Optimisation of Cutting Parameters Using Taguchi Method in Composite Materials


Kiran Varghese, Rakesh K Rajan, Sivarajan S


Abstract— Glass fiber reinforced polymer (GFRP) composite materials have many applications in engineering and constructional fields. For achieving a better result an accurate way of optimization is required; for that optimum cutting parameters are used. In this paper the characteristic nature that involved in the turning operation of GFRP composite material using uncoated aluminum oxide ceramic inserts at various cutting operations are calculated. The cutting parameters used in this experiment are feed rate, length of the tool from tool holder, depth of cut and the cutting operations are done under constant speed of rotation at 320rpm. MINITAB is the software used for the experiment for determining the orthogonal array. The parameters mentioned above are optimized by considering multiple performance characteristics such as cutting forces and surface roughness. When cutting force was taken into consideration parameters such as the tool from tool holder and depth of cut have contributed significantly and with the help of grey base method optimum parameter combination of surface roughness is also calculated


Index TermsOptimisation, Orthogonal array, surface roughness, Taguchi method, ANOVA method, Turning parameters, Grey base method


.



1 INTRODUCTION

—————————— —————————

Glass fiber reinforced polymer materials are finding appli- cations in variety of engineering fields especially in the filed of aeronautical and automotive engineering. the machining pro- cess required for the fiber reinforced polymer composite is quite different from that of the normal metals that we use. The major problems that we have to face are rapid tool wear, de- fective sub layer with cracks, rough surface finishes on fin- ished components.at present the technology related to cutting tools and inserts are increased in a rapid way. In earlier days ferrous materials made tools. But now in recent times indus- tries employ tools made from various materials such as sin- tered carbides, ceramics and cubic boron nitride. In recent years of processing new tooling materials are being intro- duced and secondary processes are being explored to improve the tool lives, such as heat treatment and employment of sur- face coatings on tool inserts [3]. The machine tool that per- forms turning operations in which unwanted material is re- moved from a work piece rotated against a cutting tool. The rotating horizontal spindle to which the work holding device is attached is usually power driven at speeds that can be var- ied. On lathe the cutting tool is supported on a tool post and is controlled by hand. The researchers from time to time for the process planning of such turning operations adopt proce- dures. Linear programming, quadratic programming, lagran- gian multiplier, geometric program-Ming, particle swarm op- timization, genetic algorithm, Taguchi method are some of the techniques.


used for the optimizing process. Taguchi method is an ex- perimental method. It is effective methodology to find out the effective performance and machining conditions. Taguchi pa- rameter design offers a simple, systematic approach and can reduce number of experiment to optimize design for perfor- mance, quality and manufacturing cost.

Signal to noise ratio and orthogonal array are two major tools used in robust design. The process of optimization may be based on various parameters like best possible surface fin- ish, maximum production rate; minimum production cost etc. in machining operations this is possible by suitable represen- tation of the parameters in terms of objective function and constraints. It has long been recognized that conditions during cutting, such as feed rate, cutting speed and diameter of cut, should be selected to optimize the economics of machining operations. The objective of this research is to study the effect of feed, length of the tool from tool holder, depth of cut in an experimental approach. The machinability of materials is de- termined by surface finish. Surface roughness and dimension- al accuracy are the important factors required to predict ma- chining parameters of any machining operations, optimization of machining parameters not only increases the utility for ma- chining economics, but also the product quality is also in- creased. In this context, an effort has been made to estimate the surface roughness using experimental data. Since turning is the primary operation in most of the production process in the industry, surface finish has greater influence on the quality of the product. Surface finish in turning operation has been found to be influenced in varying amounts by number of fac- tors such as feed rate, material characteristics, hardness, built up edge, cutting speed, tool nose radius, cutting time, depth of cut and tool cutting edge angles, stability of machine tool and work piece setup, and chatter, and use of cutting fluids [2]. Taguchi method consists of a plan of experiments with the objective of acquiring data in a controlled way, which executes these experiments and analyzes the data, in order to obtain the information about the behavior of given process. Orthogonal array is used to define the experimental plans and the treat- ment of the experimental results is based on the analysis of variance (ANOVA).


