6

CHAPTER 2

LITERATURE REVIEW OF HARD TURNING

The process of hard turning has been studied theoretically and

experimentally for over 40 years (Tonshoff et al 2000). In the late 60s itself,

Shaw and Nakayama (1967) studied the machining characteristics of high

strength materials such as titanium alloy, beryllium, work hardened steel and

refractory materials. While machining these materials, a lower feed and speed

were recommended, since they imposed a large cutting force and high cutting

temperature. Moreover it was identified that the machining of a hard material

required high rigidity and high stability machine tools as they operate at high

energy levels.

In this chapter, the state of research in hard turning and its

monitoring is presented, under three broad headings, namely (i) The science

and technology of hard turning (ii) The various monitoring techniques that

were developed for the process monitoring of hard turning and lastly (iii)

Certain specific acoustic emission based monitoring techniques and signal

processing, employed to monitor the conventional turning operations, as it is

intended to use these techniques for monitoring hard turning.

2.1 HARD TURNING PROCESS

The various studies conducted to technically establish the

technology of hard turning and to understand the hard turning process, include

studies on developing suitable tool materials, tool geometry and tool edge

geometry, and coatings for enhanced tool life, identifying optimal cutting

7

conditions, understanding the process of chip formation and identifying the

various tool wear mechanisms.

2.1.1 Development of Suitable Tool Material

Cutting tools used for hard turning require high hardness, high

compressive strength, high resistance to abrasive wear, thermal resistance and

chemical stability at elevated temperatures. Cubic Boron Nitride (CBN) is the

preferred cutting tool material rather than poly crystalline diamond (PCD) for

machining hardened steel since it has very high hardness and hot hardness and

is chemically inert particularly in the presence of carbon absorbing materials.

Ceramic inserts are also being used for hard turning. Hodgson and Trendler

(1981) carried out hard turning experiments to identify a suitable cutting tool

material for machining hardened steel. In this study, they performed the

turning tests on three different varieties of hardened steels, such as cold work

steels (D2 and D6), HSS steel (M2) using CBN and ceramic cutting tools. The

ceramic cutting tool did not perform satisfactorily, while machining cold

worked steel (D6) resulting in gross fracture and chipping of the cutting edge.

On the other hand, CBN inserts performed better while machining hardened

steel, in terms of extended tool life even at higher cutting speeds.

Konig et al (1990) conducted hard machining experiments in

turning, milling and drilling using CBN tool inserts. In the turning of hard

materials, such as high speed steel and bearing steel, they used both the CBN

and mixed ceramics, to study the wear behavior and tool life. Even though the

CBN and ceramics were tool materials frequently used for machining

hardened steel, CBN had emerged a preferred tool material due to the slow

and uniform tool wear characteristics, compared to ceramic tool inserts

(Figure 2.1). From the surface finish point of view also, the CBN tools were

found to perform better.

8

Figure 2.1 Tool Wear in Hard Turning for CBN and Mixed Ceramic

Tool Inserts (Konig et al., 1990)

The performance of the CBN tools depends on the CBN content,

grain size, bonding material, and microstructure. In general, CBN tools can be

classified into two types. The first type contains very high CBN content with

more than 90% by volume of CBN, in which the CBN grains are directly

bonded by themselves or with a metallic binder. The second type contains a

lower CBN content, in the range of 50 to 70% by volume of CBN in which

the CBN grains are bonded by a ceramic binder such as TiC or TiN.

Kevin Chou et al (2002) experimentally investigated the wear

behavior of low and high CBN tool material. The authors performed hard

turning experiments on AISI 52100 steel material using high and low CBN

cutting tool material. The tool wear examination showed that low CBN

content tools generate a better surface finish and have lower flank wear rates

than their high CBN counterparts, even though low CBN tool material had

inferior mechanical properties compared to a high CBN tool. The

performances of the low CBN content tools were found to be better at higher

cutting speeds. The depth of cut was also found to have a minor effect on the

9

tool wear. The performance of the CBN tool was found to depend not just on

the gross mechanical properties of the tool material but the combined effect of

gross mechanical property and the microstructure properties of CBN.

