Influence of Process Parameters on Wire EDM Process for AISI 316 Stainless Steel

Transcription

1 Ifluece of Process Parameters o Wire EDM Process for AISI 316 Stailess Steel 1 K. R. Padmavathi, 2 S. Devaraj, 3 I. Joh Solomo, 4 A. Premkumar 1,2 Asso. Prof., Dept. of Mechaical Egg., Paimalar Egg. College, Cheai , Idia. Abstract The optimizatio of wire electrical discharge machiig (WEDM) parameters for AISI 316 stailess steel usig Taguchi s robust desig approach is proposed. Experimetatio is plaed as per Taguchi s L27 orthogoal array. Each experimet has to be performed uder differet process parameters of pulse o, pulse off, pulse peak curret, wire tesio ad wire feed rate. Two resposes amely material removal rate ad surface roughess (Ra) were cosidered for each experimet. The optimum machiig parameter combiatio is obtaied by usig the aalysis of sigal-to-oise (S/N) ratio. The level of importace of the machiig parameters o the material removal rate ad surface roughess is determied by usig aalysis of variace (ANOVA). Keywords: WEDM, ANOVA, MRR, Surface roughess, Sigal to oise ratio I. INTRODUCTION As ewer ad more exotic materials have bee developed i the past few decades, covetioal machiig operatios ted to reach their limitatios as relatively more complicated shaped jobs are required to be maufactured. The icreased use of wire-cut electrical discharge machiig (WEDM) i maufacturig has thus kept growig at a highly accelerated rate sice its first idustrial applicatio more tha 30 years ago. Its broad capabilities have allowed it to ecompass productio i aerospace ad automotive idustries ad virtually all areas of coductive material machiig. This is because wire EDM provides the best alterative, or somes the oly alterative, for machiig coductive, exotic ad HSTR (high stregth ad temperature resistive) material with the scope of geeratig itricate shapes ad profiles. Also, WEDM has proved to have icredible prospective i its applicability i the preset day metal cuttig idustry for achievig a sigificat dimesioal accuracy, surface fiish ad cotour geeratio trait of products or parts. Wire EDM is a special form of the traditioal EDM process i which the electrode is a cotiuously movig coductive wire. Material is wor from the work piece by a series of distict sparks betwee the work piece ad the wire electrode (tool) estraged by a thi film of dielectric fluid (deioised water) which is cotiuously force fed to the machiig zoe to flush away the eroded particles. The progress of the wire is proscribed umerically to attai the preferred three-dimesioal shape ad accuracy for the work piece. Although the average cuttig speed, relative machiig costs, accuracy ad surface fiish have bee improved may s sice the commercial iceptio of the machie, much more improvemet is still required to meet the icreasig demad of precisio ad accuracy by differet idustries. II. LITERATURE REVIEW To improve the performace amely the material removal rate, cuttig speed, dimesioal accuracy ad surface roughess of the WEDM process, several researches have attempted previously. Ramakrisha et al. [7] used Multi Respose optimizatio procedure for studyig WEDM operatios usig Taguchi s robust desig of experimets. Experimets were plaed as per Taguchi s L16 Orthogoal array. Each experimet has bee performed uder differet cuttig coditios of pulse o ; wire tesio, delay, ad wire feed speed ad igitio curret itesity. Three resposes amely material removal rate, surface roughess ad wire wear ratio (WWR) had bee cosidered for each experimet. Liao et al [3] carried out the study o the machiig parameters optimizatio of WEDM i SKD 11 alloy steel. They used 0.25mm diameter brass wire as electrode. Accordig to Taguchi quality desig cocept L 18 orthogoal arrays table was chose for coductig experimets. A total of six machiig parameters like pulse o, pulse off, table feed, wire tesio, wire speed ad flushig pressure were chose