Index Terms Wire EDM, MRR, Surface roughness, Regression, Taguchi method, ANOVA, Fractional Factorial.

ISSN: 39-5967 Experimental Investigations of HSS M Alloy on Wire Electric discharge machining Process Using Taguchi Methodology K.Hari Narayana,,Dr.B.Balu Naik, Srinivasa R Nandam 3 and A.Anand Rao 4 Research Scholar, Jawaharlal Nehru Technological University, Hyderabad - 50007 Principal & Professor, Jawaharlal Nehru Technological University 3, 4 Scientists, Mechanical Engineering Group, DMRL, DRDO, Hyderabad 500058 Abstract In this present work, investigations of cutting performance with pulse on time, pulse off time, servo voltage, wire feed, Current and Cutting speed were experimentally investigated in wire electric discharge machining (WEDM) process. Brass wire with 0.5mm diameter and HSS M alloy with 0mm dia were used as tool and work materials in the experiments. The cutting performance outputs considered in this study are material removal rate (MRR) and surface roughness. Experimentation has been completed by using Taguchi s L8 ( 5 ) orthogonal array under different conditions of parameters. Optimal combinations of parameters were obtained by this technique. This study shows that the complete problem can be solved with the minimum number of experiments when compared to full factorial design. The results obtained are analyzed for the selection of an optimal combination of WEDM parameters for proper machining of HSS M alloy to achieve better surface finish. In addition the importance of the cutting parameters on the cutting performance outputs is determined by using analysis of variance (ANOVA). Index Terms Wire EDM, MRR, Surface roughness, Regression, Taguchi method, ANOVA, Fractional Factorial. I. INTRODUCTION The demands for alloy materials having high hardness, toughness and impact resistance are increasing today in the growth of mechanical industry. Non-traditional machining methods including electrochemical machining, ultrasonic machining, electrical discharging machining (EDM) etc. are applied to machine alloy materials in order to overcome the difficulties in traditional machining methods. WEDM process with a thin wire as an electrode transforms electrical energy to thermal energy for cutting materials. With this process alloy steel, conductive ceramics and aerospace materials can be machined irrespective to their hardness and toughness. Furthermore, WEDM is capable of producing a fine, precise, corrosion and wear resistant surface. WEDM is considered as a unique adoption of the conventional EDM process, which uses an electrode to initialize the sparking process []. However, WEDM utilizes a continuously travelling wire electrode made of thin copper, brass or tungsten of diameter 0.05-0.30 mm, which is capable of achieving very small corner radii. The wire is kept in tension using a mechanical tensioning device reducing the tendency of producing inaccurate parts. During the WEDM process, the material is eroded ahead of the wire and there is no direct contact between the work piece and the wire, eliminating the mechanical stresses during machining. II. LITERATURE REVIEW WEDM is an essential operation in several manufacturing processes in some industries, which gives importance to variety, precision and accuracy. Several researchers have attempted to improve the performance characteristics namely the surface roughness, cutting speed, dimensional accuracy and material removal rate [].The machining variables included pulse-on time, pulse-off time, pulse-peak current, servo voltage and servo feed. The variables affecting the surface roughness were identified using ANOVA technique. Results showed that pulse-on time and pulse-peak current were significant variables to the surface roughness of wire-edm HSS M Alloy. The maximum prediction error of the model was less than 7% and the average percentage error of prediction was less than 3%. Ramakrishnan and Karunamoorthy (008) developed artificial neural network (ANN) models and multi response optimization technique to predict and select the best cutting parameters of wire electrodischarge machining (WEDM) process []. HSS M alloy is selected as work material to conduct experiments 330

