Tool Life Based on the Taguchi Method

8/10/2019 Tool Life Based on the Taguchi Method

1/8

ORIGINAL ARTICLE

Optimization of turning parameters for surface roughness

and tool life based on the Taguchi method

Ahmet Hasalk &Ula ayda

Received: 24 April 2007 /Accepted: 25 June 2007 /Published online: 14 September 2007# Springer-Verlag London Limited 2007

Abstract In this study, the effect and optimization of

machining parameters on surface roughness and tool life ina turning operation was investigated by using the Taguchi

method. The experimental studies were conducted under

varying cutting speeds, feed rates, and depths of cut. An

orthogonal array, the signal-to-noise (S/N) ratio, and the

analysis of variance (ANOVA) were employed to the study

the performance characteristics in the turning of commer-

cial Ti-6Al-4V alloy using CNMG 120408-883 insert

cutting tools. The conclusions revealed that the feed rate

and cutting speed were the most influential factors on the

surface roughness and tool life, respectively. The surface

roughness was chiefly related to the cutting speed, whereas

the axial depth of cut had the greatest effect on tool life.

Keywords Surface roughness . Tool life . Taguchi method .

Analysis of variance . Ti-6Al-4V

1 Introduction

Recent trends in the aerospace industry have been to increase

the use of titanium alloys because of the outstanding

mechanical properties that can be provided at critical load-

carrying locations in many military and commercial aircraft

[1]. These alloys are generally used for a component whichrequires the greatest reliability and, therefore, the surface

roughness must be maintained. Additionally, it is essential to

satisfy surface integrity requirements using the cutting tool

material in an economic time frame [2]. Turning is one of the

fundamental machining processes, especially for the finish-

ing of machined parts. Usually, the selection of appropriate

machining parameters is difficult and relies heavily on the

operators experience and the machining parameters tables

provided by the machine-tool builder for the target material.

Hence, the optimization of operating parameters is of great

importance where the economy and quality of a machined

part play a key role [3].

In recognizing the need to reduce the cost and improve

quality and productivity, companies have initiated total

quality management (TQM). TQM is a revolutionary

management commitment, employee involvement, and the

use of statistical tools. The quality engineering method of

Dr. Taguchi, employing design of experiments (DOE), is

one of the most important statistical tools of TQM for

designing high-quality systems at reduced cost. Taguchi

methods provide an efficient and systematic way to

optimize designs for performance, quality, and cost. This

method has been used successfully in designing reliable,

high-quality products at low cost in such areas as

automotive, aerospace, and consumer electronics [4].

The purpose of this paper is to demonstrate an

application of Taguchi parameter design in order to identify

the optimum surface roughness and tool life performance

with a particular combination of cutting parameters in

turning titanium alloy.

Int J Adv Manuf Technol (2008) 38:896903

DOI 10.1007/s00170-007-1147-0

A. Hasalk: U. ayda(*)Department of Manufacturing, University of Firat,

Technical Education Faculty,

23119 Elazig, Turkey

e-mail: [email protected]


2/8

2 Experimental procedure

2.1 Materials, test conditions, and measurement

Commercial Ti-6Al-4V alloy was used as the target material

in the present investigation. The nominal chemical compo-

sition and mechanical properties of the tested material is

illustrated in Tables1 and 2, respectively. Figure1 presents

the microstructure of the titanium alloy. The experimental

studies were carried out on a Johnford T-35 CNC lathe

(20 kW). A CNMG 120408-883 insert which consisted of

94 wt% tungsten carbide with 6 wt% cobalt as the binder

was used as the cutting tool material. The tool life was

determined according to the recommendation of theInternational Organization for Standardization (ISO) criteria

(the period of cutting time until the average flank wear

reached 0.3 mm or the maximum flank wear land

VBmax=0.6 mm). The surface roughness (Ra) was measured

using a Mitutoyo SJ-201 portable device within the

sampling length of 2.5 cm. The cutting tool was studied

under scanning electron microscopy (SEM). The level of

cutting parameter ranges and the initial parameter values

were chosen from the handbook recommended for the

tested material. The graphs, models, and tables presented in

the following sections were produced using the design of

experiments (DOE) and analysis modules of the softwarepackages JMP plus v.5.0. and STATISTICA 7.0.