for the cotrollig factors ad each parameter have three levels ad the desiged machiig coditios are evaluated i terms of the measured machiig performaces like gap width, material exclusio rate, surface coarseess, discharge frequecy, gap voltage, ormal discharge frequecy ratio. Chiag et al [1] used grey relatioal aalysis for their study o alumiium oxide particle reiforced material with multiple performace characteristics. Ho et al [2] reviewed the WEDM research cocerig the optimizatio of the process parameters survey the ifluece of the various factors distressig the machiig performace ad yield. Mahabatra et al.[4] optimized the WEDM process parameters usig Taguchi method. I this study the relatioship betwee cotrol factors like discharge curret, pulse duratio, pulse frequecy, wire speed, wire tesio ad dielectric flow ad resposes like material removal rate, surface roughess ad Kerf were established by meas of o liear regressio aalysis. Shaja Kuriakose et al. [11] coducted multi objective optimizatio of WEDM process by No Domiated Sortig Geetic Algorithm techique. Ozdemir et al. [5] performed exploratio o machiability of odular cast iro (GGG 40) by WEDM. Variatio of the surface roughess ad cuttig rate with machiig parameters were scietifically modeled by usig the regressio aalysis method. Puri et al [6] made a attempt to

2 model the white layer depth through respose surface methodology. Tarag et al. [10] used feed forward eural etwork approach to determie the optimal cuttig parameters while machiig SUS-304 stailess steel i WEDM. Speedig et al [10] performed the modelig process through respose surface methodology ad artificial eural etworks. A respose surface model based o a cetral composite rotatable experimetal desig, ad size back propagatio eural etworks have bee developed. Speedig et al [10] performed the modelig process through respose surface methodology ad artificial eural etworks. A respose surface model based o a cetral composite rotatable experimetal desig, ad size back propagatio eural etworks have bee developed. Shaja kuriakose et al [11] coducted multi objective optimizatio of WEDM process by o domiated sortig geetic algorithm techique. They established that ifluece of the parameters o the cuttig velocity ad the surface fiish are quiet cotrary. Tarag et al [13] used feed forward eural etwork approach to determie the optimal cuttig parameters while machiig SUS 304 stailess steel i WEDM. The disparity of surface roughess ad MRR with machiig parameters ad optimizatio of machiig settig for miimum surface roughess ad maximum MRR should be ivestigated experimetally ad the obtaied results should be iterpreted ad modeled statistically to uderstad closely the behavior of machiig rate ad accuracy i WEDM. I this cram, the upshot of the machiig parameters ad their level of cosequece o the surface roughess ad the MRR are statistically evaluated by usig aalysis of variace (ANOVA). Also, a optimizatio study is itroduced for the case of low surface roughess ad high MRR. Experimets were coducted udereath diverse machiig parameters, amely, pulse o, pulse off, discharge curret, wire tesio ad wire feed rate. The settigs of machiig parameters were determied by usig Taguchi experimetal desig method. III. EXPERIMENTAL DETAILS A. Machie tool, material ad Measuremet The experimetal studies were performed o a ELEKTRA OPTICUT 434 WEDM machie tool. Differet settigs of pulse o, pulse off, discharge curret, wire tesio ad wire feed rate were used i the experimets as specified i Table TABLE 1 EXPERIMENTAL SETTINGS TABLE 2 CHEMICAL COMPOSITION OF AISI 316 STAINLESS STEEL Elemets C Cr Ni Mo M Si P Fe % (Wt.) Balace The mea cuttig speed data (Cs) was observed directly from the machie tool moitor. Geerally durig this process the wire diameter is kept costat. Therefore, the width of cut (W) remais costat. The MRR for the WEDM operatios usig Eq. 1 which is show below: MRR = Cs