Application of Taguchi & Response Surface methodology in Optimization for Machining of Ceramics with Abrasive Jet Machining

Available online at www.sciencedirect.com

ScienceDirect Materials Today: Proceedings 2 (2015) 3308 – 3317

4th International Conference on Materials Processing and Characterization

Application of Taguchi & Response surface methodology in Optimization for machining of ceramics with abrasive jet machining D.V. Srikantha, M. Sreenivasa Raob a

Department of Mechanical Engineering, St Martin’s Engineering College, Secunderabad, T G. India b Dept. of Mechanical Engineering, JNTUH, Kukatpally, Hyderabad, Telangana. India.

Abstract Abrasive jet machining also known as micro-abrasive blasting or pencil blasting is a non-traditional machining process in which the metal removal takes place due to high velocity abrasive particles .The abrasive particles are propelled by high velocity carrier gas (commonly air). This process is specially used for edge shapes and intricate shapes. The metal cutting or drilling is performed due to high impact of abrasive particles on the work surface. This machining technology proves good results when tool life is taken into consideration, given the fact that the abrasive material can be reused several times before abrasive particles lose their cutting effect. It is already proved that this technique is effectively adopted for polishing and deburring processes. The present study highlights the influence of different parameters of Abrasive jet machining like Pressure, SOD, Abrasive Flow Rate, on the Metal removal and Kerf width on Ceramic Tiles, the type of abrasive particle used for this experiments is Al2O3. The experiments are conducted according to TAGUCHI method of L9 orthogonal array and RSM, latter compared with the Results of ANOVA using STATGRAPHICS. ©©2015 Ltd. All Elsevier rights reserved. 2014Elsevier The Authors. Ltd. All rights reserved. Selection peer-review under responsibility of theof conference committee membersmembers of the 4th of International conference on Materials on Selectionand and peer-review under responsibility the conference committee the 4th International conference Processing and Characterization.

Materials Processing and Characterization. Keywords: Pressure;SOD;Taguchi;Anova;

1.

Introduction

Ceramic is a word taken from Greek word ‘keramos’ means potter’s clay. Ceramic materials are non-metallic, inorganic compounds – primarily compounds of oxygen, but also compounds of carbon, nitrogen, boron, and silicon. Originally, the art of making pottery, now a general term for the science of manufacturing articles prepared from

2214-7853 © 2015 Elsevier Ltd. All rights reserved. Selection and peer-review under responsibility of the conference committee members of the 4th International conference on Materials Processing and Characterization. doi:10.1016/j.matpr.2015.07.149

3309

D.V. Srikanth and M. Sreenivasa Rao / Materials Today: Proceedings 2 (2015) 3308 – 3317

pliable, earthy materials that are made rigid by exposure to heat. Ceramics includes the manufacture of earthenware, porcelain, bricks, and some kinds of tile and stoneware. The minerals used to make ceramics are dug from the earth and are then crushed and ground into fine powder. Manufacturers often purify this powder by mixing it in solution and allowing a chemical precipitate (a uniform solid that forms within a solution) to form. The precipitate is then separated from the solution, and the powder is heated to drive off impurities, including water. The result is typically a highly pure powder with particle sizes of about 1 micrometre. “Abrasive jet machining utilizes the pressure of fluid stream to remove material from the surface of the job (2)”. When using air as a medium the mixture of air and abrasives are allowed to impinge on the work surface at about 200 to 400m/s through the nozzle and work material is eroded by the high velocity abrasive particles. “The inside diameters of the nozzle are about 0.04mm and standoff distance is kept about 0.7 to1.0mm (3)”. The process can be easily controlled to vary the metal removal rate which depends on flow rate and size of abrasive particles .the cutting action is cooled because the carrier gas serves as a coolant. “Till date there has been a through and detailed experiment and theoretical study on the Process (7,8,9)”.

a)

b)