2.1.2 Tool Geometry and Tool Edge Geometry

In 1980’s investigations were conducted for developing a suitable

tool geometry and tool edge geometry, to improve the tool life of the CBN

inserts. Hodgson and Trendler (1981) conducted experiments in hard turning,

to study the effect of rake angle and edge preparation of the CBN tools on

tool wear. The study showed that the negative rake angle performed better

than the positive or zero rake angle from the tool life point of view.

Providing a negative rake angle also increased the wedge strength,

thereby preventing the chipping of the cutting edge. In hard turning, as the un-

deformed chip thickness, which is the feed, is very small, the actual cutting is

carried out mostly at the chamfered region of the cutting edge. However, a

larger negative rake increases the thrust force drastically, and this was shown

in the experimental results of Nakayama et al (1988).

The influence of the tool nose radius on the surface roughness,

cutting forces, tool wear and white layer formation had been systematically

investigated by Kevin Chou and Husi song (2004). The authors conducted

finish hard turning experiments on hardened AISI 52100 steel, using mixed

ceramic tool inserts. Inserts with nose radii of 0.8, 1.6 and 2.4 mm were used

in the experiments. The experiments revealed that the larger nose radius to

contributed to a better surface finish, but resulted in increased cutting forces

and increased specific cutting energy. Along with the lower nose radius, worn

out tool wear and higher feed rates were also found to have a strong impact on

the higher depth of the white layer formation.

10

Since hard turning involves higher cutting forces and bigger shock

loads, the chamfered cutting edge of CBN insert was used. The chamfered

cutting edges had shown much higher tool life than the sharp cutting edge,

though they led to an increase in the main cutting forces. Tugur Ozel et al

(2005) performed the hard turning experiments on hardened AISI H13 steel,

using CBN inserts to study the effect of tool edge geometry, workpiece

hardness, feed rate and cutting speed on surface roughness and cutting force.

Their objective was to identify suitable edge geometry, for a better surface

finish and improved wedge strength. In the study sharp, chamfered and honed

edges were used. The honed edges were found to perform much better than

the sharp and chamfered edges in terms of surface finish, radial cutting force

and tangential cutting force.

Lalwani et al (2008) presented the effects of cutting conditions on

surface roughness. They have carried out hard turning experiments and

processed the data using response surface methodology and face-centered

central composite design approach. Their non-linear quadratic model surface

roughness showed the feed rate was the primary contributor and interactional

effect of feed and depth of cut, cutting speed and depth of cut to be the

secondary contributor for affecting the surface roughness of workpiece.

Tugrul Ozel et al (2008) introduced a ‘variable edge design’ on the

cutting edge of CBN insert and experimentally studied its influence on cutting

force, temperature distribution and tool wear. The results of experimentally

obtained cutting forces were used to validate an FEA model. The

investigation of the experiments concluded that variable edge tool inserts are

better than the normally prepared edge in terms of reduced tool wear, less heat

generation and less induced plastic strain on the work material during hard

turning.

11

2.1.3 Tool Coating

Reginaldo et al (2007) undertook studies to report the wear

characteristics of a coated PCBN insert while machining hardened steel.

Three different TiN and TiC-based coatings were used on the PCBN, to study

the wear behavior and surface finish of the work piece. Finite element

simulation model was used to estimate chip-tool interface temperature for

studying thermal resistance of the coatings. The results show that the coatings

acted as a thermal barrier between the work piece and PCBN tool, resulting in

a small reduction at the substrate temperature.

Youngsik Choi and Richard Liu (2009) used micro / nano CBN

particle coated tungsten carbide tools in hard turning. The coated tools with a

CBN grain size of less than 0.5 m were observed to produce more uniform

residual compressive stresses on the machined workpiece. This resulted in an

enhanced fatigue life of the component.

2.1.4 Identifying Optimal Cutting Conditions

Hard turning is usually carried out at low feeds and depth of cuts

with normal cutting speeds without applying any coolant. Tugrul Ozel and

Karpat (2005) attempted to predict the surface roughness and tool flank wear

using regression and artificial neural network models. The predictive models

so developed suggested the surface roughness of the work piece to depend

upon primarily, the hardness of work material, edge preparation of insert, feed

and cutting speed. The surface finish of the work piece was found to improve

by the use of chamfered and honed edges, and with the reduction in cutting

speed and feed and with increased hardness in work material. But the tool

wear was found to increase with the hardness of the work material.