2.2 The Taguchi method and design of experiments

Taguchis method of experimental design provide a simple,

efficient, and systematic approach for the optimization of

experimental designs for performance quality and cost [5].

The traditional experimental design methods are too

complex and difficult to use. Additionally, large numbers

of experiments have to be carried out when the number of

machining parameters increase [6,7]. Taguchi [8] designed

certain standard orthogonal arrays by which the simulta-

neous and independent evaluation of two or more param-eters for their ability to affect the variability of a particular

produ ct or process characteristics can be done in a

minimum number of tests. A loss function is then defined

to calculate the deviations between the experimental value

and the desired value. This loss function is further

transferred into a signal-to-noise (S/N) ratio, . Usually,

there are three S/N ratios available, depending on the type

of characteristic; the lower-the better (LB), the higher-the

better (HB), and the nominal-the better (NB). In turning,

lower surface roughness and higher tool life are indications

of better performance. Therefore, for obtaining optimum

machining performance, the LB and HB ratios wereselected for surface roughness and tool life, respectively.

The S/N ratios for each type of characteristic can be

calculated as follows:

Nominal is the best : S=NNB 10 logjy

s2y

! 1

Larger is the better maximize :

S=NLB

10 log 1

nXni1

1

y2i !

2

Table 1 Chemical composition of Ti-6Al-4V alloy

Ti Al V Fe O C N H

89.464 6.08 4.02 0.22 0.18 0.02 0.01 0.0053

Table 2 Workpiece material properties

Workpiece material Ti-6Al-4V

Hardness (HV) 600

Melting point (C) 1660

Ultimate textile strength (Mpa) 832

Yield strength 745

Impact-toughness (J) 34

Modulus of elasticity (106 Mpa) 11.3

Fig. 1 Typical microstructure of Ti-6Al-4V alloy (the light area is

predominant of the phase rich in aluminum and the dark area is

predominant of the phase rich in vanadium)

Table 3 Machining settings used in the experiments

Symbol Cutting parameter Level 1 Level 2 Level 3

A Cutting speed (m/min) 30 60 90

B Feed rate (mm/rev) 0.15 0.25 0.35

C Depth of cut (mm) 0.5 1 2

Int J Adv Manuf Technol (2008) 38:896903 897


3/8

Smaller is the better minimize :

S=NSB 10 log 1

nX

n

i1

y2i ! 3

where y is the average of the observed data, s2y is the

variance ofy, n is the number of observations, and y is the

observed data. Regardless of the category of the perfor-

mance characteristics, the greater S/N ratio corresponds to

the better performance characteristics. Therefore, the

optimal level of the process parameters is the level with

the highest S/N ratio, [9].

The machining parameters (control factors) chosen for the

experiments were: (i) cutting speed, (ii) feed rate, and (iii)

depth of cut. Each parameter has three levels, denoted 1, 2,

and 3. Table3indicates the factors and their levels. To select

an appropriate orthogonal array for the experiments, the total

degrees of freedom (DOF) need to be computed. The DOF

for the orthogonal array should be greater than or at least

equal to those for the design parameters. In this study, an L9orthogonal array with four columns and nine rows was used

(Table4). This array has eight DOF and it can handle three-

level design parameters. The surface roughness and tool life

results were subject to the analysis of variance (ANOVA).

3 Analysis and the discussion of experimental results

3.1 Analysis of the signal-to-noise ratio

Table 4 shows the experimental results for the surface

roughness and tool life and corresponding S/N ratios using

Eqs.2 and3. The mean S/N ratio for each level of the other

machining parameters was calculated in a similar mannerand the results are shown in Tables 5 and 6, respectively.

Additionally, the total mean S/N ratio is computed by

averaging the total S/N ratios. Based on the data presented

in Table 5, the optimal machining performance for the

surface roughness was obtained at 90 m/min cutting speed

(Level 3), 0.15 mm/rev feed rate (Level 1), and 0.5 mm

depth of cut (Level 1) settings. Figure2presents plots of the

S/N ratio for the three control parameters A, B, and C,

studied at three levels for the surface roughness. The S/N

ratio corresponds to the smaller variance of the output

characteristics around the desired value. Figure 3 shows the

comparison of the predicted and actual and residual andactual S/N ratios for the surface roughness using regression

analysis. Most of the points are close to the line and the

deviations are very small and negligible. With respect to the

coefficient of multiple determinations (R2) of the fitted

model (Fig. 3), also known as R2-statistics, the value

obtained for the surface roughness is 0.92. This means that

the model as fitted explains 92% of the variability of surface

roughness. The equation of this model proposed for the

surface roughness by the variables of cutting speed, feed

rate, and axial depth of cut parameters is shown in Eq.4:

