Case Study

Proc IMechE Part B:

J Engineering Manufacture
1–12
Experimental study on tool condition Ó IMechE 2015
Reprints and permissions:
monitoring in boring of AISI 316 sagepub.co.uk/journalsPermissions.nav
DOI: 10.1177/0954405414565139

stainless steel pib.sagepub.com

K Venkata Rao1, BSN Murthy2 and N Mohan Rao3

Abstract
In this article, effect of cutting parameters, namely, cutting speed, feed rate and tool nose radius, on the tool life was
examined experimentally. The experiments were performed on boring of AISI 316 steel with cemented carbide tool
inserts. Design of experiments was prepared, and eight experiments were performed with two levels of the cutting para-
meters. The effects of cutting parameters were analyzed by evaluating the amplitude of workpiece vibration, surface
roughness and volume of metal removed. The experimental data were taken for evaluation of tool life for a flank wear of
0.6 mm in all the eight trials. A laser Doppler vibrometer was used for online data acquisition of workpiece vibration,
and a high-speed fast Fourier transform analyzer was used to process the acousto-optic emission signals for the work-
piece vibration. Taguchi, analysis of variance and regression analysis methods were used to identify significant cutting
parameters affecting the workpiece vibrations, surface roughness and volume of metal removed. All selected analysis
methods used in this study predicted similar cutting parameter.

Keywords
Tool wears, boring of steels, acousto-optic emission, analysis of variance, Taguchi, laser Doppler vibrometer

Date received: 8 April 2014; accepted: 19 November 2014

Introduction best L/D ratio was 3, which results in least vibration in

boring process. Considering this, in this work, the L/D
Stainless steels are widely used in commercial and ratio was taken as 3 in order to minimize vibrations of
industrial applications due to their excellent corrosive tool and workpiece.
resistance. But it is difficult to machine them due to Tool failure can be identified by observing higher
their hardening tendency in machining process.1 Boring power consumption, poor surface finish, dimensional
is defined as enlarging a hole size that has been already inaccuracy, appearance of a burnishing band on
made by drilling or casting to the designed dimension. machined surface, tool vibrations, workpiece vibration
Therefore, control of dimensional tolerance and surface and so on. Andren et al.7 explained different deforma-
roughness is important.2 The cutting parameters like tion zones in the machining process such as primary,
feed rate, spindle speed, depth of cut, tool nose radius, secondary and tertiary deformation zones. In the pri-
nose angle and vibration data constitute input informa- mary zone, the deformation results in considerable
tion for evaluation of tool life.3 In metal cutting, vibra- stress and strain rates. In the secondary and tertiary
tion of tool and workpiece is the important limiting deformation zones, normal and shear traction loads are
factors for metal removal rate and machining effi-
ciency. In boring operations, the length of boring bar is 1
Department of Mechanical Engineering, Vignan University, Guntur, India
kept long; therefore, it is subjected to vibrations and 2
Department of Mechanical Engineering, GIT, GITAM University,
results in tool failure and poor surface finish and may Visakhapatnam, India
also introduce chatter.4 Length–diameter (L/D) ratio of 3
Department of Mechanical Engineering, University College of
boring bar is one of the important factors which causes Engineering Vizianagaram, JNTUK, Vizianagaram, India
tool vibration. Boring bars with a high L/D ratio tend
Corresponding author:
to chatter. The vibrations also depend on boring bar K Venkata Rao, Department of Mechanical Engineering, Vignan University,
fixation, machine tool condition and selection of cut- Vadlamudi, Guntur 522213 AP, India.
ting conditions.5 Korkut and Kucuk6 proved that the Email: kvenkat_rama@rediffmail.com

Downloaded from pib.sagepub.com at UNIV OF CONNECTICUT on May 12, 2015

2 Proc IMechE Part B: J Engineering Manufacture

Figure 1. Single-degree-of-freedom system.