12

Experimental work carried out by Neo et al (2003) and Mohamed

Athmane Yallese et al (2009) showed the surface roughness generated in

finish hard turning to be mainly influenced by the cutting parameters such as

cutting speed, cutting feed and depth of cut. Among these cutting parameters,

cutting feed is the dominant factor contributing to the surface roughness. In

their experimental study the authors had shown the cutting forces to increase

rapidly with increase in feed rate which in turn affected the stability of the

machine tool. The study also showed that cutting speed beyond 280 m/min to

produce sparks which resulted in very poor surface finish and very rapid tool

wear. The authors pointed out that the use of recommended range of cutting

speed and cutting feed for better surface quality and for extended tool life.

In the hard-turning process, tool geometry and cutting conditions

determine the time and cost of production, which ultimately affect the quality

of the final product. Optimization of the machining process not only improved

the overall machining economics, but also the product quality to a great

extent. In this context, Dibag singh et al (2007a) made an effort to estimate

the optimum tool geometry and machining conditions using genetic

algorithms to get the best possible surface finish within the constraints. A

surface roughness model was developed using response surface methodology

by incorporating nose radius, effective negative rake angle, cutting speed and

feed on machining of bearing steel. The typical optimum machining

conditions obtained were 200 m/min cutting speed, 0.1 mm feed, 6° effective

rake angle, and 1.2 mm nose radius, leading to a minimum surface roughness

value.

The cost of manufacturing finished component is directly related

machining time and how long the tool can be used for continuous machining

with the pre-requisite surface quality. Hence the deeper understanding of tool

13

wear behavior, tool life, and surface finish by selecting appropriate cutting

condition is the important criteria for the machining of hard material.

Therefore the machinability characteristics of work-tool combination are

required for all possible combinations and this is truer for the hard machining.

Paulo Davim and Luis Figueria (2007) used the orthogonal array experiments

to determine the influence of cutting speed, feed and cutting time on the

specific pressure and surface roughness and tool flank wear. The authors were

carrying out experiments on cold work tool steel D2 (AISI) using ceramic

cutting tool. The study showed the tool wear to be primarily influenced by

cutting speed and the surface roughness to be influenced by feed and cutting

time which is an indirect measure of tool wear.

Hard turning is usually performed without any cutting fluid as dry

turning. However Vikram Kumar and Ramamoorthy (2007) had conducted

experiments with a minimum fluid application technique to result in a

sustainable hard turning process. The authors had reported a reduction in

cutting force, temperature and surface roughness.

2.1.5 Cutting Force and Temperature

When turning a soft steel material by feeding the cutting tool along

the workpiece axial direction, the principal (tangential) cutting force is the

largest among the three force components. However, because of the small

depth of cut and large nose radius commonly used in hard turning, chip

formation takes place only near the nose radius, and on the chamfer or hone

of the cutting tool. As a result, the thrust (radial) force component can

increase drastically and become the largest force component as the tool wear

progresses.

14

Figure 2.2 Cutting Geometry and Cutting Forces in Hard Turning

(Tonshoff at al 2000)

Figure 2.2 shows the typical results of cutting forces during hard

turning (Tonshoff 2000). The experimental results clearly show the radial

force component (FP) to be the largest, especially after a long distance of cut.

A relatively steep rise in the radial force (FP) can be observed, compared with

the other two force components, due to a progressive increase in the tool flank

wear. Young-woo park (2002) had also reported the radial force during hard

machining to be the largest among the cutting forces, regardless of cutting

conditions and tool material.

The wear behavior of the CBN in machining hardened steel is

determined by the workpiece hardness and cutting temperature. During hard

machining, the energy consumed by the process is largely converted into heat.

The generation of heat during machining will increase the temperature in the

cutting zone. A higher cutting temperature creates a softening of the work

material, but it promotes the accelerated diffusion wear rates on tool material.