ISSN: 39-5967 and brass wire of 0.5mm diameter used as a tool electrode. Experiments were planned as per Taguchi L-8 orthogonal array using fractional factorial design. Experiments were performed under different cutting conditions of pulse on time, pulse off time, wire feed, speed and ignition current. It was found that the pulse on time, delay time and ignition current had more influence than wire feed speed on the performance characteristics considered in this study. MRR is improved with increase in pulse on time and ignition current. But the surface quality of the work specimen is affected adversely with the increased value of pulse on time and ignition current. Pradhan et. al. (009). Optimized micro-edm process parameters for machining Ti-6Al-4V super alloy. The influence of machining process parameters such as peak current, pulse-on-time, dielectric flushing pressure and duty ratio on performance criteria like MRR, TWR, over cut and taper have been examined[3] Manna and Kumar (009) investigated the effects of various cutting parameters of WEDM on wire crater depth, electrode wear rate and surface roughness using Taguchi methods based on L8 mixed orthogonal array[4] A. Taguchi Method III. METHODOLOGY The quality engineering method proposed by Taguchi is commonly known as the Taguchi method or Taguchi approach. This approach provides a new experimental strategy in which a modified and standardized form of design of experiment (DOE) is used. In other words, the Taguchi approach is a form of DOE with special application [5]. The concept of the Taguchi method is that the parameter design is performed to reduce the sources of variation on the quality characteristics of product, and reach a target of process robustness Taguchi design of experiments using specially constructed tables known as orthogonal array ( OA ). It utilizes the orthogonal arrays from experimental design theory to study a large number of variables with a small number of experiments [6]. This technique helps to study effect of many factors (variables) on the desired quality characteristic most reasonably. By studying the effect of individual factors on the results, the best factor combination can be determined. The standardized Taguchi-based experimental design used in this study is an L8 orthogonal array. B. Signal-to-Noise ratios (S/N ratio) In the Taguchi method, the term signal represents the desirable value (mean) for the output characteristic and the term noise represents the undesirable value (S.D) for the output characteristic [3]. Therefore, the S/N ratio is the ratio of the mean to the S.D. S/N ratio is used to measure the quality characteristic deviating from the desired value (8-9). The S/N ratio η is defined as. Larger the Better:. Smaller the Better: Where n = no of repetition --------------------------() -----------------------------() IV.EXPERIMENTAL SETUP Experiments have been performed on five axes CNC Wire cut EDM (ULTRACUT from ELECTRONICA)[] at Defence Metallurgical Research Laboratory, Hyderabad(AP). The photographic view of CNC Wire cut EDM and experimental set-up are shown in Figure. 33

ISSN: 39-5967 Fig : An Experimental Set up of CNC Wirecut EDM A 0.5 mm diameter brass wire is used in this experiment as a cutting tool and HSS M grade alloy steel in form of a round rod 0mm dia mounted on WEDM machine tool and specimens of 5 mm length is cut according to Taguchi L8 design. Chemical composition of material is given in Table. Table : Chemical composition of HSS M alloy steel (%) C Si Mn P S Cr Mo Ni V W HSS M 00.84 0.39 0.308 0.06 0.0 04.8 04.86 0.0 0.88 06.07 A. Material Removal Rate (MRR) Fig : Weight Balance The mean cutting speed data (Cs) is observed directly from the computer monitor who is directly attached to the machine tool. Generally, during this process the wire diameter is kept constant. Therefore, the width of cut (W) 33

ISSN: 39-5967 remains constant. Therefore, the MRR for the WEDM operation is calculated using Eq. 3 [9] which is shown below: MRR = W-W/TIME mg/sec --------------------------------------- (3) Where W= Initial Weight of the work specimen, W= Final weight of the work specimen Precision balance (Fig ) is used to measure the weight of the work piece. This machine capacity [4] is 300 gram and accuracy 0.00 gram with a make Brand: SHINKO DENSHI Co. LTD, JAPAN, and Model: DJ 300S B. Surface Roughness measurement It was measured on Taylor Hobson Talysurf [0] at DMRL ( Fig 3) giving Ra value in microns. Ra is measured along four different lines on the surface and the average value is considered for further analysis. Fig 3: Taylor Hobson Talysurf Instrument V. SELECTION OF CUTTING PARAMETERS Eight machining parameters were selected as control factors.all the five parameters have two levels denoted by and. The experimental design was based on L8 orthogonal array using Taguchi method. Minitab 6 software is used for graphical analysis of the obtained data [7]. Table : Wire EDM parameters and their levels S. No Parameters Symbol Level Level Units Pulse On time T ON 0 0 μsec Pulse Off time TOFF 45 60 μsec 3 Peak Current PC Amp 4 Servo voltage SV 0 30 volt 5 Servo Feed SF 0 40 mm/sec Table 3: Experimental design using L8 orthogonal array S.NO TON (A) TOFF (B) PC (C) SV (D) SF(E) 0 45 0 0 0 45 30 40 3 0 60 0 0 333