R a 3:1955V 4:2311 105f 20703:7302 a

4:2334 105Vf 20826:5222V a

6:0964 105f a

4

On the other hand, the estimated response surfaces ofRa,

as a function of the factors of cutting speed and feed rate

Table 4 Experimental design using the L9 orthogonal array and

experimental results

Exp.

no.

A B C Surface

roughness

(m)

Tool

life

(s)

S/N ratio

for surface

roughness

S/N

ratio for

tool life

1 1 1 1 1.198 2458 5.6570 67.8116

2 1 2 2 3.365 2176 10.5397 66.7532

3 1 3 3 11.428 1531 21.1594 63.6995

4 2 1 2 2.785 742 8.8965 57.4081

5 2 2 3 5.136 670 14.2125 56.5215

6 2 3 1 8.634 360 18.7242 51.1261

7 3 1 3 4.900 893 13.8039 59.0170

8 3 2 1 2.131 200 6.5717 46.0206

9 3 3 2 4.416 344 12.9006 50.7312

Table 5 Response table mean signal-to-noise (S/N) ratio for surface

roughness factor and significant interaction

Symbol Cutting

parameter

Mean S/N ratio

Level 1 Level 2 Level 3 max

min

A Cutting speed 12.452 13.944 11.092a 1.36

B Feed rate 9.452a 10.441 17.595 8.143

C Depth of cut 10.318a 10.779 16.392 6.074

Total mean S/N ratio= 12.496aOptimum level

Table 6 Response table mean S/N ratio for tool life factor and

significant interaction

Symbol Cutting parameter Mean S/N ratio

Level 1 Level 2 Level 3 max

min

A Cutting speed 0a 19.561 22.656 22.656

B Feed rate 10.333a 20.294 21.540 11.207

C Depth of cut 20.646 17.334 14.437a 6.209

Total mean S/N ratio= 57.676aOptimum level

898 Int J Adv Manuf Technol (2008) 38:896903


4/8

and cutting speed and axial depth of cut, with the other

factors held constant at their central values, are shown in

Figs.4and 5, respectively. As shown in Figs. 4and5, thesurface roughness has a tendency of increasing when both

feed rate and axial depth of cut factors are increased within

the work interval considered for this study. There is no

significant change observed due to the cutting speed. But,

especially at the end of a tools life, the surface finish

became rougher. This is probably due to deformation on the

flank face or adherence of workpiece material at the tool

nose. Figure6shows a typical side view of the cutting tool.

The chip can be observed welded on the surface of the

cutting tool and covering the cutting edge (BUE).

From Table 6, the optimal machining performance for tool

life was obtained at 30 m/min cutting speed (Level 1),0.15 mm/rev feed rate (Level 1), and 2 mm depth of cut

(Level 3) settings. The mean S/N ratio graph for tool life is

shown in Fig.7. In Fig.8, the actual and predicted S/N ratios

and residual and predicted S/N ratios for tool life were

compared using regression analysis. Here, the R2 of the

model is 0.96. The equation of this model proposed for the

tool life is shown in Eq.5:

tool life 151:5023V 2:2852 107f 4:105

106a 2:2812 107V f

4:2013 106V a 4:9128

107f a 5

In Figs.9 and 10, the estimated response surface of tool

life as a function of cutting speed and feed rate and as a

function of cutting speed and axial depth of cut, respec-

tively, are depicted. Under constant axial depth of cut and

feed rate conditions, an increase in cutting speed leads to asharpen decrease in tool life. This phenomenon can be

attributed to the very high chemical reactivity of titanium

alloy and a very high temperature occurring close to the

cutting edge as a result of plastic deformation under

compressive stress during machining [10].

