1107
1108
1109
1110
1111
1112
1113
1114
1115
1116
1117
1118
1119
1120
1121
1122
Li, S. and Elbestawi, M.A., Fuzzy clustering for automated tool condition monitoring in machining,
Mech. Syst. Signal Process., 10, 533–550, 1996.
Li, X.Q., Wong, Y.S., and Nee, A.Y.C., Tool wear and chatter detection using the coherence function of
two crossed accelerations, Int. J. Mach. Tools Manuf., 37, 425–435, 1997.
Liu, Q. and Altintas, Y., On-line monitoring of flank wear in turning with multilayerd feed-forward
neural network, Int. J. Mach. Tools Manuf., 39, 1945–1959, 1999.
Lou, K. and Lin, C., An intelligent sensor fusion system for tool monitoring on a machining centre, Int.
J. Adv. Manuf. Technol., 13, 556–565, 1997.
Luo, G., Osypiw, D., and Irle, M. 2002. Tool wear monitoring by on-line vibration analysis with
wavelet algorithm. In Metal Cutting and High Speed Machining, pp. 393–405, Kluwer Academic/
Plenum Publishers, New York.
Mdlazi, L., Marwala, T., Stander, C.J., Scheffer, C., and Heyns, P.S. 2003. The principal component
analysis and automatic relevance determination for fault identification in structures.
In Proceedings of the 21st International Modal Analysis Conference (IMAC), Kissimmee, FL,
Paper 37.
Merchant, M.E., Mechanics of the cutting process, J. Appl. Phys., 16, 318–324, 1945.
Morimoto, Y., Ichida, Y., and Sata, R. 2000. Excitation technique by 2-axes shaker of an CNC lathe,
pp. 1643–1648. In Proceedings of the 18th International Conference on Modal Analysis, San Antonio,
TX.
Pai, P.S., Nagabhushana, T.N., and Rao, P.K.R., Tool wear estimation using resource allocation network,
Int. J. Mach. Tools Manuf., 41, 673–685, 2001.
Ruiz, A., Guinea, D., Barrios, L.J., and Betancourt, F., An empirical multi-sensor estimimation of tool
wear, Mech. Syst. Signal Process., 7, 105–199, 1993.
Scheffer, C. 1999. Monitoring of tool wear in turning operations using vibration measurements, Masters
dissertation (MEng), Department of Mechanical and Aeronautical Engineering, University of
Pretoria, South Africa.
Scheffer, C. 2002. Development of a tool wear monitoring system for turning using artificial intelligence,
Ph.D. thesis, Department of Mechanical and Aeronautical Engineering, University of Pretoria,
South Africa.
Scheffer, C. and Heyns, P.S. 2000a. Synthetic diamond tool wear monitoring using vibration
measurements, pp. 245–251. In Proceedings of the 18th International Modal Analysis Conference,
San Antonio, TX, 7–10 February.
Scheffer, C. and Heyns, P.S. 2000b. Development of an adaptable tool condition monitoring system,
pp. 361–370. In Proceedings of the 13th International Congress on Condition Monitoring and
Diagnostic Engineering Management (COMADEM 2000), Houston, TX, 3–8 December.
Scheffer, C. and Heyns, P.S. 2001a. Tool condition monitoring systems — an overview, pp. 316–323.
International Conference on Competitive Manufacturing (COMA ’01), Stellenbosch, South Africa,
31 January–2 February.
Scheffer, C. and Heyns, P.S., Wear monitoring in turning operations using vibration and strain
measurements, Mech. Syst. Signal Process., 15, 1185–1202, 2001b.
Scheffer, C. and Heyns, P.S. 2002a. A robust and cost-effective system for conducting cutting experiments
in a production environment, pp. 329–334. In Proceedings of 3rd CIRP International Conference
on Intelligent Computation in Manufacturing Engineering (ICME 2002), Ischia (Naples), Italy,
3–5 July.
Scheffer, C. and Heyns, P.S. 2002b. Neural Network approaches for sensor-based tool wear monitoring, In
Proceedings of Metalworking Tools & Fluids, South African Institute of Tribology, Johannesburg,
South Africa, 7 November.
Scheffer, C. and Heyns, P.S., An effective tool condition monitoring system for turning, Mech. Syst. Signal
Process., 2004.Q6
Scheffer, C., Kratz, H., Heyns, P.S., and Klocke, F., Development of a tool wear monitoring system for
hard turning, Int. J. Mach. Tools Manuf., 43, 973–985, 2003.