ISSN: 39-5967 4 0 60 30 40 5 0 45 0 40 6 0 45 30 0 7 0 60 0 40 8 0 60 30 0 A. Effect of process parameters on MRR VI. RESULTS AND DISCUSSION In order to see the effect of process parameters on the MRR, experiments were conducted using L8 OA (Table 3). The experimental data is given in Table 4. Figure 4 show that the MRR increases with the increase in pulse on time, peak current and decreases with increase in pulse off time and servo voltage [8]. The effects of servo feed on MRR is not very significant. B. Selection of optimal levels Analysis of Variance (ANOVA) table 5[3] shows that the significance of the process variables towards MRR. The response table 6 shows the average of response characteristic for each level of each factor. The tables include ranks based on delta statistics, which compare the relative magnitude of effects. The delta statistic is the highest minus the lowest average for each factor. Minitab assigns ranks based on delta values; rank to the highest delta value, rank to the second highest, and so on. The ranks indicate the relative importance of each factor to the response. The ranks and the delta values show that pulse on time have the greatest effect on MRR and is followed by servo voltage, servo feed, peak current and pulse off time in that order. As MRR is the higher the better type quality characteristic, it can be seen from Figure 4 that the second level of pulse on time (A), first level of pulse off time(b), second level of peak current(c), first level of servo voltage(d), second level of servo feed(e) provide maximum value of MRR. Table 4 Experimental results for MRR and Surface Roughness S.No. MRR (mg/sec) S/N Ratio Surface Roughness S/N Ratio.83.45.04-6.9.6 0.95.839-5.9 3 3.43.963-5.85 4.5.93.855-5.36 5 43.6 3.80 3.067-9.73 6.6 6.9.978-9.47 7 3 30.0.867-9.4 8 6.5 4.3.57-8.0 A General Regression Analysis for MRR Vs. On time, Off Time, Peak Current,Servo Voltage and Servo Feed is described as below General Regression Analysis: MMR versus OnTime, OffTime, Current, Voltage, Feed Regression Equation MMR = -3.6 + 0.098444 OnTime - 0.04877 OffTime + 0.9644 Current -0.08389 Voltage + 0.0673 Feed Coefficients Term Coef SE Coef T P Constant -0.0080644 0.007466 -.08065 0.393 OnTime 0.000393 0.0000609.877 0.49 OffTime -0.0000550 0.0000406 -.3547 0.308 334

Mean of SN ratios ISSN: 39-5967 Current 0.0008854 0.0006088.4549 0.83 Voltage 0.000066 0.0000304 0.54669 0.639 Feed -0.00007 0.0000304 -.5636 0.630 Summary of Model S = 0.000860989 R-Sq = 83.05% R-Sq(adj) = 40.68% PRESS = 0.0000377 R-Sq(pred) = -7.8% Table 5: Analysis of Variance for MRR Source DF Seq SS Adj SS Adj MS F P % Regression 5 0.0000073 0.0000073 0.000005.96008 0.3740 6.6667 OnTime 0.0000039 0.0000039 0.0000039 5.336 0.49404 44.508 OffTime 0.000004 0.000004 0.000004.8356 0.30846 5.6053 Current 0.000006 0.000006 0.000006.497 0.83083 7.9837 Voltage 0.000000 0.000000 0.000000 0.9887 0.639435.543 Feed 0.000000 0.000000 0.000000 0.3766 0.6978.700 Error 0.000005 0.000005 0.0000007 Total 7 0.0000087 Table 6: Response Table for Signal to Noise Ratios (Taguchi Analysis: MRR versus A, B, C, D, E ) Larger is better Level A B C D E -0.3500 3.0558.9798 3.9548.6939 5.7709.365 3.44.466 3.770 Delta 6.09 0.6907.464.4887.033 Rank 5 4 3 Main Effects Plot for SN ratios Data Means 6.0 A B C 4.5 3.0.5 0.0 6.0 D E 4.5 3.0.5 0.0 Signal-to-noise: Larger is better Fig 4: Effect of Control parameters on MRR C. Effect of process parameters on Surface Roughness In order to see the effects of process parameters on the surface roughness, experiments were conducted using L8 OA (Table 3). The experimental data are given in Table 4. It is seen from the Figure 3 that surface roughness decreases with the increase of pulse on time, peak current and surface roughness increases with the increase in servo voltage and pulse off time. There is no significant changes in servo feed rate. 335