applied on the tool at workpiece–tool–chip interface divided into regenerative chatter and nonregenerative
due to friction and plastic deformation. The boring bar chatter. The regenerative chatters occur due to undula-
is the weakest link in the boring bar fixing or clamping tion of the workpiece surface, produced during previ-
system of the lathe turret. Due to the above loads ous successive cuts. The nonregenerative chatters are
applied on the tool, the boring bar will vibrate. The induced due to eigenfrequencies of tool holder excita-
vibration of boring bar results in poor machining, sur- tion, plastic deformation of workpiece vibration, fric-
face finish and reduced tool life. Tamizharasan et al.8 tion at tool–chip interface, cutting forces and tool
have used audible acoustic emission signal for online geometry.
monitoring of tool wear. They recorded acoustic emis- Tobias10 explained the two chatters such that the
sion signals and corresponding flank wear for machin- regenerative chatters occur due to undulation of the
ing time of 300 s. Using design of experiments (DOEs) workpiece surface, produced during previous successive
and simulated annealing algorithm for optimization of cuts. The nonregenerative chatters are induced due to
cutting parameters, they found correlation between eigenfrequencies of tool holder excitation, plastic defor-
acoustic signals and flank wear. mation of workpiece vibration, friction at tool–chip
Vibration is defined as the repetitive motion of an interface, cutting forces and tool geometry. Hamdan
object or objects relative to a stationary frame referred to and Bayoumi11 studied the effects of tool rake and
as the equilibrium of the vibration. Inman9 stated that flank clearance angles on the stability of cutting tool.
vibrations exist everywhere and may have a great impact They carried out experimental and analytical investiga-
on the surrounding environment. One general phenom- tion on the effect of tool angles. The two methods
enon of vibration is the ‘‘self-oscillation’’ or resonance. A proved that the larger rake and clearance angles
simplest vibrating system with a single-degree-of-freedom increase the stability of vibration.
system is shown in Figure 1. Vibrations, which are pres- A closed-loop feedback circuit is used in online
ent due to an impressed force on the system, are called as vibration control system to measure the relative vibra-
forced vibrations. Vibrations of air compressors, internal tion between workpiece and cutting tool.12 Xiao et al.13
compression engines, machine tools and other machinery studied the effect of tool nose radius on chatter vibra-
are examples for forced vibrations. tions and surface roughness. They introduced vibration
Vibration became a regular problem in machining cutting method with large nose radius to reduce vibra-
processes like turning, boring, drilling and milling. tions, surface roughness and improve tool strength.
Surface roughness, dimensional accuracy, tool life, pro- Laser Doppler vibrometers (LDVs) are being used to
duction rate and productivity are influenced by the observe high-frequency vibrations during machining
vibrations. A configuration of boring, using a long process. LDVs will observe the chatter behavior of the
overhang boring bar, is shown in Figure 2. The work- end-mill shank and the chatter vibrations at the high
piece is fixed in the headstock chuck, and the boring cutting speed.14 A fast Fourier transform preprocessor
bar is fixed in the turret of computer numerical control is used for generating features from an online acousto-
(CNC) machine tool. In boring process, the workpiece optic emission (AOE) signals to develop a database for
is rotated, and the boring bar is moved against the appropriate decisions. The AOE signals can be trans-
rotating workpiece. Forces which are developed by formed into time domain with different frequency
metal deformation result in vibration of tool and work- zones by the fast Fourier transform. This AOE tech-
piece. Vibrations may be measured in terms of displace- nique is being used in tool condition monitoring.15,16
ment, velocity or acceleration. Surface roughness and tool wear are strongly affected
In turning process, vibration of tool is denoted as by the vibration amplitude and frequency. Improper
self-excited chatter or tool vibration. Based on the driv- tool geometry and the nose radius are more significant
ing force of the tool vibration, the tool vibrations are than depth of cut to cause vibration in metal cutting.17

Downloaded from pib.sagepub.com at UNIV OF CONNECTICUT on May 12, 2015

Rao et al. 3

Table 1. Test trials for boring.

Trial no. S (m/min) F (mm) N (mm)

1 210 0.10 0.8

2 210 0.16 0.8
3 170 0.10 0.8
4 170 0.16 0.8
5 210 0.10 0.4
6 210 0.16 0.4
7 170 0.10 0.4
Figure 2. A configuration of internal turning. 8 170 0.16 0.4