Fig. 2 Mean signal-to-noise

(S/N) ratio graph for surface

roughness

Fig. 3 Comparison of actual

and predicted and residual and

predicted S/N ratios of surface

roughness using regressionanalysis



5/8

3.2 Analysis of variance

The purpose of the statistical analysis of variance

(ANOVA) is to investigate which design parameter signif-

icantly affects the surface roughness and tool life. Based on

the ANOVA, the relative importance of the machining

parameters with respect to surface roughness and tool life

was investigated to determine more accurately the optimum

combination of the machining parameters. The analysis is

carried out for the level of significance of 1% (the level of

confidence is 99%) [11, 12]. Tables 7 and 8 show the

results of the ANOVA analysis for the machining outputs,

respectively. The last columns of Tables 7 and 8 indicate

the percentage contribution of each factor on the total

variation, indicating their degree of influence on the results.

The greater the percentage contribution, the greater the

influence a factor has on the performance. According to

Table 7, the feed rate was found to be the major factor

affecting the surface roughness (54.50%), whereas the axial

depth of cut and cutting speed factors affect the surface

roughness by 31.57% and 5.62%, respectively. The change

of cutting speed in the range given in Table 3 has aninsignificant effect on the surface roughness. Table 8shows

the results of the ANOVA for tool life. It can be found that

the cutting speed is the most significant cutting parameter

for affecting tool life (73.69%). The feed rate affects the

tool life by 14.42%. The axial depth of cut has an

insignificant effect on tool life (7.91%).

3.3 Confirmation tests

The purpose of the confirmation test is to validate conclusions

drawn during the analysis phase [13]. Once the optimal level

of the process parameters is selected, the final step is topredict and verify the improvement of the performance

characteristics using the optimal level of the process

parameters. The predicted S/N ratio using the optimal

level of the process parameters can be calculated as [7,9]:

m Xqi1

ji m

6

where hm is the mean of the S/N ratio,i is the mean of the

S/N ratio at the optimal level, and q is the number of

12

10

8

6

4

2

0

V(m/min)

f(mm/rev)

Ra

(m)

Fig. 4 Estimated response surface ofRa vs. Vand f

12

10

8

6

4

2

V(m/min)

a (mm)

Ra

(m

)

Fig. 5 Estimated response surface ofRa vs. Vand a

Fig. 6 Scanning electron microscopy (SEM) image of the cutting tool

showing the built up edge (BUE) after machining Ti-6Al-4 V alloy

(V=60 m/min, a=1 mm, f=0.25 mm/rev)



6/8

Fig. 7 Mean S/N ratio graph for

tool life

Fig. 8 Comparison of actual

and predicted and residual and

predicted S/N ratios of tool lifeusing regression analysis

3000

2500

2000

1500

1000

500

f(mm/rev)V(m/min)

Toollife(s)

Fig. 9 Estimated response surface of tool life vs. Vand f

3000

2500

2000

1500

1000

5000

V(m/min)a(mm)

Toollife(s)

Fig. 10 Estimated response surface of tool life vs. Vand a



7/8

Table 8 Results of ANOVA

for tool life Machining

parameter

Degrees of freedom Sum of squares Mean square F ratio Contribution (%)

Cutting speed 2 332.770 166.385 71.104 73.69Feed rate 2 65.129 32.564 13.916 14.42

Depth of cut 2 35.720 17.860 7.632 7.91

Error 2 4.680 2.340 3.99

Total 8 438.299 100

Table 7 Results of analysis of

variance (ANOVA) for surface

roughness

Machining

parameter

Degrees of freedom Sum of squares Mean square F ratio Contribution (%)

Cutting speed 2 12.2 6.1 2.606 5.62

Feed rate 2 118.4 54.2 23.162 54.50

Depth of cut 2 68.6 34.3 14.658 31.57

Error 2 4.680 2.340 8.31

Total 8 203.88 100

Table 9 Results of the confir-mation experiment for surface

roughness

Initial cutting parameters Optimal cutting parameters

Prediction Experiment

Level A2B2C2 A3B1C1 A3B1C1

Surface roughness (m) 4.215 1.726

S/N ratio (dB) 12.4960 5.2200 4.7408

Improvement os S/N ratio 7.7552

Table 10 Results of the con-

firmation experiment for tool

life

Initial cutting parameters Optimal cutting parameters

Prediction Experiment

Level A2B2C2 A1B1C3 A1B1C3

Surface roughness (m) 520 1,746

S/N ratio (dB) 54.3201 68.8079 64.8409

Improvement of S/N ratio 10.520



8/8

process parameters that significantly affect the perfor-

mance characteristics.