15

Liu et al (2002) studied the performance of poly crystalline Cubic

Boron Nitride (PCBN) tool in finish hard turning of GCr15 bearing steel with

different hardness values. They examined the influence of work material

hardness and cutting conditions on the cutting temperature and tool wear. A

natural thermocouple was used to measure tool temperature during

machining. The cutting temperature was found to increase along with the

workpiece hardness till the workpiece hardness reached a value of 50 HRC

and then as the hardness was further increased the temperature was found to

reduce. This was attributed to material softening at elevated temperatures. The

increase in cutting speed and feed was also found to increase the tool

temperature.

Ren et al (2004) had conducted experiments by hard turning the

hard-faced high chromium layers using PCBN tool material. The temperature

at the tool - shim interface was measured using thermocouple and the

temperature at the wear zone was estimated using finite element analysis. The

cutting temperature was found increase with increasing cutting speed and

feed.

2.1.6 Mechanism of Chip Formation

In hard turning a localized plastic deformation was found to occur at

the primary shear zone (Nakayama et al 1988). Since the work material didn’t

have enough ductility, the crack was initiated at the surface where there was

no hydrostatic pressure. This was the primary reason identified for the

formation of saw-toothed chips, observed in the cutting experiments of

hardened 70-30 brass materials.

The high brittleness of the work material prevented the plastic flow

under the high compressive stress and led to the formation of a crack. The

crack enabled the release of the pent up energy and provided a sliding surface

16

for the segmented material. Simultaneously the plastic deformation and the

heating of work material occurred at the vicinity of the cutting edge of the

tool, facilitating the sliding of chip segment. This cycle repeated continuously

leading to the severely segmented chip formation (Konig et al 1990).

Through experimental investigations Shaw et al (1993, 1998)

showed that saw toothed, wavy and segmented chips were formed in

machining of hardened steel, due to catastrophic shear and plastic flow. Rapid

tool wear is the basic difficulty in hard turning and wear significantly affects

on the quality of surface finish.

Davies et al (1997) conducted the experiments to study the

mechanism of segmented chip formation in AISI 52100 bearing steel and

electro plated nickel-phosphorous on copper substrates using CBN tool with

relatively low cutting speed. The average spacing of shear band in the chip

during machining was found to increase with the increasing cutting speed and

reached a limiting value determined by the cutting feed and depth of cut and

work material properties. The authors used a simple one dimensional thermo

mechanical model of a continuous homogeneous material getting sheared by

an impinging rigid wedge to explain this behavior.

2.1.7 Tool Wear Mechanism

Since rapid tool wear is the basic problem in hard turning, efforts

are being made to understand the wear mechanism operating between the tool

and work material, in conditions prevailing during hard turning. One major

difficulty in turning of hardened steels is the tool-wear caused by the hardness

of the material. The wear behavior of the CBN is influenced by many factors,

such as the composition of the CBN material, work piece hardness, cutting

condition and nature of cutting operation etc. The experimental investigation

of Kevin Chou et al (1997) showed the wear rate of the CBN tool to depend

17

upon the carbide particle in the work material and size of the CBN grain in

tool. The results of the experiments clearly indicated the wear rate to be

directly proportional to the size of the carbide particle in the work material

and inversely proportional to the size of the CBN grains.

Finish hard turning experiments were carried out by Penalva et al

(2002) to understand the effect of tool wear on surface roughness. Finish

turning was performed on AISI 52100 hardened steel using CBN insert. The

examination of tool wear profile and surface roughness profiles showed the

replication of worn-out cutting edge on the surface of the work piece.

The tool wear of a cutting tool is also dependent on the type of

hardened steel being machined, namely the physical and chemical interactions

between the workpiece material and the tool. Tool and die steels typically

contain a variety of carbide particles based on alloying elements like tungsten,

molybdenum, vanadium and chromium that can greatly accelerate abrasive

wear.

Gerard Poulachon et al (2003) carried out hard turning experiments

on specimens of hardened steel having different material composition but

having same hardness to understand wear behavior of CBN. Dry turning

experiments were conducted on four different hardened steels namely

X155CrMoV12 cold work steel (AISI D2), X38CrMoV5 (AISI H11) hot

work steel, 35NiMo16 hot work steel and 100Cr6 bearing steel (AISI

52100).Their results suggested the various carbides present in steel such as

primary carbides M7C3 and secondary carbides M3C in the microstructure to

influence the tool wear. Higher tool wear rates were attributed to the presence

of high quantities of primary carbides in steels.