In this work, tool life was evaluated by analyzing

machining characteristics like surface roughness, ampli- variations of experimental design. The word signal
tude of workpiece vibration and volume of metal means the desirable value, and the word noise means
removed. So far, accelerometers are used to measure the undesirable value. The formulae for S/N ratio are
vibration tool while machining is in progress. But in designed such that the experimentalist can always select
this work, an attempt is made to measure vibration of the larger factor-level settings to optimize the quality
rotating workpiece with LDV. characteristics of an experiment. Then, the selection of
calculating the S/N ratio depends on the characteristics
such as smaller-the-better, larger-the-better or normal-
Methodology the-better.20
In this work, Taguchi, analysis of variance (ANOVA) S/N ratios for different categories are given as
and regression analysis methods were used to analyze follows:
the experimental data to find out the effect and contri-
bution of cutting parameters on machining characteris- Normal is the best
tics (surface roughness, amplitude of workpiece   
S=N = 10 log ^y s2 y ð1Þ
vibration and volume of metal removal). Regression
equations were arrived at trials to represent the relation Smaller is the best
between dependent and independent variables.  X 
S=N = 10 log ðI=nÞ y2 ð2Þ
DOE and selection of cutting parameters Larger is the best
A specially designed orthogonal array of Taguchi was  X   
used to investigate the effects of the entire machining S=N = 10 log ðI=nÞ I y2 ð3Þ
parameters through the small number of experiments,
and it takes less time for the experimental investiga-
tions. Salgado et al.18 studied the cutting parameters, where ^y is the average of observed data y, sy2 is the var-
feed rate, spindle speed, depth of cut, tool nose radius, iance of y and n is the number of observations.21
nose angle and vibration data which are the input infor-
mation for evaluation of tool life. Two levels of spindle ANOVA
speed, feed rate and nose radius were taken in their
xek et al.19 have taken two
work to evaluate tool life. Cxic In this study, ANOVA was used to find out the effect
levels of cutting parameters in their work to determine of cutting speed, nose radius and feed rate on tool life
the optimal combinations of drilling parameters and to by analyzing the surface roughness, amplitude of work-
minimize machining costs and time in drilling of AISI piece vibration and amount of metal removed until the
316 stainless steel. tool failed. In the analysis, the percentage contribution
In this study, the cutting parameters, such as nose of each cutting parameter was determined and was used
radius (N), rotational cutting speed (S) and feed rate to measure the corresponding effects on the metal cut-
(F), were selected as control factors with two levels. An ting. The performed experimental plan was evaluated
appropriate orthogonal array was selected to determine at 95% of confidence level.19–21
the optimal cutting parameters and to analyze the
effects of these parameters. The cutting parameters Regression analysis
were assigned to each column, and eight combinations
of parameters were derived as shown in Table 1. Multiple regression analysis was also used to find out
significant cutting parameter. An equation between
dependent and independent variables was developed to
Signal-to-noise ratio characteristics. The Taguchi method predict the error. The coefficient of determination (R2)
uses signal-to-noise (S/N) ratio to measure the is calculated for the dependent variables based on which

Downloaded from pib.sagepub.com at UNIV OF CONNECTICUT on May 12, 2015

4 Proc IMechE Part B: J Engineering Manufacture

Table 2. Chemical composition of AISI 316.

Elements C Mn Si P S Cr Ni N Fe

Percentage 0.08% maximum 2.0% 0.75% 0.045% 0.03% 16%–18% 10%–14% 0.1% Balance

Table 3. Tool geometry of DNMG150608 and DNMG150604.

Cutting Cutting Thickness Hole, Side

edge length point smallest clearance
angle diameter

15.5 mm 55° 6.35 mm 5.16 mm 0°

Figure 3. Workpieces.

the decision could be taken to suggest the equation.

Empirical models were developed for the prediction of
surface roughness in turning. The model considers
workpiece hardness, feed, cutting tool point angle,
depth of cut, spindle speed and cutting time.22

Tool life criteria

A sharp cutting tool is expected to give more cutting
ability for long time with an effective and smooth man-
ner. If it is not giving satisfactory work, then it can be
said as tool failure. Gradual wear is one of the tool fail- Figure 4. Tool inserts.
ure modes that occurs due to loss of mass or weight in
metal cutting on face and flank of tool. The wear on the
face and flank is called as crater wear and flank wear or Experimental procedure
wear land, respectively. According to International The experiment was conducted on CNC lathe DX200
Standards Organization,23 the maximum wear land model. The metal used in this experiment is AISI 316
width (VBmax) should be less than 0.6 mm. This criter- with length of 90 mm, outer diameter of 100 mm and
ion was adopted in this study. inner diameter of 56 mm.
The following sequential procedure was used to
carry out the experiment under dry condition:
Workpiece material and tool inserts
The chemical composition of AISI 316 stainless steel 1. Each test was started with a fresh cutting edge with
is shown in Table 2 with percentages of all elements. one test condition (trial), and machining is stopped
It has good corrosion resistance when exposed to cor- at the end of each pass. After each pass, the depth
rosive atmosphere, and it is regarded as marine grade of cut was increased by 0.2 mm (fixed depth of cut
due to its corrosion resistance. The machinability of was given in each pass) until the tool failed.
this steel is rated at 45% with respect to the AISI 2. Vibration signals from the rotating workpiece were
1112. The workpieces used in the experiment are measured in the machining process using LDV.
shown in Figure 3. The setup of experiment is shown in Figure 5.
Physical vapor deposition (PVD)-coated tungsten 3. After each pass, the tool insert is removed, and
carbide tool inserts were used in this experiment with flank wear is measured with machine vision system
two nose radii of 0.8 mm (DNMG150608) and 0.4 mm (Figure 6). Optical microscopes of crater and flank
(DNMG150604). The insert geometry is shown in wear on the two tool inserts are shown in Figures
Table 3 and Figure 4. 7 and 8, respectively.