IJSER © 2013

http://www.ijser.org

LITERATURE SURVEY

Traditionally, the selection of cutting conditions for metal cutting is left to the operator. In certain cases, experience of such operator plays a major role, but even for a skilled opera- tor it is difficult to attain the optimum values each time. Ma- chining parameters mainly used in turning operations are feed rate, depth of cut and cutting speed. The setting of these pa- rameters determines the quality characteristics of turned work piece. Following the pioneering work of Taylor (1907) and his famous tool life equation, different experimental and analyti- cal approaches for the optimization of machining parameters have been investigated. J.paulo davim [1] this paper presents a study of the influence of cutting conditions (cutting velocity and feed) and cutting time on turning metal matrix compo- sites. Experiments, based on the techniques of Taguchi, were performed and machining with cutting conditions is prefixed in work pieces. M. Nalbant, H. Go ̈kkaya, G. Sur [2] investigat- ed the signal to noise ratio, the analysis of variance, and or- thogonal array and are employed to study the performance characteristics in turning operations of AISI 1030 steel bars using Tin coated tools. V.N. Gaitonde, S.R. Karnik, J. Paulo Da- vim[3] reported the Minimum quantity of lubrication (MQL) in machining is established alternative to completely dry or flood lubricating system from the viewpoint of cost, human health issues and ecology. Hence, it is necessary to take a proper MQL and cutting conditions in order to enhance machinability for a given work material. The work aims at determining the optimum amount of MQL and the most appropriate cutting speed and feed rate during turning of brass using K10 carbide tool. Yu-Hsin Lin, Yung-Kuang Yang, Ming-Chang Jeng [4] investigated the optimization of CNC turning operation pa- rameters for SKD11 (JIS) using the Grey relational method. Nine experiments based on an orthogonal array of Taguchi method were performed. Yung-Tien Liu , Wei-Che Chang , Yutaka Yamagata [5] in this research, the optimization process of compensation cutting for eliminating the residual form er- ror of an aspheric surface using the Taguchi method was per- formed.


T.G Ansalam Raj and V.N Narayanan Namboothiri [6] formed an improved genetic algorithm for the prediction of surface finish in dry turning of SS 420 materials..Kantesh Balani, Sandip P. Harimka , Anup Keshri , Yao Chen


,Narendra B. Dahotre , Arvind Agarwalng[7] had studied the multiple length scale of plasma-sprayed Al2O3– carbon nanotube (CNT) Nano composite coating. CNT content and dispersion have been shown to enhance the macro-wear resistance (pin-on-disk) by more than 49 times, and Nano- wear (scratch) resistance up to 19 times. CNTs showed a way to reduce the wear of Al2O3 matrix by (i) increasing densifica- tion, (ii) CNT bridging and (iii) CNT lubrication. Dependence of wear volume loss on the micro structural features is de- scribed at different length scales, and disparity of the material loss which is observed in the macro- and Nano-wear tests are explained. A wear model has been established inorder to nu-

merically quantify wear loss dependence in terms of bulk ma- terial properties and correlating these with wear parameters from Nano-scratch testing. C. Z. Huang1, Wang and X. Ai[8] had carried out an experimental investigation to coat two types of carbide powders TiC and (W, Ti)C, with an alumina ceramic using a solution-gel technology. The coated carbide powder is fabricated into two kinds of new ceramic tool mate- rials by the hot pressing method. A scanning electron micro- scope (SEM) observation reveals that in general the matrix (carbide) grains are uniformly coated with the alumina ceram- ic and the microstructure of the new tool materials is more homogeneous than that of conventionally made uncoated aluminum oxide ceramic. The tests of mechanical properties and wear resistance in machining are finally conducted. It is shown that when machining mild carbon steel the new tool materials can increase the tool-life by up to 100% as compared to other two ceramic tool materials that have the same matrix but fabricated in the conventional way, and the fracture toughness of the material is improved by up to 33%. When compared with hard coated carbide tool, the new materials posses superiority ability in maintaining the wear resistance during the entire tool-life.