An extensive literature review was presented by Cora Lahiff et al

(2007) on the wear modes and wear mechanism of various grades of CBN

18

inserts. Tool wear mechanism is a complex phenomena and no single wear

mechanism can satisfactorily provide a full explanation. The authors

identified two body abrasion caused by hard carbide particles and martensite

in the workpiece, three body abrasion caused by the loosened CBN grains,

diffusion wear and adhesion because of high temperature to be the

predominant wear modes. They also observed the tool wear to depend on the

composition CBN content and binder, tool edge geometry and machine tool

stability.

Yong Huang et al (2005) had modeled the depth of crater wear as a

function of cutting temperature, stress and other cutting conditions taking in

to consideration tool work material properties. The model predicted the crater

depth and this predicted crater depth was validated experimentally. According

to the authors modeling of crater depth is important as increased crater depth

will weaken the cutting edge leading to micro chipping and catastrophic

fracture of the cutting edge. The authors claim this model to help the user to

select the cutting conditions and tool manufacturer to optimize the tool

geometry for improved wear resistance.

Lin et al (2008) in their experimental study on hard turning

machined AISI 4340 alloy steel using cubic boron nitride tool. They used

CBN inserts having 45- 50 % of CBN and TiC based binder. The experiments

were conducted at various cutting speeds in the range of 58 to 180 m/min. In

all the experiments the feed rate and depth of cut remained constant at 0.1

mm/min and 0.2 mm respectively. An infrared photography unit was mounted

on side of the carriage adjacent to tool tip to measure the temperature of the

back side chip. In their experiments cutting forces were measured and tool

wear was studied using SEM analysis. During the experiments the authors

observed the average cutting temperature to be low at very low cutting speeds

and it gradually increased with increase in cutting speed. But the Principal

19

cutting force decreased as the cutting speed was increased. In low speed

cutting, the binder of the hard particles of the cutting tool was observed to be

removed from the substrate due to high cutting force resulting from low

cutting temperature.

2.1.8 Surface Roughness Model

Since hard turning is the finishing process, the quality of the surface

finish is the most expected outcome of the process. The surface finish in hard

turning is influenced by the cutting conditions, tool geometry, edge

preparation, tool wear and rigidity of the machine tool. The generation of the

surface roughness in turning process has been modeled for several decades.

Many researchers had developed theoretical and empirical models to predict

the surface roughness (Tugrul Ozel and Karpet 2005 and Lalwani et al 2008).

Knuefermann and McKeown (2004) attempted to develop a

numerical model of surface roughness in hard turning based on variables such

as machine tool vibration, tool geometry, material deformation, chip removal

and tool path. In their model, the authors employed the material partition

equation for defining chip removal. Cutting experiments were carried out in

an ultra precision lathe using CBN tool to compare measured surface

roughness with simulated roughness. The analysis of results found the

simulated surface roughness to be fairly comparable with measured

roughness.

2.2 MONITORING HARD TURNING

As hard turning is an advanced manufacturing process close

monitoring and control of the process is essential. Tool wear is a complex

phenomenon occurring due to different mechanisms and is stochastic. In

general, a worn tool affects the surface quality of the work material, and

20

therefore there is a need to alert the operator to stop further machining to

avoid undesirable consequences. The modern machine tool has many sensing

systems as an integral part of it. Using the data from these sensors to obtain

useful information about the process is an important domain of current

research. One of the major issues that need to be addressed is evolving an

online, reliable and robust monitoring methodology, is to predict the process

condition in hard turning.

2.2.1 Monitoring Strategies

The basic strategy of process monitoring can be classified into (i)

signal based method (ii) Model based method and (iii) classifying method,

depending on the complexity of manufacturing process (Tonshoff et al.,

2000a)

Figure 2.3 Classification of Monitoring methods (Tonshoff et al 2000a)

The three approaches are schematically explained in Figure 2.3.