x L (1) Where Cs = cuttig speed i mm/mi., L = thickess of the material i mm. A Talysurf at 0.8mm cut-off value was applied to measure the surface roughess (Ra) of each specime. B. Desig of experimet based o Taguchi method To evaluate the effects of machiig parameters o performace characteristics (surface roughess ad MRR), ad to idetify the performace characteristics uder the optimal machiig parameters, a specially desiged experimetal procedure is required. Covetioal ivestigatioal desig methods are too complex ad difficult to use. Additioally, large umber of experimets has to be carried out whe umber of machiig parameters icreases. I this study, Taguchi method, a powerful tool for parameter desig of performace characteristics, was used to determie optimal machiig parameters for miimum surface roughess ad maximum MRR i WEDM. I Taguchi method, process parameters which ifluece the products are separated ito two mai groups: cotrol factors ad oise factors. The cotrol factors are used to select the best coditios for stability i desig of maufacturig process, whereas the oise factors deote all factors that cause variatio. Taguchi projected to get hold of the characteristic data by meas of orthogoal array, ad to explore the recital measure from the data to decide the optimal process parameters. This scheme uses a uusual pla of orthogoal arrays to cram the whole costrait space with petite umber of experimets oly. I this study, five machiig parameters were used as cotrol factors ad each parameter was desiged to have three levels, deoted 1, 2 ad 3 [14, 15, 16]. Accordig to the Taguchi quality desig cocept, a L27 orthogoal array table with 27 rows, aalogous to the umber of experimets as preferred for the experimets ad the cosequeces are preseted i Table 3. Sl. No. Cotro l factors 1 Pulse o 2 Pulse off 3 Pulse peak curre t 4 Wire tesio Uit Rage Level Symbol Actual Deci ded µsec A µsec B amps C gm D Volume Wire 6, Issue m/mi 02 E Published 5 by, feed rate

3 L 27 (3 13 ) TABLE 3 IV. ANALYSIS AND DISCUSSION OF EXPERIMENTAL RESULTS The aalysis of variace (ANOVA) was used to foud statistically oteworthy machiig parameters ad the percet cotributio of these parameters o the surface roughess ad the MRR. I Taguchi method [8], a loss fuctio is used to estimate the variatio amog the ivestigatioal value ad the preferred value. This loss fuctio is further malformed i to a sigal-to-oise (S/N) ratio. There are umerous S/N ratios existig depedig o the type of characteristics; lower the better (LB), omial is the best (NB), ad higher is better (HB). I WEDM, the lower surface roughess ad higher MRR are the sig of better performace. For the HB ad LB, the defiitio of the loss fuctio (L) for machiig performace results (MRR, surface roughess) of repetitive umber are : L L HB LB A Pulse o µsec 1/ 1/ i1 i1 B Pulse off µsec 1/ Y Y 2 SF 2 MRR C Pulse peak Curret amps D Wire tesio gm E Wire feed rate m/mi Where Y MRR ad Y SF are the respose for material removal rate ad surface fiish correspodigly ad (2) (3) M RR m m 2 / mi SR (µm) ORTHOGONAL ARRAY FOR L27(313) TAGUCHI DESIGN deotes the umber of experimets. The S/N ratio ca be calculated as a logarithmic coversio of the loss fuctio as show below: S/N ratio for MRR = -10 log 10 (LHB) (4) S/N ratio for SF = -10 log 10 (LLB) (5) Despite the cosequeces of the sort of the performace characteristics, a greater S/N value correspods to a better performace. Therefore, the best possible level of the machiig parameters is the level with the greatest S/N ratio value. By applyig the Eqs. (2) (5), the S/N ratio values for each experimet of L27 (Table 3) was calculated (Table 4 ad 5). Based o the aalysis of S/N ratio, the optimal machiig performace for the MRR was obtaied at 6 µsec. pulse o (Level 3), 4 µsec. pulse off (Level 1), 5 amps. Pulse peak curret (Level 3), 1000 g. wire tesio (Level 3) ad 4 m/mi. wire feed rate (Level 2) settigs. Fig. 1 