Fig 1 (a) Process Diagram of Abrasive jet machining

(b) Set up of Abrasive Jet machine

Design of Experiments (DOE) techniques enables designers to determine simultaneously the individual and interactive effects of many factors that could affect the output results in any design. DOE also provides a full insight of interaction between design elements; therefore, it helps turn any standard design into a robust one. Simply put, DOE helps to pin point the sensitive parts and sensitive areas in designs that cause problems in Yield. Designers are then able to fix these problems and produce robust and higher yield designs prior going into production. Taguchi methods (namely Taguchi orthogonal arrays)(4,5) can be used in the design of experiments in order to reduce the variations but still give statistically valid results on individual content elements.. Taguchi's method for experimental design is straightforward and easy to apply to many engineering situations, making it a powerful yet simple tool. Response surface methodology (RSM) consists of a group of mathematical and statistical techniques used in the development of an adequate functional relationship between a response of interest, y, and a number of associated control (or input) variables denoted by x1, x2... xk. In general, such a relationship is unknown but can be approximated by a low-degree polynomial model of the form. ‫ ݕ‬ൌ ݂ሺ‫ݔ‬ሻߚ ൅ ߳

(1)

where X=(x1,x2,...,xk), f(x) is a vector function of p elements that consists of powers and cross-products of powers of x1,x2,...,xk up to a certain degree denoted by d (≥1), β is a vector of p unknown constant coefficients’ referred to as parameters, and is a random experimental error assumed to have a zero mean. “This is conditioned on the belief that model (1)”” provides an adequate representation of the response. In this case, the quantity f(x)β represents the mean response, that is, the expected value of y, and is denoted by μ(x). Two important models are commonly used in RSM.

3310

D.V. Srikanth and M. Sreenivasa Rao / Materials Today: Proceedings 2 (2015) 3308 – 3317

These are special cases of model(1) and include the first-degree model (d=1).

(2) And the second degree model (d=2)

(3) STATGRAPHICS is an easy-to-learn, easy-to-use PC-based statistics software package with an intuitive design that will make your everyday data analysis easier than you ever imagined. An important feature of STATGRAPHICS is the ability to construct expressions that create or transform data “on-the-fly”. 2.

Experimentation & Analysis Experiments were conducted according to Design of Experimentation Method i.e. TAGUCHI (6) Analysis of Orthogonal Array L9 of three factors and nine levels. The MRR values obtained after experimentation were implemented in Taguchi and compared with ANOVA .The Table 1 indicates the levels of Experiments on factors (Pressure, AFR, and SOD), MRR, S/N ratio. The experimental work was carried on a test rig which was designed and manufactured in the workshops of the Mechanical Engineering Department, St Martin’s Engineering College, Secunderabad. Table 1: Process parameters and levels Machining Parameters

Level 1

Level2

Level3

Pressure

4

6

8

AFR

3.5

4.5

5.5

Stand of Distance

10

15

20

Table 1 represents the Factors used for the Taguchi Experimentation. The Process parameters of Abrasive jet machining are considered as the Factors of Analysis. The outputs i.e. MRR was considered as Responses. The ceramic tiles are employed for experimentation with Abrasive Jet Machining set up at various level designed according to Taguchi. The type of abrasives used for experimentation are Al2O3, Silicon Carbide of different grit sizes. The Abrasive particles mixed with air were impinged on the ceramics with diff Pressures to obtain cutting. Table 2: Experimental Design Matrix and Results S.NO MACHINING PARAMETERS

MRR

PSNRA1

10

0.0216

-33.9428

15

0.0231

-31.7547

5.5

20

0.0469

-26.9178

3.5

15

0.0411

-28.0644

6

4.5

20

0.0657

-24.2805

6

6

5.5

10

0.0476

-25.4748

7

8

3.5

20

0.0648

-22.7954

8

8

4.5

10

0.0582

-25.0428

9

8

5.5

15

0.0874

-21.8016

Pressure

AFR

1

4

3.5

2

4

4.5

3

4

4

6

5

SOD

D.V. Srikanth and M. Sreenivasa Rao / Materials Today: Proceedings 2 (2015) 3308 – 3317

3311

Table 3: Response Table for Signal to Noise Ratios(Larger is better) S.NO Pressure AFR SOD 1