Ilhan Asilturk, Harun Akkus [9] they conducted ex- periments on hard turning operations in lathe by the orthogo- nal array of L9 method. LB Abhang and Hameedullah [10] carried out the experiment on a steel turning operation on the basis of Taguchi method. LB Abhang and Hameedullah [11] carried out the experiment on a steel turning operation on the basis of Taguchi method. For analyzing significance of each parameter they used analysis of variance method known as the ANOVA method in their experiment. Ali R. Yildiz [12]he proposed the optimization approach can be applied to two case studies for multi-pass turning operations to illustrate the effectiveness and robustness of the proposed algorithm in ma- chining operations. Fabrício José Pontes, Anderson Paulo de Paiva, Pedro Paulo Balestrassi, João Roberto Ferreira,Messias Borges da Silva [13]discussed the study on the applicability of radial base function (RBF) neural networks for prediction of Roughness Average (Ra) in the turning process of SAE 52100 hardened steel, taguchisorthogonal array is used as a tool to design parameters of the network.


EXPERIMENTAL CONCEPT

Traditional method of doing experiments are too complex and a large number of experiments must be carried out in it. When we consider a large number of parameters and large number of levels the number of experiment increases and a lot of time is consumed for doing that work. To solve this prob- lem, the Taguchi method uses a special design of orthogonal arrays to study the entire parameter space with only a small number of experiments. The experiments were carried out with three independent factors (feed rate, length of the tool from the tool holder and depth of cut at three levels each. Here

a standard L27 orthogonal array is used. The various factors and their levels are shown in table I.


TABLE I. Different parameters and levels


Factors

Level 1

Level 2

Level 3

The Feed rate (mm/rev) “A”

0.130

0.260

0.520

Length of the tool from tool holder (mm) “B”


10


20


30

Depth of cut

(mm) “C”


1


2


3


Using MINITAB15 software orthogonal array re- quired for the experiment is calculated. Experiment is con- ducted on the level based values from the orthogonal array it is mentioned in the table II


TABLE II. Orthogonal array using MINITAB


SL NO.

Feed rate (A)

Length of the tool from tool holder (B)

Depth of cut (C)

1

1

1

1

2

1

1

2

3

1

1

3

4

1

2

1

5

1

2

2

6

1

2

3

7

1

3

1

8

1

3

2

9

1

3

3

10

2

1

1

11

2

1

2

12

2

1

3

13

2

2

1

14

2

2

2

15

2

2

3

16

2

3

1

17

2

3

2

18

2

3

3

19

3

1

1

20

3

1

2

21

3

1

3

22

3

2

1

23

3

2

2

24

3

2

3

25

3

3

1

26

3

3

2

27

3

3

3


Work piece material

The work piece material used is glass reinforced fiber polymer (GFRP) in the form of cylindrical bar of diameter 30mm and length of 150 mm.


Cutting tool material

The cutting tool material used in this experiment is uncoated aluminum oxide ceramic insert.


Machine tool

The turning operation is carried out on a rigid lathe with 2.25kw

(spindle speed 54-1200rpm) motor drive.


Turning Dynamometer

Cutting tool is mounted rigidly on the tool post. The termi- nals from the respective amplifiers are connected to the dy- namometer display unit. Initially the reading is set to a zero value, such that the errors are eliminated.


Surface roughness tester (perthometer) Test principle: inductance type Measurement range: 160 μm

Stylus tip radius: 2 μm

Stylus tip material: diamond

Measurement force: 4MN


Constraints

Range of depth of cut (1 to 3mm)

Range of cutting speed (at constant speed of 320 rpm)

Range of feed rate (0.130-0.520)

Range of length of the tool rom tool holder (10 to 30mm)


RESULTS AND DISCUSSIONS


Experiments are conducted according to the standard or- thogonal array of L27 with the help of MINITAB15. The cut- ting force in x and y direction is measured thrice using a dy- namometer of each work piece. The results obtained are tabu- lated in table III and analysis of variance of the data with the cutting forces with the objective of the analyzing the influence of each variable on the total variance of the results is per- formed and the results obtained are tabulated in table IV for Fx and Fy in table V. It shows percentage contribution of each parameter towards the cutting forces.