Downloaded from pib.sagepub.com at UNIV OF CONNECTICUT on May 12, 2015

Rao et al. 5

Figure 5. Experimental setup for boring.

Figure 7. Crater wear.

Figure 6. Machine vision system. Figure 8. Flank wear.

4. After each pass, the workpiece was also removed,

and its surface roughness is measured on Talysurf
(Figure 9).

The steps 1–4 were to be continued until the tool failed.

After this, two or three passes were performed on the
workpiece to observe the behavior of tool wear.

5. In each trial, surface roughness, volume of metal

removed, time and amplitude of workpiece vibra-
tions are identified when the tool failed.
6. The workpiece was changed, and steps were
repeated with a new trial.

Figure 9. Talysurf.
Results and discussion
In each trial, eight passes were conducted on work- occurred for 0.4-mm nose radius insert, and the corre-
piece with new tool insert. After each pass, surface sponding time- and frequency-domain spectrographs
roughness, flank wear and hardness were measured. are shown in Figure 12.
Volume of metal removed was also noted and shown There was good correlation between amplitude of
in Table 4. workpiece vibration (Y), surface roughness (Ra) and
Figures 10 and 11 represent spectrographs at the first flank wear (VB) for all the trials. VB value of 0.6 mm is
pass and seventh pass (when tool failed) in both time the criterion for tool failure.23 In this study, the machin-
and frequency domains. In trial 8, mechanical chipping ing characteristics observed at the tool failure are

Downloaded from pib.sagepub.com at UNIV OF CONNECTICUT on May 12, 2015

6 Proc IMechE Part B: J Engineering Manufacture

Table 4. Observations in trial 1.

Pass Ra (mm) rms (mm/s) y (mm) VB (mm) Volume (mm3) Hardness HRB

1 1.50 0.7601 0.3512 0.1254 17.639 95.00

2 1.65 0.8823 0.2859 0.1652 35.326 95.25
3 1.80 0.8610 0.2911 0.1972 53.074 95.50
4 1.98 0.8469 0.2904 0.1257 70.883 95.50
5 2.10 0.9228 0.3065 0.2135 88.753 95.75
6 2.30 0.9115 0.3214 0.4521 106.692 96.00
7 3.20 0.8393 0.3251 0.5974 124.693 96.50
8 3.60 0.9240 0.3354 0.6841 142.754 96.75

rms: root mean square.

Figure 10. Time- and frequency-domain spectrographs in boring at the first pass.

Figure 11. Time- and frequency-domain spectrographs in boring when the tool failed.

represented in boldface in Table 4. Regression equa- The machining characteristic values corresponding
tions for surface roughness and amplitude of workpiece to the tool fail failure for all the eight trials are pre-
for all the trials are given in Table 5. sented in Table 6 along with their calculated S/N ratios.
The variations of flank wear, surface roughness and
amplitude of workpiece vibration for two different nose
radii of cutting tools in all trials are shown in Figure Analysis of workpiece vibrations for tool life
13. It is observed that the surface roughness and ampli- In this section, tool life is evaluated by analyzing the
tude of wok piece vibration increase with the increase amplitude of workpiece vibration with Taguchi and
in flank wear as shown in Figure 14. ANOVA methods.

Downloaded from pib.sagepub.com at UNIV OF CONNECTICUT on May 12, 2015

Rao et al. 7

Figure 12. Time- and frequency-domain spectrographs when the tool failed due to mechanical chipping.

Table 5. Regression equations for all trials.