Any monitoring system attempts to measure the condition of a machine tool

or of the process itself. In the signal based method, the measured signal values

are compared with the predefined signal or the ranges of signals. In model-

based monitoring, an empirical model of the process is built, and this is

21

compared with actual experimentally measured quantities, to detect any

process deviation. In a classifier based monitoring system, a feature vector is

identified which is used for subsequent discrimination. Depending on the

actual system being monitored, a suitable strategy needs to be identified and

used.

2.2.2 Monitoring using Vibration Signal

Figure 2.4 On-Line Roughness Measuring Technique using Cutting

Vibration in Hard Turning (Dong Young et al 1996)

Dong Young Jang et al (1996) conducted hard turning experiments

to understand the correlation between the cutting tool vibration and the

surface roughness of the work material in hard turning. The experimental

setup is shown in Figure 2.4. The relative vibration between the tool and the

work piece was measured using an inductive pickup. This vibration signal

was used to predict the surface roughness. The predicted surface roughness

was found to correlate well with the actually measured roughness values.

22

Thus, the authors had proposed an online surface roughness monitoring

system using an inductive vibration sensor.

2.2.3 Monitoring using Cutting Forces and Acoustic Emission

Scheffer et al (2003) proposed a tool wear monitoring system for

hard turning using a three-directional cutting force dynamometer, an AE

sensor and a temperature sensor. The authors formulated a monitoring

methodology using artificial intelligence (AI) techniques for monitoring tool

flank and crater wear of CBN insert. Self-Organizing Map analysis was used

to classify and isolate some of the common noises present during the process

such as floor vibration, work piece clamping and temperature conditions.

Zhou et al (2003) measured the cutting force in the radial direction

of the workpiece using a Kistler three component dynamometer and found it

to be sensitive to tool wear in hard turning process.

Figure 2.5 A Proposed ANN Model for Tool Wear Monitoring

(Zhou et al 2003)

23

In hard turning the radial force is much higher than feed force and

the cutting force due to very small cutting depth, very small feed rate and

negative rake angle of the tool. From the experiments, the authors observed

both frequency and amplitude to show a tendency to increase as the flank

wear progressed The authors had developed an ANN based wear prediction

model using radial force, frequency energy and accumulated cutting time as

inputs The ANN architecture is shown in Figure 2.5. The predicted wear was

corrected using tool holder temperature. The predicted wear was observed to

be very close to the measured flank wear values.

Mehdi Remadna and Rigal (2006) had conducted turning

experiments on tempered steel of hardness 52 HRC using CBN inserts and

monitored the three component cutting forces. The authors observed the

specific cutting force measured along the radial direction of the workpiece to

correlate well with the tool wear and surface roughness.

An online process monitoring system for hard turning using three-

component force sensors was developed by Dongfeng Shi and Gindy (2007).

Experiments were carried out for machining Inconel 718 using ceramic tools.

The static and dynamic components of the forces were measured using

multiple sensors and a power sensor was used to trigger the data acquisition.

The acquired signal was subsequently processed using wavelet transforms.

This signal processing enabled to distinguish between static and dynamic

components, and to obtain features of tool malfunctions such as tool wear,

tool chipping and tool breakage. The analysis of force signals showed the

gradual wear of the cutting tool to be related to the static force components

and the transient occurrences like tool chipping got revealed in the dynamic

components.

Acoustic emission based online tool failure detection system waspresented by Liao et al (1995) in face milling of high chromium material

24

using PCBN tool insert. The experiments were carried out to extract thefailure features from AE count rate, AE peak and AERMS. During the cuttingexperiments the authors observed the pattern of AERMS to change verysharply beyond the point of tool fracture. Further the authors proposed amoving range control chart for online tool failure monitoring in hard millingoperation.