-30.87

-28.27

-28.15

2

-25.94

-27.03

-27.21

3

-23.21

-24.73

-24.66

Delta

7.66

3.54

3.49

Rank

1

2

3

Main Effects Plot for SN ratios Data Means

Pressure

AFR

-24

Mean of SN ratios

-26 -28 -30 -32 4

6 SOD

8

10

15

20

3.5

4.5

5.5

-24 -26 -28 -30 -32

Signal-to-noise: Larger is better

Fig 2: Graphs indicates the Effect of Pressure, AFR, SOD on MRR

Optimal Range in the MRR is observed by Taguchi method of Optimization is Pressure-8 kg/cm2, AFR 5.5 gm/min and SOD 20 mm observed from the Graphs. Response values were obtained by STATGRAPHICS. Experiments were conducted at different levels (41 levels) and Analysed with Analysis of Variance. STATGRAPHICS is an easy-to-learn, easy-to-use PC-based statistics software package with an intuitive design that will make your everyday data analysis easier than you ever imagined. An important feature of STATGRAPHICS is the ability to construct expressions that create or transform data “on-the-fly”. An important technique for analysing the effect of categorical factors on a response is to perform an Analysis of Variance. An ANOVA decomposes the variability in the response variable amongst the different factors. Depending upon the type of analysis, it may be important to determine: (a) which factors have a significant effect on the response, and/or (b) how much of the variability in the response variable is attributable to each factor. STATGRAPHICS Centurion provides several procedures for performing an analysis of variance. The General Linear Models procedure is used whenever the Basic procedures of ANOVA like one-way, Two-way variance etc. are not appropriate. It can be used for models with both crossed and nested factors, models in which one or

3312

D.V. Srikanth and M. Sreenivasa Rao / Materials Today: Proceedings 2 (2015) 3308 – 3317

more of the variables is random rather than fixed, and when quantitative factors are to be combined with categorical ones. Designs that can be analysed with the GLM procedure include partially nested designs, repeated measures experiments, split plots, and many others. 2.1 Response Surface Design Attributes x Design class: Response Surface x Design name: 3-level factorial design: 3^3 x File name: C:\Users\Srikanth\Desktop\Ceramics\stat ceramics.sfx 2.2 Base Design x Number of experimental factors: 3 x Number of blocks: 1 x Number of responses: 1 x Number of runs: 41, including 14 centre points per block x Error degrees of freedom: 31 x Randomized: Yes Table 4 : RSM experiments (41)of Factors and Objective using stat graphics table EXPERIMENT NO 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39

Pressure kg/cm2 6 6 8 4 8 6 6 6 6 4 6 8 8 4 6 8 6 6 6 8 6 4 6 6 6 6 6 4 6 6 8 4 6 6 6 8 4 4 4

AFR gm/min 4.5 4.5 4.5 3.5 3.5 4.5 5.5 4.5 4.5 3.5 4.5 5.5 4.5 5.5 4.5 3.5 4.5 4.5 3.5 4.5 4.5 5.5 3.5 4.5 4.5 4.5 5.5 4.5 3.5 4.5 3.5 5.5 4.5 4.5 5.5 5.5 3.5 4.5 4.5

SOD mm 15 15 15 10 20 15 20 15 15 15 15 20 20 10 15 15 15 10 20 10 15 15 15 15 15 15 10 15 10 15 10 20 15 15 15 15 20 20 10

MRR gm/sec 0.0534 0.0534 0.0742 0.0216 0.0648 0.0534 0.0861 0.0534 0.0534 0.0327 0.0534 0.0841 0.0795 0.0225 0.0534 0.0476 0.0534 0.0416 0.0548 0.0582 0.0534 0.0317 0.0411 0.0534 0.0534 0.0534 0.0476 0.0231 0.0374 0.0534 0.0758 0.0469 0.0534 0.0534 0.0672 0.0874 0.041 0.0327 0.0209

3313

D.V. Srikanth and M. Sreenivasa Rao / Materials Today: Proceedings 2 (2015) 3308 – 3317

40 41

6 8

4.5 5.5

20 10

0.0657 0.0648

Table 5 :Factors and the range of values &units Factors Low High Units Continuous Pressure

4.0

8.0

kg/cm2

Yes

AFR

3.5

5.5

gm/min

Yes

SOD

10.0

20.0

mm

Yes

3-level factorial design: 3^3 design which will study the effects of 3 factors in 41 runs. The design is to be run in a single block. The order of the experiments has been fully randomized. This will provide protection against the effects of lurking variables. Table 6 : Estimated effects for MRR (gm/sec) Effect Estimate

Stnd. Error

V.I.F.