Trials Regression equations for Ra and Y

1 Ra = 6.3 H20.053 T0.11 L20.091 V0.441 VB1.25 (4)

Y = 9.72 H0.0987 T20.019 L20.V0.157 VB0.0575 (5)
2 Ra = 233.1 H0.353 T0.45 L20.181 V0.81 VB0.639 (6)
Y = 20.67 H0.0094 T20.08 L0.0145 V20.059 VB0.268 (7)
3 Ra = 2285 H2.84 T8.79 L0.255 V23.67 VB9.6 (8)
Y = 254.0 H0.541 T1.07 L0.0468 V20.521 VB0.736 (9)
4 Ra = 290.1 H0.881 T0.4 L20.282 V1.22 VB21.84 (10)
Y = 210.2 H0.101 T0.238 L0.017 V20.116 VB0.054 (11)
5 Ra = 57 H20.6 T0.85 L20.119 V0.41 VB3.72 (12)
Y = 28.49 H0.0924 T20.0799 L0.002 V0.0096 VB20.262 (13)
6 Ra = 100 H21.02 T5.02 L20.161 V0.08 VB0.41 (14)
Y = 211.2 H0.115 T20.241 L0.0193 V20.058 VB0.148 (15)
7 Ra = 243.6 H0.44 T1.89 L20.026 V20.384 VB1.17 (16)
Y = 23.9 H20.234 T1.67 L20.0528 V20.208 VB20.536 (17)
8 Ra = 103 H20.992 T21.52 L20.553 V2.83 VB24.72 (18)
Y = 2160 H1.55 T23.96 L0.0856 V1.02 VB1.27 (19)

H: hardness; T: machining time; L: length of workpiece; V: volume of metal; VB: flank wear.

Figure 13. Flank wear, amplitude of vibration and surface roughness for eight trials.

Taguchi. In this method, measured values of amplitude characteristic for all the trials as shown in Table 6.
of workpiece vibration were taken in Table 6, and S/N According to Taguchi results, the cutting speed has
ratio was measured with ‘‘smaller is better’’ more influence on the amplitude and is shown with first

Downloaded from pib.sagepub.com at UNIV OF CONNECTICUT on May 12, 2015

8 Proc IMechE Part B: J Engineering Manufacture

Figure 14. Correlation of surface roughness and amplitude of vibration with flank wear.

Table 6. S/N ratios of experimental results for surface roughness, amplitude of vibration and metal removal rate.

Trial no. N (mm) S (m/min) F (mm/rev) Ra (mm) S/N ratio Y (mm) S/N ratio V (mm3) S/N ratio

1 0.8 210 0.10 3.20 210.1030 0.3251 9.75966 124.693 41.9168

2 0.8 210 0.16 3.21 210.1301 0.3227 9.82402 131.286 42.3644
3 0.8 170 0.10 3.20 210.1030 0.5371 5.39890 155.372 43.8275
4 0.8 170 0.16 3.42 210.6805 0.5373 5.39566 99.8100 39.9835
5 0.4 210 0.10 3.80 211.5957 0.3239 9.79178 142.754 43.0918
6 0.4 210 0.16 4.20 212.4650 0.3903 8.17203 131.286 42.3644
7 0.4 170 0.10 4.20 212.4650 0.5419 5.32162 116.152 41.3005
8 0.4 170 0.16 4.55 213.1602 0.8633 1.27677 119.964 41.5810

S/N: signal-to-noise.

Table 7. Taguchi analysis. ANOVA. In the ANOVA, at a confidence level of 95%,

the experimental results and the Taguchi S/N ratios
Level N S F were analyzed as shown in Table 8. ANOVA also deter-
1 6.141 4.348 7.568
mined the contribution of individual cutting parameters
2 7.595 9.387 6.167 on the amplitude of workpiece vibration. Based on both
Delta 1.454 5.039 1.401 ANOVA for experimental results (amplitude of work-
Rank 2 1 3 piece vibration) as well as ANOVA for S/N ratio, the
cutting speed is identified as significant parameter with
the maximum contribution of 65.48% and 77.72%,
rank in Table 7. Abuthakeer et al.24 also found the respectively.
influence of cutting speed on the tool vibration in exter- According to Figure 15, to minimize the amplitude
nal turning process. of vibration, the best combination of three cutting

Table 8. Analysis of variance.

Source DF Sequential SS Adjusted SS Adjusted MS F p Percentage of contribution

Analysis of variance for Y

N 1 0.01972 0.01972 0.01972 1.79 0.252 8.27
S 1 0.15613 0.15613 0.15613 14.19 0.020 65.48
F 1 0.01859 0.01859 0.01859 1.69 0.264 7.80
Error 4 0.04401 0.04401 0.01100 18.45
Total 7 0.23844 100.00
Analysis of variance for S/N ratio
N 1 4.228 4.228 4.228 2.64 0.179 6.47
S 1 50.776 50.776 50.776 31.74 0.005 77.72
F 1 3.925 3.925 3.925 2.45 0.192 6.01
Error 4 6.399 6.399 1.600 9.80
Total 7 65.328 100.00

DF: degree of freedom; SS: sums of squares; MS: mean square; S/N: signal-to-noise.