ISSN: 39-5967 D. Selection of Optimal Levels Analysis of Variance (ANOVA) table 7 [3] shows that the significance of the process variables towards surface roughness. The response table 8 shows the average of response characteristic for each level of each factor. The Table includes ranks based on delta statistics, which compare the relative magnitude of effects. The delta statistic is the highest minus the lowest average for each factor. Minitab assigns ranks based on delta values; rank to the highest delta value, rank to the second highest, and so on. The ranks indicate the relative importance of each factor to the response. The ranks and the delta values for various parameters show that pulse on time has the greatest effect on surface roughness and is followed by servo voltage, pulse off time, peak current, and servo feed rate in that order. As surface roughness is the lower the better type quality characteristic, from Figure 5 it can be seen that the first level of pulse on time(a), first level of pulse off time(b), second level of peak current (C), first level of servo voltage(d), first level of servo feed (E) result in minimum value of surface roughness. A General Regression Analysis [] for Surface Roughness Vs. On time, Off Time, Peak Current, Servo Voltage and Servo Feed are described as below: General Regression Analysis: Surface Finish versus OnTime, OffTime, Current, Voltage and feed Regression Equation Surface Finish = 0.375 + 0.09465 OnTime - 0.0333 OffTime + 0.365 Current Coefficients - 0.008675 Voltage + 0.00095 Feed Term Coef SE Coef T P Constant 0.37500 0.6954 0.977 0.86 OnTime 0.094650 0.005673 6.6834 0.004 OffTime -0.033 0.00378 -.9436 0.099 Current 0.36500 0.056733.4060 0.38 Voltage -0.008675 0.00837-3.058 0.09 Feed 0.000950 0.00837 0.3349 0.770 Summary of Model S = 0.08035 R-Sq = 99.34% R-Sq(adj) = 97.70% PRESS = 0.0599 R-Sq(pred) = 89.48% Table 7: Analysis of Variance for Surface Finish Source DF Seq SS Adj SS Adj MS F P Regression 5.94569.94569 0.3894 60.45 0.06353 OnTime.797.797.797 78.337 0.003574 OffTime 0.05578 0.05578 0.05578 8.665 0.098630 Current 0.0376 0.0376 0.0376 5.789 0.37896 Voltage 0.0600 0.0600 0.0600 9.353 0.09350 Feed 0.0007 0.0007 0.0007 0. 0.76956 Error 0.087 0.087 0.00644 Total 7.95857 Table 8 Response Table for Signal to Noise Ratios for surface roughness (Taguchi Analysis: Surface Roughness versus A, B, C, D, E) Smaller is better Level A B C D E -5.680-7.674-7.09-7.736-7.435-9.4-7.46-7.6-7.085-7.385 Delta 3.46 0.58 0.403 0.65 0.050 Rank 3 4 5 336

Mean of SN ratios ISSN: 39-5967 Main Effects Plot for SN ratios Data Means A B C -6-7 -8-9 D E -6-7 -8-9 Signal-to-noise: Smaller is better Fig 5 : Effect of Control parameters on Surface Finish VII. CONFIRMATION EXPERIMENT The confirmation experiment is the final step in any design of experiment process. Table 9 and Table 0 show the comparison of the predicted value with the new experimental value for the selected combinations of the machining parameters. As shown in these tables, the experimental values agree reasonably well with predictions because an error of 4.44 % for the S/N ratio of MRR and 7.54 % for the S/N ratio of surface roughness is observed when predicted results are compared with experimental values. Hence, the experimental result confirms the optimization of the machining parameters using Taguchi method for enhancing the machining performance. However, the error in MRR and surface roughness can be further expected to reduce if the number of measurements is increased. Table 9 result of the confirmation experiment for MRR Predicted value Experimental value % error Optimal level ABCDE ABCDE MRR 45.79 40.3 S/N ratio for MRR 6.7768 5.586 4.44 Table 0 result of the confirmation experiment for surface roughness Predicted value Experimental value % error Optimal level ABCDE ABCDE Surface roughness.7944.4 S/N ratio for Surface roughness -4.37546-4.04576 7.54 VIII. CONCLUSION In this work, it is intended to study factors pulse on time, pulse off time, servo voltage, servo feed for maximizations of MRR and minimization of surface roughness in WEDM process using Taguchi Method. Analysis of the results leads to conclude that factors at level ABCDEcan be set for maximization of MRR. 337