TABLE III. Experimental design of L27 orthogonal array


Sl no

A


(mm/rev)

B


(mm)

C


(mm)

Fx


(N)

Fy


(N)

Fz


(N)

Fx


(N)

Fy


(N)

Fz


(N)

Fx


(N)

Fy


(N)

Fz


(N)

1

1

1

1

0.130

10

1

35

155

165

35

150

165

35

155

170

2

1

1

2

0.130

10

2

40

80

165

45

85

165

55

80

170

3

1

1

3

0.130

10

3

55

125

255

55

130

260

60

125

260

4

1

2

1

0.130

20

1

50

110

160

50

115

165

55

110

165

5

1

2

2

0.130

20

2

80

270

360

90

270

340

90

280

330

6

1

2

3

0.130

30

3

20

120

130

20

120

120

30

130

120

7

1

3

1

0.130

30

1

50

230

260

50

230

250

50

230

250

8

1

3

2

0.130

30

2

90

320

350

90

310

345

85

320

345

9

1

3

3

0.130

30

3

20

210

255

25

210

255

20

220

260

10

2

1

1

0.260

10

1

60

140

185

60

145

185

60

140

185

11

2

1

2

0.260

10

2

40

70

155

45

70

155

40

75

155

12

2

1

3

0.260

10

3

50

90

185

55

90

185

50

90

185

13

2

2

1

0.260

20

1

45

145

190

45

150

190

45

145

190

14

2

2

2

0.260

20

2

70

120

150

70

120

125

75

125

155

15

2

2

3

0.260

20

3

55

320

355

55

320

320

55

320

360

16

2

3

1

0.260

30

1

30

120

145

35

120

120

35

120

145

17

2

3

2

0.260

30

2

80

130

140

80

130

135

85

135

140

18

2

3

3

0.260

30

3

60

260

300

65

260

295

60

260

300

19

3

1

1

0.520

10

1

55

120

165

55

120

155

55

120

165

20

3

1

2

0.520

10

2

80

110

210

80

115

215

80

115

220

21

3

1

3

0.520

10

3

40

130

190

40

130

210

45

135

195

22

3

2

1

0.520

20

1

70

230

280

75

230

285

70

230

285

23

3

2

2

0.520

20

2

20

230

265

20

235

270

25

230

265

24

3

2

3

0.520

20

3

35

165

185

35

165

165

35

170

190

25

3

3

1

0.520

30

1

80

255

320

80

260

260

80

255

310

26

3

3

2

0.520

30

2

55

175

215

55

175

225

55

175

210

27

3

3

3

0.520

30

3

40

130

185

40

130

190

45

130

185


TABLE IV. ANOVA table for Fx


Factor

DOF

SS

MSS

P=(SS/SST)×100

A

2

743.20

371.5

1.65%

B

2

789.50

394.75

2.76%

C

2

2822.987

1411.4935

9.87%

AB

4

3296

824

11.52%

BC

4

7391

1847.75

25.84%

AC

4

13552.461

3388.115

47.39%

Total

28595.148

8237.608

100%

TABLE V. ANOVA table for Fy


Factor

DOF

SS

MSS

P=(SS/SST)×100

A

2

8889

4444.5

0.2%

B

2

126678.39

63339.195

3.9%

C

2

3584.432

1792.21

0.1%

AB

4

306619.85

76654.96

9.6%

BC

4

2618645.66

654661.415

82.04%

AC

4

127287.02

31821.755

3.9%

Total

3191704.31

832714.035

100%


From the table IV, it is observed that the interaction fac- tors between feed rate and depth of cut (47.39%) have great influence on cutting force Fx. In the case of parameters depth of cut (9.87%) has great influence on cutting force Fx. But feed rate (1.65%) and length of the tool from tool holder (2.76%) have less significant contribution on cutting force Fx..


DOF = degrees of freedom SS = sum of squares

MSS= mean sum of squares

From the table V, it is observed that the interaction factors be- tween length of the tool from tool holder and depth of cut (82.04%) have great influence on cutting force Fy. In the case of parameters length of the tool from tool holder (3.9%) has great influence on cutting force Fy . But feed rate (0.2%) and depth of cut (0.1%) have less significant contribution on cutting force Fy


From the table V, it is observed that the interaction factors be- tween length of the tool from tool holder and depth of cut (82.04%) have great influence on cutting force Fy. In the case of parameters length of the tool from tool holder (3.9%) has great influence on cutting force Fy . But feed rate (0.2%) and depth of cut (0.1%) have

2

SST= ∑XI

– T2 N

less significant contribution on cutting force Fy

T= sum of forces in x direction

N= total number of experiments


.