Downloaded from pib.sagepub.com at UNIV OF CONNECTICUT on May 12, 2015

Rao et al. 9

Figure 16. Main effects of plot for S/N ratios for surface
Figure 15. Main effects of plot for S/N ratios for vibration of roughness.
workpiece.

ANOVA. In the ANOVA, at a confidence level of 95%,

Table 9. Taguchi analysis. the experimental results and the Taguchi S/N ratios
were analyzed as shown in Table 10. ANOVA also
Level N S F determined the contribution of individual cutting para-
meters on the surface roughness. According to the
1 212.42 211.60 211.07
2 210.25 211.07 211.61
ANOVA for experimental results and ANOVA for S/N
Delta 2.17 0.53 0.54 ratio, the nose radius is the significant parameter, and
Rank 1 3 2 its contribution is 84.51% and 86.43%, respectively.
Marimuthu and Chandrasekaran1 and Muthukrishnan
and Paulo Davim25 also found that the feed has more
parameters of nose radius, cutting speed and feed rate is influence on the surface roughness in external turning.
0.8 mm, 210 m/min and 0.1 mm, respectively. According to Figure 16, to minimize the surface
roughness, the best combination of three cutting para-
meters of nose radius, cutting speed and feed rate is
Analysis of surface roughness for tool life 0.8 mm, 210 m/min and 0.1 mm, respectively
In this section, tool life is evaluated by analyzing the
surface roughness with Taguchi and ANOVA methods.
Analysis of volume of metal removal for tool life
Tool life can also be defined as volume of material
Taguchi. Measured values of roughness for all the trials
removed between two successive sharpening of tools.26
were taken in Table 6, and S/N ratios were measured
In this section, tool life is evaluated by analyzing the
with ‘‘smaller is better’’ characteristic for all the trials.
volume of metal removal with Taguchi and ANOVA
Taguchi results show that the nose radius has more
methods. In past research, influence of metal removal
influence on the surface roughness and is shown with
rate on the tool life was studied. But in this work, effect
first rank in Table 9.
of volume metal removed on the tool life was studied.

Table 10. Analysis of variance.

Source DF Sequential SS Adjusted SS Adjusted MS F p Percentage of contribution

Analysis of variance for Ra

N 1 1.72980 1.72980 1.72980 84.48 0.001 84.51
S 1 0.11520 0.11520 0.11520 5.63 0.077 5.63
F 1 0.12005 0.12005 0.12005 5.86 0.073 5.86
Error 4 0.08190 0.08190 0.02048 4.00
Total 7 2.04695 100.00
Analysis of variance for S/N ratio
N 1 9.3945 9.3945 9.3945 114.90 0.000 86.43
S 1 0.5591 0.5591 0.5591 6.84 0.059 5.14
F 1 0.5882 0.5882 0.5882 7.19 0.055 5.42
Error 4 0.3271 0.3271 0.0818 3.01
Total 7 10.8689 100.00

DF: degree of freedom; SS: sums of squares; MS: mean square; S/N: signal-to-noise.

Downloaded from pib.sagepub.com at UNIV OF CONNECTICUT on May 12, 2015

10 Proc IMechE Part B: J Engineering Manufacture

Table 11. Taguchi analysis.

Level N S F

1 42.08 42.75 42.87

2 42.76 42.10 41.97
Delta 0.68 0.65 0.90
Rank 2 3 1

Table 12. Analysis of variance.

Source DF Sequential SS Adjusted SS Adjusted MS F p Percentage of contribution

Analysis of variance for volume of metal removed

N 1 0.1 0.1 0.1 0.00 0.986 0.01
S 1 387.4 387.4 387.4 0.53 0.506 19.39
F 1 600.8 600.8 600.8 1.14 0.346 30.07
Error 4 909.8 909.8 352.5 45.53
Total 7 1998.2 100.00
Analysis of variance for S/N ratio for volume of metal removed
N 1 0.008 0.008 0.008 0.00 0.949 0.08
S 1 2.159 2.159 2.159 0.71 0.446 22.70
F 1 2.946 2.946 2.946 1.14 0.346 30.97
Error 4 4.399 4.399 1.625 46.25
Total 7 9.512 100.00

DF: degree of freedom; SS: sums of squares; MS: mean square; S/N: signal-to-noise.

Table 13. Analysis of variance with interactions of cutting parameters.