average

0.0530576

0.00141615

A:Pressure

0.0403667

0.00282104

1.0

B:AFR

0.0135

0.00282104

1.0

C:SOD

0.0183556

0.00282104

1.0

AA

-0.00998747

0.00430428

1.30595

AB

0.00705

0.00345505

1.0

AC

-0.00433333

0.00345505

1.0

BB

0.00521253

0.00430428

1.30595

BC

0.0094

0.00345505

1.0

CC

0.0021792

0.00430428

1.30595

Standard errors are based on total error with 31 d.f. This table (6) shows each of the estimated effects and interactions. Also shown is the standard error of each of the effects, which measures their sampling error. Note also that the largest variance inflation factor (V.I.F.) equals 1.30595. For a perfectly orthogonal design, all of the factors would equal 1. Factors of 10 or larger are usually interpreted as indicating serious confounding amongst the effects. To plot the estimates in decreasing order of importance, select Pareto Charts from the list of Graphical Options. To test the statistical significance of the effects, select ANOVA Table from the list of Tabular Options. You can then remove insignificant effects by pressing the alternate mouse button, selecting Analysis Options, and pressing the Exclude button. Table 7: Analysis of Variance for MRR Source Sum of Squares

Df

Mean Square

F-Ratio

P-Value

A:Pressure

0.00733261

1

0.00733261

204.75

0.0000

B:AFR

0.000820125

1

0.000820125

22.90

0.0000

C:SOD

0.00151617

1

0.00151617

42.34

0.0000

AA

0.000192815

1

0.000192815

5.38

0.0271

AB

0.000149108

1

0.000149108

4.16

0.0499

AC

0.0000563333

1

0.0000563333

1.57

0.2191

BB

0.0000525203

1

0.0000525203

1.47

0.2350

BC

0.00026508

1

0.00026508

7.40

0.0106

3314

D.V. Srikanth and M. Sreenivasa Rao / Materials Today: Proceedings 2 (2015) 3308 – 3317

CC

0.0000091796

1

0.0000091796

Total error

0.00111018

31

0.0000358122

Total (corr.)

0.011454

40

0.26

0.6162

R-squared = 90.3075 percent R-squared (adjusted for d.f.) = 87.4936 percent Standard Error of Est. = 0.00598433 Mean absolute error = 0.0035334 Durbin-Watson statistic = 2.03099 (P=0.5224) Lag 1 residual autocorrelation = -0.0302185 The ANOVA table(7) partitions the variability in MRR into separate pieces for each of the effects. It then tests the statistical significance of each effect by comparing the mean square against an estimate of the experimental error. In this case, 6 effects have P-values less than 0.05, indicating that they are significantly different from zero at the 95.0% confidence level. The R-Squared statistic indicates that the model as fitted explains 90.3075% of the variability in MRR. The adjusted R-squared statistic, which is more suitable for comparing models with different numbers of independent variables, is 87.4936%. The standard error of the estimate shows the standard deviation of the residuals to be 0.00598433. The mean absolute error (MAE) of 0.0035334 is the average value of the residuals. The Durbin-Watson (DW) statistic tests the residuals to determine if there is any significant correlation based on the order in which they occur in your data file. Since the P-value is greater than 5.0%, there is no indication of serial autocorrelation in the residuals at the 5.0% significance level. 3. Results & Discussions Optimize Response x Goal: maximize MRR x Optimum value = 0.0939292 Table 8 : Optimum Range of Factors Maximizes Machining Factor Low High Optimum Pressure

4.0

8.0

AFR

3.5

5.5

8.0 5.5

SOD

10.0

20.0

20.0

This table (8) shows the combination of factor levels which maximizes MRR over the indicated region. Use the Analysis Options dialog box to indicate the region over which the optimization is to be performed. You may set the value of one or more factors to a constant by setting the low and high limits to that value. Contours of Estimated Response Surface SOD=15.0 0.08

5.5 5.1

0.07

AFR

4.7 0.03 4.3 3.9

MRR 0.0 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08 0.09 0.1

0.04 0.05

0.06

3.5 4

5

6 Pressure

7

8

Fig 3 :Contour of Estimated Response surface

D.V. Srikanth and M. Sreenivasa Rao / Materials Today: Proceedings 2 (2015) 3308 – 3317 Estimated Response Surface SOD=15.0

0.1

MRR

0.08 0.06 0.04 0.02 0 4

5

6

7

Pressure

8

3.5

3.9

4.3

4.7

5.1

5.5

AFR

Fig 4: Estimated Response Surface

Normal Probability Plot for MRR 99.9 99

percentage

95 80 50 20 5 1 0.1 -3

0

3

6 Standardized effects

9

12

15

Fig 5 :Normal Probability plot for MRR

Standardized Pareto Chart for MRR

A:Pressure

+ -

C:SOD B:AFR BC AA AB AC BB CC 0

3

6 9 Standardized effect

12

Fig 6 : Standardized Pareto chart for MRR

15

3315

3316

D.V. Srikanth and M. Sreenivasa Rao / Materials Today: Proceedings 2 (2015) 3308 – 3317