Sl

Orthogonal

array

A

B

C

Fx

Fy

Fz

Fx

Fy

Fz

Fx

Fy

Fz

6

1

2

3

0.130

20

3

20

120

130

20

120

120

30

130

120

TABLE VI. Optimized solution of Fx for the experiment


TABLE VII. Optimized solution of Fy for the experiment


Sl

Orthogonal

array

A

B

C

Fx

Fy

Fz

Fx

Fy

Fz

Fx

Fy

Fz

10

2

1

1

0.260

10

1

60

140

185

60

145

185

60

140

185


Dividing the values of three level parameters by the number of experiments carries out optimization procedure and in this case the number of experiments is 27. From the divided values the least values are taken into consideration and their respec- tive level is identified

Optimized solution for Fx can be obtained from table VI and optimized value for Fx are feed (0.130mm), length of the tool from tool holder (2mm), depth of cut (3mm).

Optimized solution for Fx can be obtained from table VI and optimized value for Fx are feed (0.260mm), length of the tool from tool holder (1mm), depth of cut (1mm).


GREY BASE


The surface roughness of each work piece is measured using a surface roughness measuring instrument (perthometer). Roughness value is initially measured twice and after that mean of that value is considered. The results obtained are tab- ulated in table VIII and grey base method on surface rough- ness with the objective of the analyzing the influence of each variable on the total is performed and the results obtained are tabulated in table IX. It shows contribution of each parameter towards the surface roughness.


Grey relational co efficient =

𝜉ij = ΔMin+(ξΔMax÷(Δij + ξΔMax)) ΔMax = 1 ΔMin = 0 ξ = 0.5


By applying above equation grey relational co efficient is ob- tained and are tabulated in table IX. The value of ‘ξ’ always remains same as 0.5 in all the three cases of Fx ,Fy and Ra

.


Sl.No

Orthogonal

Array

Fx(N)

Fy(N

)

Ra (µm)

1

1

1

1

35

150

4.32

2

1

1

2

45

85

5.40

3

1

1

3

55

130

2.39

4

1

2

1

50

115

3.67

5

1

2

2

90

270

3.65

6

1

2

3

20

120

4.36

7

1

3

1

50

230

5.65

8

1

3

2

90

310

3.86

9

2

3

3

25

210

3.90

10

2

1

1

60

145

3.04

11

2

1

2

45

70

4.45

12

2

1

3

55

90

5.40

13

2

2

1

45

150

6.64

14

2

2

2

70

120

6.42

15

2

2

3

55

320

3.80

16

2

3

1

35

120

5.92

17

2

3

2

80

130

4.25

18

2

3

3

65

260

4.19

19

3

1

1

55

120

5.20

20

3

1

2

80

115

4.61

21

3

1

3

40

130

3.54

22

3

2

1

75

230

2.88

23

3

2

2

20

235

5.99

24

3

2

3

35

165

4.43

25

3

3

1

80

260

3.66

26

3

3

2

55

175

3.21

27

3

3

3

40

130

4.42

TA BLE VIII.

Grey base

TABLE IX. Grey relational co- efficient and grade


The grade value is calculated by adding three values of Fx ,Fy and surface roughness and taking theirs average give the grade according to the grade grey ordering is done .


CONCLUSION

For solving machining optimization problems, several conven- tional techniques had been used so far, but they are not robust in nature and have problems when applied to the turning pro- cess, which involves a number Of variables and constraints. To overcome the above problems, Taguchi method is used in this work. Since Taguchi method is experimental method it is realistic in nature. From this experiment the prime factor af- fecting cutting forces are length of the tool from the tool hold- er and depth of cut and in order to obtain the best surface fin- ish on composite material the optimal parameter combination obtained is feed rate(0.260mm/rev),length of the tool from tool holder(10mm), depth of cut (2mm).


REFERENCES


[1] [1] J. Paulo Davim, “Design of optimization of cutting parameters for turning metal matrix composites based on the orthogonal arrays” Journal of Materials Processing Technology 132 (2003) 340-344.

[2] M. Nalbant , H. Go ̈kkaya, G. Sur, “Application of Taguchi method in the optimization of cutting parameters for surface roughness in

turning” Materials and Design 28 (2007) 1379–1385.