Source DF Sequential SS Adjusted SS Adjusted MS F p Percentage of contribution

Analysis of variance for volume of metal removed

N 1 0.015 0.048 0.048 0.03 0.887 0.15
S 1 1.996 0.970 0.970 0.52 0.546 20.94
F 1 3.095 1.631 1.631 0.87 0.449 32.47
N3S 1 0.972 0.809 0.809 0.43 0.578 10.20
N3F 1 0.735 0.903 0.903 0.48 0.559 7.68
S3F 1 1.772 1.772 1.772 0.95 0.433 18.59
Error 2 0.945 0.945 1.872 9.97
Total 8 9.529 100

DF: degree of freedom; SS: sums of squares; MS: mean square.

Taguchi. In this method, volume of metal removed until parameters of nose radius, cutting speed and feed rate
tool failed was taken and S/N ratio was measured with is 0.4 mm, 210 m/min and 0.1 mm, respectively.
‘‘larger is better’’ characteristic for all the trials.
According to Taguchi results, the feed rate has more
influence on the volume of metal removed and is shown
Regression analysis
with first rank in Table 11. Multiple regression predictive equations were generated
for workpiece vibration, surface roughness and volume
of metal removed based on the control factor as
ANOVA. In the ANOVA, at a confidence level of 95%,
the experimental results and the Taguchi S/N ratios Y = 1:75  0:248N  0:00698S + 1:61F ð20Þ
were analyzed as shown in Table 12. ANOVA also Ra = 5:73  2:32N  0:00600S + 4:08F ð21Þ
determined the contribution of individual cutting para-
V = 112 + 0:6N + 0:242S  236F ð22Þ
meters on the metal removal rate. According to the
ANOVA for experimental results and ANOVA for S/N The regression analysis determined that the spindle
ratio, the feed rate shows more contribution of 30.07% cutting speed has more effect on the surface roughness
and 30.97%, respectively. Interaction of the cutting (0.02 of p value), nose radius has more effect on the
parameters is shown in Table 13. surface roughness (0.001 of p value) and feed rate has
According to Figure 17, to maximize the metal more effect on the surface roughness (0.346 of p value)
removal rate, the best combination of three cutting at confidence level of 95%.

Downloaded from pib.sagepub.com at UNIV OF CONNECTICUT on May 12, 2015

Rao et al. 11

Funding
This research received no specific grant from any fund-
ing agency in the public, commercial or not-for-profit
sectors.