Guo and Ammula (2005) had attempted to monitor the hard turningprocess of AISI 52100 steel of about 62 HRC turned with CBN inserts usingacoustic emission monitoring. The authors had attempted to monitor primarilythe surface integrity which was measured in terms of (i) The thickness of thewhite layer formed and (ii) The surface roughness. The authors had attributedboth the increase in the thickness of the white layer formation and thedeterioration in the surface roughness to the increased flank wear and theassociated increased rubbing between the flank and the workpiece. Theauthors had measured the conventional Acoustic Emission parameters likeAERMS, frequency and count rate and observed them to correlate well with thethickness of the white layer and surface roughness. The sharp tool, because oflow contact area with the workpiece was observed to produce less dampingand increased AERMS. As the tool wears off the contact area between the toolflank and the workpiece was found to increase resulting in relativelyincreased damping between the two leading to lowering of AERMS. But whenthe tool flank wear land crossed the limit the increased rubbing was found toproduce highly brittle white layer. The brittle white layer was observed tocontribute low damping leading to increased AERMS values.

2.3 ACOUSTIC EMISSION BASED MONITORING INTURNING

Tool wear in conventional machining can be monitored usingvarious indirect variables that are sensitive to tool wear. The frequently usedindirect parameters are the three components of cutting force, Acoustic

25

Emission (AE) signal, motor current and vibration Teti et al (2010). Unlikedirect methods these methods can be used for on-line tool wear monitoring

AE signals are transient elastic strain waves, generated duringplastic deformation, and by disturbances occurring at the level of the atomicstructure. Plastic deformation, fracture, wear and rubbing are all importantsources of acoustic emission signals.

2.3.1 Sources of AE in Metal Cutting

The AE monitoring is one of the most effective methods for processmonitoring the conventional metal cutting operations (Dornfeld 1992). Themajor advantage of using the AE to monitor the machining operation is that,the frequency range of the AE signal is much higher than environmental noisein a non intrusive process. The friction between the cutting tool andworkpiece generates a continuous AE signal, which gives rich information onthe cutting process. In process monitoring, using acoustic emission, if aspecific threshold and band width can be established for a particular processconfiguration, then the AE signal and AERMS may be effectively used formonitoring the event of tool wear (Dornfeld and Asibu 1980). The threemajor sources of the AE signal in metal cutting process are (Dornfeld 1992):

(i) Continuous deformation in the primary shear zone andfracture of work material in the secondary, tertiary shear zone.

(ii) Fracture of the cutting tool between chip-tool and tool-workpiece interfaces.

(iii) Collision, entangling and breakage of chips.

An AE signal can be classified into two types, namely continuousand burst AE signal. Continuous signals are generated with shearing in theprimary zone and wear on the tool flank. The burst types are observed duringcrack growth in the material, tool fracture and chip breakage. The importanceof AE signal processing is to eliminate unwanted noise and to extract feature

26

signals, which can be used to correlate the process parameters. To effectivelydetect tool wear and fracture of the cutting tool one has to understand thethree basic elements of the monitoring system, they are compatibility of theAE signal, AE sensor and the pre amplifier (Krzysztof Jemielniak 2001).

2.3.2 AE Based Tool Condition Monitoring

A brief review on AE based tool wear monitoring methodology inturning was presented by Xiaoli Li (2002). In his review, the author haspresented the state of research on various AE signal processing techniquesand feature extraction. The various AE signal processing techniques widelyused include time domain analysis, Fast Fourier Transform and wavelettransform. For feature extraction and tool wear estimation, techniques likepattern classification, group method of data handling (GMDH) methodology,fuzzy classifier, and neural network with data fusion technology are beingused. Quantifiable characteristics such as (i) ring down count, (ii) AE event,(iii) rise time, (iv) peak amplitude, and (v) RMS voltage can also be used tocharacterize the wear of the tool insert. According to the author, careful signalprocessing and integrations with other sensors will be an effective approachfor AE-based tool condition monitoring.

Dornfeld and Asibu (1980) conducted the turning experiments usingHSS cutting tool. They have presented a brief report about the generation ofAE during metal cutting process. The objective of the study was to establishthe relationship between AE root mean square (AERMS) value with the metalcutting parameters such as strain rate, cutting speed and feed in turningoperation and to validate this relationship by experimental investigation. Theexperimental data was presented as plots between AERMS signal and cuttingparameters for two different tool rake angles. They observed the strain rate tobe influenced by the cutting speed and this in turn increased RMS voltage.They have also identified the RMS of the AE signal suitable for processmonitoring in turning.