Figure no 3,4,5,6 represents Response plots, Pareto chart and Probability plots of the influence of Factors on MRR .The combination of the Processes and their standardised effect on Metal Removal Rate are clearly indicated in Pareto chart (fig 6).In Estimate Response surface Graph(fig 4) it is Observed that by increase in pressure the MRR is also increased. Table (7) shows Analysis of Variance for MRR. F Value (204.75) of the parameter indicates the Air Pressure is significantly contributing more towards cutting performance. F value (22.90)of parameter indicates the Contribution of SOD is less comparing with Pressure and AFR. The reviews highlights that the effect of Nozzle diameter on MRR.As increase in diameter of the Nozzle will increase the MRR mean while the variation in depth of cut may be increased.

Fig 7 : Ceramic tiles drilled at different Pressure, SOD, and AFR

Fig (7) indicates the specimens of ceramic tiles drilled at various Pressures, SOD’s and AFR .These experiments are conducted on the Test rig at the workshop of St. Martin’s Engineering, Hyderabad 4.

Conclusions: Taguchi Method of Design was initially applied for investigating the effect of different parameters on Metal Removal Rate. For the Analysis Design of experiment of MINI TAB was used and analyzed the optimal value for Large MRR. Nine Experiments were conducted according to the levels on Taguchi and obtained the S/N Ratio. Optimized Response in Taguchi is shown in the plots of S/N Ratio (fig3)The optimal values are Pressure(8 kg/cm2),AFR(5.5 gm/min),SOD (20 mm).On the other side by using STARGRAPHICS Centurion 41 levels of Experiments were conducted with RSM and analyzed by using Analysis of Variance (ANOVA).The Optimum Range of machining(table 4.5)of Stat graphics ANOVA is similar to the results obtained by Taguchi as well RSM. Further there is much scope in developing the statistical modelling on Surface characteristics, Kerf Analysis by applying various types of new and promising software’s.

D.V. Srikanth and M. Sreenivasa Rao / Materials Today: Proceedings 2 (2015) 3308 – 3317

References [1] [2] [3] [4] [5] [6] [7] [8] [9]

U.D.Gulhani,P.P.Patkar,S.P.Patel,A.A.Patel et al , Analysis of AJM Parameters on MRR,Kerf width of Hard and Brittle Materials like ceramics,--IJDMT-April-2013. A. P. Verma and G. K. Lal Publication(1984)“An experimental study of abrasive jet machining”, International Journal of Machine Tool Design and Research, Volume 24, Issue 1,pp 19-29. Chandra, Bhaskar, Singh Jagtar , A Study of effect of Process Parameters of Abrasive jet machining, International Journal of Engineering Science and Technology 01/2011 vol 3 pp 504-513. .Roy, R.K.: Design of experiments using the Taguchi approach: 16 step sto product and process improvement, Wiley-Inter science, New York, 2001. Cukor, G., Jurković, Z., Sekulić, M.: Rotatable Central Composite Design of Experiments versus TaguchiMethod in the Optimization of Turning, Metalurgija/Metallurgy, 50(2011)1, 17-20. Tsai, F.C., Yan, B.H., Kuan, C.Y., Huang, F.Y.(2008). Taguchi and experimental investigation into the optimal processing conditions for the abrasive jetpolishing of SKD61 mold steel, International Journal ofMachine Tools & Manufacture, Vol. 48, Nos. 7-8, pp.932–945. R. Balasubramaniam, J. Krishnan and N.Ramakrishnan(1999) , “An experimental study onthe abrasive jet deburring of cross drilled holes” ,Publication: Journal of Materials ProcessingTechnology, Volume 91, Issues 1-3, pp 178-182. R. Balasubramaniam, J. Krishnan and N.Ramakrishnan(2002) “A study on the shape ofthe surface generated by abrasive jet machining”,Publication: Journal of Materials Processing Technology, Volume 121, Issue 1, 14February 2002, pp 102-106. A. P. Verma and G. K. Lal Publication(1984)“An experimental study of abrasive jet machining”, International Journal of Machine Tool Design and Research, Volume 24, Issue 1,pp 19-29.

3317