References
1. Marimuthu P and Chandrasekaran K.Experimental
study on stainless steel for optimal setting of machining
parameters using Taguchi and neural network. ARPN J
Eng Appl Sci 2011; 6(10): 119–127.
2. Beauchamp Y, Thomas M, Youssef YA, et al. Investiga-
tion of cutting parameter effects on surface roughness in
lathe boring operation by use of a full factorial design.
Figure 17. Main effects of plot for S/N ratios for volume of Comput Ind Eng (18th International Conference on Com-
metal removed. puters and Industrial Engineering) 1996; 31(3–4): 645–
651.
3. Salgado DR, Alonso FJ, Cambero I, et al. In-process
Conclusion surface roughness prediction system using cutting vibra-
tions in turning. Int J Adv Manuf Tech 2009; 43: 40–51.
In this work, experiments were conducted on lathe 4. Yussefian NZ, Moetakef-Imani B and El-Mounayri H.
using two cutting tool inserts with nose radii of 0.8 and The prediction of cutting force for boring process. Int J
0.4 mm for boring of AISI 316 steel. According to Mach Tool Manu 2008; 48: 1387–1394.
selected orthogonal array, eight experiments (trials) 5. Kuster F and Gygax PE. Cutting dynamics and stability
were conducted with two levels of cutting parameters of boring bars. CIRP Ann-Manuf Techn 1990; 39: 361–
such as workpiece cutting speed, tool insert nose radius 366.
and feed rate. In each trial of experiment, a strong cor- 6. Korkut I and Kucuk Y. Experimental analysis of the
relation among the dependent and independent vari- deviation from circularity of bored hole based on the
ables was found. In this work, a noncontact monitoring Taguchi method. J Mech Eng 2010; 56(5): 340–346.
7. Andren L, Hakansson L, Brandt A, et al. Identification
system was used with LDV to observe vibration of
of dynamic properties of boring bar vibrations in a con-
workpiece during machining. Tool life was evaluated tinuous boring operation. Mech Syst Signal Pr 2004;
by analyzing surface roughness, amplitude of work- 18(4): 869–901.
piece vibration and metal removal rate with the help of 8. Tamizharasan T, Kingston Barnabas J and Pakkirisamy
Taguchi, ANOVA and regression analysis. The follow- V. Optimization of turning parameters by using design of
ing points can be concluded from this work: experiments and simulated annealing algorithm based on
audible acoustic emission signals. Proc IMechE, Part B:
1. Setup of LDV is simple, and it is easy to measure J Engineering Manufacture 2012; 226(7): 1159–1173.
9. Inman DJ. Engineering vibration. 2nd ed. Upper Saddle
vibration of workpiece. So, this study could impart
river, NJ: Prentice Hall, 2001, pp.258–269.
a better technique for vibration measurement in
10. Tobias SA. Machine-tool vibration. Glasgow; London:
different materials from industry and academic Blackie and Sons Ltd, 1965.
perspectives. 11. Hamdan MN and Bayoumi AE. An approach to study
2. Cutting speed is found as a significant parameter the effects of tool geometry on the primary chatter vibra-
with contribution of 77.72% on the amplitude of tion in orthogonal cutting. J Sound Vib 1989; 128(3):
vibration. 451–469.
3. In the analysis of surface roughness, it was found 12. Taskesen A. Computer aided nonlinear analysis of
that nose radius shows more contribution of machine tool vibrations and developed computer soft-
86.43%. ware. Math Comput Appl 2005; 3: 377–385.
4. In the analysis of volume of metal removal, it was 13. Xiao M, Sato K, Karube S, et al. The effect of tool nose
found that feed rate and cutting speed have signifi- radius in ultrasonic vibration cutting of hard metal. Int J
cant contribution of 32.47% and 22.7%, Mach Tool Manu 2003; 43(13): 1375–1382.
14. Nakagawa H, Kurita Y, Ogawa K, et al. Experimental
respectively.
analysis of chatter vibration in end-milling using laser
5. It was found that the workpiece cutting speed of
Doppler vibrometers. Int J Autom Tech 2008; 2(6): 431–
210 m/min and feed rate of 0.1 mm are optimum 438.
for good tool life. 15. Prasad BS, Sarcar MMM and Ben BS. Development of a
6. Same cutting parameter was found as significant in system for monitoring tool condition using acousto-optic
all the three methods considered under study. emission signal in face turning—an experimental
approach. Int J Adv Manuf Tech 2010; 51: 57–67.
16. Venkatarao K, Murthy BSN and Mohanrao N. Experi-
Declaration of conflicting interest mental study on surface roughness and vibration of work-
The authors declare that there is no conflict of interest. piece in boring of AISI 1040 steels. Proc IMechE, Part B:

Downloaded from pib.sagepub.com at UNIV OF CONNECTICUT on May 12, 2015

12 Proc IMechE Part B: J Engineering Manufacture

J Engineering Manufacture. Epub ahead of print 8 May parameters optimization of hard machining while
2014. DOI: 10.1177/0954405414531247. machining hardened steel. J sci Ind Res 2009; 68: 686–
17. Chang FX. An experimental study of the impact of turn- 695.
ing parameters on surface roughness. In: Proceedings of 22. Feng CX and Wang X. Development of empirical models
the industrial engineering research conference, Dallas, TX, for surface roughness prediction in finish turning. Int J
USA, May 2001, paper no. 2036. Adv Manuf Tech 2002; 20: 348–356.
18. Salgado DR, Alonso FJ, Cambero I, et al. In-process sur- 23. ISO 3865:1993. Tool-life testing with single-point turning
face roughness prediction system using cutting vibrations tools.
in turning. Int J Adv Manuf Tech 2009; 43: 40–51. 24. Abuthakeer SS, Mohanram PV and Mohan Kumar G.
19. Cxic
xek A, Kıvak T and Samtasx G. Application of Taguchi Prediction and control of cutting tool vibration in CNC
method for surface roughness and roundness error in lathe with ANOVA and ANN. Int J Lean Think 2011;
drilling of AISI 316 stainless steel. J Mech Eng 2012; 583: 2(1): 1–23.
165–174. DOI: 10.5545/sv-jme.2011.167. 25. Muthukrishnan N and Paulo Davim J. Optimization of
20. Yang WH and Tarng YS. Design optimization of cutting machining parameters of Al/SiC-MMC with ANOVA
parameters for turning operation based on Taguchi and ANN analysis. J Mater Process Tech 2009; 209(1):
method. J Mater Process Tech 1998; 84: 127–129. 225–232.
21. Gopalsamy BM, Mondal B and Ghosh S. Taguchi 26. Sharma PC. A text book on production engineering. 18th
method and ANOVA: an approach for process ed. New Delhi, India: S. Chand, 2009, pp.414–439.

Downloaded from pib.sagepub.com at UNIV OF CONNECTICUT on May 12, 2015