Vibration-Based Tool Condition Monitoring Systems 26-1171
1123
1124
1125
1126
1127
1128
1129
1130
1131
1132
1133
1134
1135
1136
1137
1138
1139
1140
1141
1142
1143
1144
1145
1146
1147
1148
1149
1150
1151
1152
1153
1154
1155
1156
1157
1158
1159
1160
1161
1162
1163
1164
1165
1166
1167
1168
1169
1170
1171
1172
1173
Sick, B., On-line tool wear monitoring in turning using neural networks, Neural Comput. Appl., 7,
356–366, 1998.
Sick, B., Online and indirect tool wear monitoring in turning with artificial neural networks: a review of
more than a decade of research, Mech. Syst. Signal Process., 16, 487–546, 2002.
Sick, B. and Sicheneder, A. 1997. Time-delay neural networks for on-line tool wear classification and
estimation in turning, pp. 461–466. In Proceedings of the Third Conference on Neural Networks and
Their Applications, Kule, Poland, 14–18 October.
Sick, B., Sicheneder, A., and Lindinger, H. 1998. A comparative evaluation of different neural network
paradigms for tool wear classification in turning, pp. 139–146. In Proceedings of the 3rd
International Workshop Neural Networks in Applications (NN ’98), University of Magdeburg,
Germany, 12–13 February.
Silva, R.J., Rueben, R.L., Baker, K.J., and Wilcox, S.J., Tool wear monitoring of turning operations by
neural network and expert system classification of a feature set generated form multiple sensors,
Mech. Syst. Signal Process., 12, 319–332, 1998.
Silva, R.J., Baker, K.J., Wilcox, S.J., and Reuben, R.L., The adaptability of a tool wear monitoring system
under changing cutting conditions, Mech. Syst. Signal Process., 14, 287–298, 2000.
Tarmal, G.J. and Opavsky, P. 2000. Signal processing in measurement of milling forces, pp. 389–397. In
Proceedings of the 13th International Congress on Condition Monitoring and Diagnostic Engineering
Management (COMADEM 2000), Houston, TX.
Teti, R., A review of tool condition monitoring literature database, Ann. CIRP, 44, 659–667, 1995.
Venkatesh, K., Zhou, M., and Caudill, R.J., Design of artificial neural networks for tool wear monitoring,
J. Intell. Manuf., 8, 215–226, 1997.
Warnecke, G. and Siems, S. 2002. Dynamics in high speed machining. Metal Cutting and High Speed
Machining, pp. 21–30, Kluwer Academic/Plenum Publishers, New York.
Xiaoli, L., Yingxue, Y., and Zhejun, Y., On-line tool condition neural network with improved fuzzy neural
network, High Technol. Lett., 3, 30–38, 1997.
Yao, Y., Fang, X.D., and Arndt, G., Comprehensive tool wear estimation in finish-machining via
multivariate time-series analysis of 3-D cutting forces, Ann. CIRP, 39, 57–60, 1990.
Zawada-Tomkiewicz, A., Classifying the wear of turning tools with neural networks, J. Mater. Process.
Technol., 109, 300–304, 2001.
Zhou, Q., Hong, G.S., and Rahman, M., A new tool life criterion for tool condition monitoring using a
neural network, Eng. Appl. Artif. Intell., 8, 579–588, 1995.
Vibration and Shock Handbook26-1172
1174
1175
1176
1177
1178
1179
1180
1181
1182
1183
1184
1185
1186
1187
1188
1189
1190
1191
1192
1193
1194
1195
1196
1197
1198
1199
1200
1201
1202
1203
1204
1205
1206
1207
1208
1209
1210
1211
1212
1213
1214
1215
1216
1217
1218
1219
1220
1221
1222
1223
1224
Author QueriesJOB NUMBER: 8568
CHAPTER: Vibration-Based Tool Condition Monitoring Systems
Q1 Kindly note that Dimla (2001) is not present in the reference list, please add or delete.
Q2 Please check: the percentages only total 80%.
Q3 Kindly note that Bonifacio et al. (1994) has been changed to Bonifacio and Diniz, 1994 as per
the reference list, please check.
Q4 Please check change to "AERMS".
Q5 Kindly note that Li and Tzeng (2000) has been changed to James and Tzeng, 2000 to match with
the reference list, please check.
Q6 Kindly update Scheffer and Heyns, 2004.
Q7 Please check the permissions for figures 4, 17, 18, 19, 21, 23 and 24. Please check if permission is
required. If so, then whether the information is correct.
1225
1226
1227
1228
1229
1230
1231
1232
1233
1234
1235
1236
1237
1238
1239
1240
1241
1242
1243
1244
1245
1246
1247
1248
1249
1250
1251
1252
1253
1254
1255
1256
1257
1258
1259
1260
1261
1262
1263
1264
1265
1266
1267
1268
1269
1270
1271
1272
1273
1274
1275