shows the effect of machiig performace o the MRR. The optimum machiig performace for the surface fiish was obtaied at 5 µsec. pulse o (Level 2), 4 µsec. pulse off (Level 1), 2 amps. Pulse peak curret (Level 1), 800 g. wire tesio (Level 2) ad 3 m/mi. wire feed rate (Level 1) settigs. Fig. 2 shows the upshot of machiig performace o the surface fiish. Machiig parameter Overall mea = db TABLE 4 S/N RATIO VALUES MRR Mea S/N ratio by factor level (db) Level 1 Level 2 Level 3 A B C D E Machiig parameter TABLE 5 S/N RATIO VALUES FOR SURFACE FINISH Mea S/N ratio by factor level (db) Level 1 Level 2 Level 3 A B C D E Overall mea = db A better feel for the relative effect of the differet machiig parameters o the MRR ad the surface fiish was obtaied by disitegratio of variace, which is called aalysis of variace (ANOVA). The virtual sigificace of the machiig parameters with respect to the MRR ad the surface fiish was ivestigated to determie more accurately the optimum combiatios of the machiig

4 S/N ratio for MRR S/N ratio for Ra parameters by usig ANOVA. The results of ANOVA for the machiig outputs are presetedi Tables 6 ad 7. Statistically, F-test provides a verdict at some cofidece level as to whether these estimates are sigificatly differet. Larger F-value idicates that the variatio of the process parameter makes a big chage o the performace characteristics. F-values of the machiig parameters are compared with the fittig cofidece table. Accordig to this aalysis, the most effective parameter with respect to MRR is pulse peak curret. Pulse o ad pulse off are foud as statistically sigificat factors, whereas the effect of wire tesio ad wire feed rate o the MRR was isigificat. Percet cotributio idicates the relative power of a factor to reduce variatio. For a factor with a high percet cotributio, a small variatio will have a great ifluece o the performace. The percet cotributios of the machiig parameters o the MRR are show i Table 6. Accordig to Table 6, pulse peak curret was foud to be the major factor affectig the MRR (67.089%), whereas pulse o was foud to be the secod rakig factor (11.573%) ad pulse off was start to be the third grade reaso (10.718). The percet ivolvemet of wire tesio ad wire feed rate are mior, beig % ad 4.642% correspodigly. Accordig to F-test aalysis, the mometous parameters o the MRR are ope circuit voltage, pulse o ad pulse off. Durig the above examiatio the optimal machiig parameters idetified for MRR are A3B1C3D3E2. The percet cotributios of the machiig parameters o the surface fiish are show i Table 7. Accordig to Table 7, pulse o is foud to be the major factor affectig the surface fiish (44.667%). The pulse peak curret, wire tesio ad pulse off are cosidered as sigificat factors with the percet cotributio o the surface fiish are , ad % respectively. The wire feed rate cosidered statistically isigificat, because it has a very low F value ad percet cotributio (2.3958%). Therefore this is icluded i the error term. Based o the above results the optimal cuttig coditios obtaied for surface fiish are A2B1C1D2E Pulse o Curret Wire feed speed off Wire tesio Grad mea Machi ig Param eters Pulse o Curret Wire feed speed Parameter level Off Wire tesio Grad mea A1 A2 A3 B1 B2 B3 C1 C2 C3 D1 D2 D3 E1 E2 E3 Figure 2. The effect of machiig parameters o surface fiish Degre es of freedom TABLE 6 RESULTS OF ANOVA FOR MRR Sum of Mea A B C D TABLE 7 RESULTS OF ANOVA FOR SURFACE FINISH F test 2.4 % Cotributio E Machiig ers Paramet ees Degr of freedom A 2 B 2 C 2 D 2 E 2 of Sum Mea test F 18.6 % Cotributio A1 A2 A3 B1 B2 B3 C1 C2 C3 D1 D2 D3 E1 E2 E3 Parameter level Figure 1. The effect of machiig parameters o MRR