[3] V.N. Gaitonde, S.R. Karnik, J. Paulo Davim Selection of optimal

MQL and cutting conditions for enhancing machinability in turning of brass” journal of materials processing technology 204 (2008) 459– 464

[4] Chorng-Jyh Tzeng, Yu-Hsin Lin, Yung-Kuang Yang, Ming- ChangJeng “Optimization of turning operations with multiple per- formance characteristics using the Taguchi method and Grey rela- tional analysis“journal of materials processing technology 2 0 9 ( 2 0 0

9 ) 2753– 2759

[5] Yung-Tien Liu, Wei-Che Chang, Yutaka Yamagata “A study on optimal compensation cutting for an aspheric surface using the Taguchi method” CIRP Journal of Manufacturing Science and Tech- nology 3 (2010) 40–48

[6] T. G.Ansalam Raj and V.N. Narayanan Namboothiri “An improved genetic algorithm for the prediction of surface finish in dry turning of SS 420 materials “manufacturing technology today 47,pp 313- 324,2010.

[7] Kantesh Balani, Sandip P. Harimka “ studied the multiple length scale of plasma-sprayed Al2O3”

[8] C .Z. Huang, Wang and X. Ai “experimental investigation to coat two types of carbide powders TiC and (W, Ti) C, with an alumina ceramic using a sol-gel technology”.

[9] lhan Asiltürk, Harun Akkußs “Determining the effect of cutting


parameters on surface roughness in hard turning using the Taguchi method” Measurement 44 (2011) 1697–1704

I

[10] _lhan Asiltürk, Süleyman Neßseli “Multi response optimization of

CNC turning parameters via Taguchi method-based response surface


Sl no

Orthogonal Array

Grey relational co- effi- cient

A

B

C

Fx

Fy

Ra

Grade

Grey

order

1

1

1

1

0.699

0.423

0.478

0.523

15

2

1

1

2

0.582

0.347

0.632

0.520

16

3

1

1

3

0.500

0.396

0.333

0.409

26

4

1

2

1

0.538

0.378

0.417

0.444

21

5

1

2

2

0.333

0.714

0.415

0.487

19

6

1

2

3

1

0.384

0.482

0.622

6

7

1

3

1

0.538

0.581

0.683

0.600

9

8

1

3

2

0.333

0.555

0.433

0.440

23

9

2

3

3

0.874

0.531

0.437

0.614

8

10

2

1

1

0.466

0.416

0.371

0.417

25

11

2

1

2

0.582

0.352

0.492

0.646

1

12

2

1

3

0.500

0.384

0.632

0.505

17

13

2

2

1

0.437

0.423

1

0.620

7

14

2

2

2

0.636

0.384

0.907

0.642

2

15

2

2

3

0.500

1

0.420

0.640

3

16

2

3

1

0.778

0.384

0.747

0.636

4

17

2

3

2

0.778

0.396

0.470

0.548

13

18

2

3

3

0.583

0.675

0.464

0.574

10

19

3

1

1

0.500

0.384

0.596

0.493

18

20

3

1

2

0.778

0.378

0.511

0.555

12

21

3

1

3

0.411

0.396

0.406

0.404

27

22

3

2

1

0.700

0.581

0.361

0.547

14

23

3

2

2

0.333

0.595

0.766

0.564

11

24

3

2

3

0.389

0.446

0.490

0.441

22

25

3

3

1

0.778

0.675

0.416

0.623

5

26

3

3

2

0.500

0.462

0.382

0.448

20

27

3

3

3

0.411

0.396

0.489

0.432

24

analysis” Measurement 45 (2012) 785–794

[11] LB Abhang and Hameedullah “optimization of maching parameters in steel turning operation by Taguchi method” Procedia Engineering 38 (2012) 40 – 48

[12] Ali R. Yildiz “Hybrid Taguchi-differential evolution algorithm for optimization of multi-pass turning operations” Applied Soft Computing (2012)

[13] Fabrício José Pontes, Anderson Paulo de Paiva, Pedro Paulo Ba- lestrassi, João Roberto Ferreira, Messias Borges da Silva “Optimiza- tion of Radial Basis Function neural network employed for predic- tion of surface roughness in hard turning process using Taguchi’s orthogonal arrays” Expert Systems with Applications 39 (2012) 7776–7787.