Table9shows the results of the confirmation experiments

using the optimal machining parameters for surface rough-

ness. The increase of S/N ratio from the initial machining

parameters to the level of optimal machining parameters is

7.755 dB. The surface roughness is decreased by 2.44 times.

So, the surface roughness is greatly improved by using theapproach adopted in this paper. Table10shows the results of

the confirmation experiments using the optimal machining

parameters for tool life. The improvement of S/N ratio from

the initial machining parameters to the optimal machining

parameters is 10.520 dB. The tool life is increased by 3.35

times. From the confirmation tests, good agreement between

the predicted machining performance and the actual machin-

ing performance were observed. Additionally, the experi-

mental results confirmed the validity of the applied Taguchi

method for enhancing the machining performance and the

optimizing the turning parameters. The surface roughness

and tool life are greatly improved by using the approach.

4 Conclusion

In this study, an investigation on the surface roughness and

tool life based on the parameter design of the Taguchi

method in the optimization of turning operations has been

investigated and presented. Summarizing the mean exper-

imental results of this study, the following generalized

conclusions can be drawn:

1. Based on the analysis of variance (ANOVA) results, thehighly effective parameters on both the surface roughness

and tool life were determined. Namely, the feed rate

parameter is the main factor that has the highest

importance on the surface roughness and this factor is

about 1.72 times more important than the second ranking

factor (depth of cut). The cutting speed does not seem to

have much of an influence on the surface roughness.

2. The tool life is affected strongly by the cutting speed

(73.69%), whereas the feed rate and depth of cut have a

significant statistical influence.

3. Based on the signal-to noise ratio results in Tables 5

and6, we can conclude that the A3B1C1 (V=90 m/min,f=0.15 mm/rev,a=0.5 mm) and A1B1C3 (V=30 m/min,

f=0.15 mm/rev, a=2 mm) settings are the optimal

machining parameters for surface roughness and tool

life, respectively.

4. The improvement of the surface roughness from the

initial machining parameters to the optimal machining

parameters is about 244%, whereas the tool life is

improved by 335%.

5. The regression results showed that the deviationsbetween the actual and predicted S/N ratios of both

surface roughness and tool life are small for each

parameter.

References

1. Ramulu M, Branson T, Kim D (2001) A study on the drilling of

composite and titanium stacks. Compos Struct 54:6777

2. Haron CHC, Jawaid A (2005) The effect of machining on surface

integrity of titanium alloy Ti-6Al-4V. J Mater Process Technol

166:188

1923. Saravanan R, Sankar RS, Asokan P, Vijayakumar K, Prabhaharan

G (2005) Optimization of cutting conditions during continuous

finished profile machining using non-traditional techniques. Int J

Adv Manuf Technol 26:3040

4. Unal R, Dean EB (1991) Taguchi approach to design optimization

for quality and cost: an overview. In: Proceedings of the 13th

Annual Conference of the International Society of Parametric

Analysts, New Orleans, Louisiana, May 1991

5. Shaji S, Radhakrishnan V (2003) Analysis of process parameters

in surface grinding with graphite as lubricant based on the

Taguchi method. J Mater Process Technol 141:5159

6. Yang WH, Tarng YS (1998) Design optimization of cutting

parameters for turning operations based on the Taguchi method. J

Mater Process Technol 84:122129

7. Lin TR (2002) Optimization technique for face milling stainlesssteel with multiple performance characteristics. Int J Adv Manuf

Technol 19:330335

8. Taguchi G (1990) Introduction to quality engineering. Asian

Productivity Organization (APO), Tokyo, Japan

9. Nalbant M, Gkkaya H, Sur G (2006) Application of Taguchi

method in the optimization of cutting parameters for surface

roughness in turning. Mater Des 28:13791385

10. Nabhani F (2001) Machining of aerospace titanium alloys. Robot

Comput Integr Manuf 17:99106

11. Ross PJ (1998) Taguchi technique for quality engineering.

McGraw-Hill, New York

12. Roy RK (1990) A primer on Taguchi method. Van Nostrad

Reinhold, New York

13. Tosun N, Cogun C, Tosun G (2004) A study on kerf and material

removal rate in wire electrical discharge machining based onTaguchi method. J Mater Process Technol 152:316322


Documents

Tool Life Based on the Taguchi Method