ISSN: 39-5967 Similarly, it is recommended to use the factors at levels such as ABCDE for minimization of surface roughness. The effects of pulse on time, pulse off time, peak current, servo voltage and servo feed on MRR and surface roughness were experimentally investigated in machining of HSS M alloy using CNC Wire-cut EDM process. The results of confirmation experiment agree well the predicted optimal settings as an error of 4.44 % is found with MRR. Similarly, an error of 7.54 % was observed for surface roughness ACKNOWLEDGEMENT This work has been carried out by the Defence Metallurgical Research Laboratory, Hyderabad AP. The authors are extremely thankful to Shri G.Malakondaiah, Director who has technically assisted in experimental work. REFERENCES [] Kanlayasiri,K. Boonmung,S. (007), An investigation on effects of wire-edm machining parameters on surface roughness of newly developed DC53 die steel, Journal of Materials Processing Technology, 87 88, 6 9. [] Ramakrishnan, R. and Karunamoorthy,L. (008), Modeling and multi-response optimization of Inconel 78 on machining of CNC WEDM process, Journal of materials processing technology, 0 7, 343 349. [3] Mahapatra, S. S. and Patnaik, A. (007), Optimization of wire electrical discharge machining (WEDM) process parameters using Taguchi method, International Journal of Advanced Manufacturing Technology, 34, 9-95. [4] Manna, A. and Bhattacharyya, B. (006), Taguchi and Gauss elimination method: A dual response approach for parametric optimization of CNC wire cut EDM of PRAlSiCMMC, International Journal of Advanced Manufacturing Technology, 8, 67 75. [5] Sarkar, S. Sekh, M., Mitra, S., Bhattacharyya, B. (007), Modeling and optimization of wire electrical discharge machining of γ-tial in trim cutting operation, Journal of Material Processing Technology, 05, 376-387. [6] Scott, D., Boyina, S., Rajurkar, K.P. (99), Analysis and optimization of Parameter Combination in Wire Electrical Discharge Machining, International Journal of Production Research, 9, 89-07. [7] Rao P.S. et al, Effect of WEDM conditions on surface roughness, a parametric optimization using Taguchi method, IJAEST, 0Vol.No, 6, IssueNo., pp.4-48. [8] Jaganathan P, Naveen kumar T, Dr. R.Sivasubramanian Machining parameters optimization of WEDM process using Taguchi Method, International Journal of Scientific and Research Publications, Volume, Issue, December 0 ISSN 50-353. [9] S Sivakiran, C. Bhaskar Reddy, C. Eswara reddy Effect Of Process Parameters On Mrr In Wire Electrical Discharge Machining Of En3 Steel International Journal of Engineering Research and Applications (IJERA) ISSN: 48-96 www.ijera.com Vol., Issue 6, November- December 0, pp.-6. [0] A Manual on Surface Roughness measurement by Defence Metallurgical Research Laboratory, Hyderabad, A.P. [] Parametric optimization of wire electrical discharge machining (WEDM) process using taguchi methods. S. Mahapatra I ; Amar Patnaik I Journal of the Brazilian Society of Mechanical Sciences and Engineering Print version ISSN 678-5878J. Braz. Soc. Mech. Sci. & Eng. vol.8 no.4 Rio de Janeiro Oct./Dec. 006. [] An Experimental Manual on Installation and Maintenance of Ultracut S CNC Wire cut EDM machine by Electronic Machine Tools Limited. [3] I. K. Chopde, Chandrasekhar Gogte, Dhobe Milind Modeling and Optimization of WEDM Parameters for Surface Finish Using Design of Experiments Proceedings of the 04 International Conference on Industrial Engineering and Operations Management Bali, Indonesia, January 7 9, 04. [4] Installation Manual of Shinko Denshi Precision Balance. 338