27

2.3.3 Traditional quantitative measurement in AE signal

The most widely used signal measurement parameters are count,peak amplitude, duration and rise time. The figure2.5 illustrates theseparameter under filtered signal envelop. These parameters detected andcaptured through AE sensor and digital storage oscilloscope in terms ofapplied voltage with respect to time. The terminology of these terms asexplained as follows.

Peak amplitude - The peak voltage of AE signalDuration - The time from the first threshold crossing to

the end of the last threshold crossingCount - The number of AE signal exceeds thresholdCount rate - Number of count per unit timeRise time - The time from the first threshold crossing to

the maximum amplitude.

Figure 2.6 Characteristics of Acoustic emission signal (Robert and

Talebzadeh, 2003)

28

2.3.4 Acoustic Emission Distribution Parameters

One of the methods of processing AE signals, is by using

parameters like the ring down count and total count (Teti and Micheletti,

1989), as metrics for monitoring tool wear. There are several techniques to

extract the features in the time domain AE signal. The commonly applied

techniques are (i) using measures of central tendency like the mean, median,

mode and root mean square value (ii) using measures of dispersion like

standard deviation or variance and (iii) using higher order moments, like the

Skew and Kurtosis.

In their study on tool wear monitoring in conventional turning,

Kannatey-Asibu and Dornfeld (1982) had assumed -distribution to

characterize the AERMS signals and evaluated the central moments and

distribution function parameters. The skew is the normalized third order

central moment and kurtosis is the fourth order central moment of the

assumed -distribution function.

The probability density function of the AERMS can be expressed as

in equation (2.1)

r 1 s 1x (1 x)f (x)(r,s) (2.1)

Where (r, s) can be expressed as

1 r 1 s 1

0(r,s) x (1 x) dx (2.2)

The distribution parameters of the distribution of ‘r’ and ‘s’ are

given by the equations 2.3 and 2.4.

29

2 22r ( ) (2.3)

2 22

(1 )s ( ) (2.4)

The skew indicates the level of symmetry about its mean value and

the kurtosis indicates the sharpness of the peak of assumed -distribution. The

values of the skew (SB) and kurtosis (KB) can be obtained from the equations

2.5 and 2.6 given below.

12

B2(s r) r s 1)Sr s 2 rs (2.5)

2

B6{(r s) (r s 1) rs(r s 2)}K

rs(r s 2)(r s 3) (2.6)

In their study Kannatey-Asibu and Dornfeld (1982) had

demonstrated the possibility of using the distribution parameters of AERMS

signal, like skew and kurtosis, for effectively monitoring the tool wear status.

As can be seen in Figure 2.6, the skew was observed to decrease as the tool

flank wear increased. Also, the kurtosis was observed to increase as the tool

flank wear increased (Figure 2.7).

30

Figure 2.7 Skew of AERMS Vs Flank Wear in Conventional Turning

(Kannatey-Asibu and Dornfeld 1982)

Figure 2.8 Kurtosis of AERMS Vs Flank Wear in Conventional Turning

(Kannatey-Asibu and Dornfeld 1982)

31

Though various works had been carried out to monitor the process

of hard turning, the quest for a novel sensor monitoring system that is robust,

reconfigurable, reliable, intelligent and inexpensive, meeting the demands of

advanced manufacturing technology, continues.

2.4 OBJECTIVE AND SCOPE

The overall objective of this study is to develop a reliable process

monitoring technique for the hard turning process using acoustic emission

signals.

The hard turning of AISI – D3 steel using cubic boron nitride

inserts, is taken as the application domain.

2.5 CHAPTER SUMMARY

This chapter dealt with the literature review of hard turning process

and the various techniques adopted in monitoring it. Specific emphasis was

given in studying the influence of the various process parameters on hard

turning, mechanism of chip formation and tool wear. Section 2.2 of this

chapter dealt with the various monitoring strategies adopted in hard turning.

The application of acoustic emission monitoring in conventional turning was

also reviewed, as it is very relevant to the present work on monitoring hard

turning using acoustic emission signal. Section 2.4 indentified the objective

and scope of the study.