5 V. CONFIRMATION EXPERIMENT The cofirmatio experimet is the fial step of the Taguchi s desig of experimet process after selectig the optimal parameters. The purpose of the cofirmatio experimet is to validate the coclusios draw durig the aalysis phase. I this learig, behid determiig the most favorable situatio ad predictig the rejoider uder these coditios, a ew experimet was plaed ad coducted with the optimum levels of the machiig parameters. The cocludig step is to evisage ad validate the ehacemet of the performace characteristic. The predicted S/N ratio η opt usig the best possible levels of the machiig parameters ca be calculated as: opt m p i1 ( ) i m where η m is total mea of S/N ratio, η opt is the mea of S/N ratio at the optimal level, ad p is the umber of chief machiig parameter that appreciably cocer the performace. The results of experimetal cofirmatio usig optimal machiig parameters are show i Tables 8 ad 9. Table 8 shows the assessmet of the predicted MRR with the authetic MRR usig most favorable machiig parameters. The improvemet i S/N ratio from the startig machiig parameters to the level of optimal machiig parameters is db. The MRR is icreased by s. So, the MRR is greatly improved by usig the approach. Table 9 shows the compariso of the predicted surface fiish with the actual surface fiish usig the optimal machiig parameters. The ehacemet i S/N ratio from the startig machiig parameters to the optimal machiig parameters is db. The surface roughess is decreased by s. The experimetal results iveterate the validity of used Taguchi method for ehacig the machiig performace ad optimizig the machiig parameters. The MRR ad the surface fiish are sigificatly ehaced by usig the approach. TABLE 8 RESULTS OF THE CONFIRMATION EXPERIMENT FOR MRR Iitial Optimal cuttig parameters Level Material removal (mm2/mi) (db) S/N rate ratio cuttig 2E2 parameters A2B2C2D 3E2 Predictio A3B1C3D 3E Experimet A3B1C3D Improvemet i S/ ratio for MRR = db TABLE 9 RESULTS OF THE CONFIRMATION EXPERIMENT FOR SURFACE FINISH Level Iitial cuttig 2E2 parameters A2B2C2D 2E1 Optimal cuttig parameters Predictio A2B1C1D 2E1 Experimet A2B1C1D Surface roughess (µm) S/N ratio (db) (6) Improvemet i S/N ratio for surface fiish = db VI. CONCLUSION This paper has offered a exploratio o the optimizatio ad the cosequece of machiig parameters o the MRR ad the surface fiish i WEDM operatios. The outcome of various machiig parameter such as pulse o, pulse off, pulse peak curret, wire tesio, ad wire feed speed has bee premeditated durig the machiig of AISI 316 stailess steel. The level of cosequece of the machiig parameters o the MRR ad the surface fiish is determied by usig ANOVA. Based o ANOVA method, the greatly effectual parameters o both the MRR ad the surface fiish were foud as pulse peak curret ad pulse o other tha the added parameters cosidered i this study. The results showed that pulse peak curret was about six s imperative tha the secod graded factor (pulse o ) for cotrollig the MRR, whereas pulse o for cotrollig the surface fiish was about two s more sigificat tha the secod rakig factor (pulse peak curret). A optimum parameter bled for the maximum MRR ad miimum surface roughess was obtaied by usig the aalysis of sigal-to-oise(s/n) ratio. The cofirmatio tests idicated that it is probable to augmet MRR ad dimiish surface roughess appreciably by usig the projected statistical techique. The experimetal results iveterate the validity of the used Taguchi method for ehacig the machiig performace ad optimizig the machiig parameters i WEDM operatios. REFERENCES [1] Chiag Ko Ta, Fu-Pig Chag, Optimizatio of the WEDM process of particle reiforced material with multiple performace characteristics usig grey relatioal aalysis, J of Matrl. Proces. Tech., 2006, Vol 180, pp [2] Ho K H, Newma S.T, State of the art i wire electrical discharge machiig, It J of Mach tools ad Mfg., 2004, Vol 44, pp [3] Liao Y.S., Huag J.T., Su H.C. A study o the machiigparameters optimizatio of wire electrical discharge machiig Joural of Materials processig techology, 1997, Vol 71, pp [4] Mahapatra.S.S., Amar Pataik, Optimizatio of wire electrical discharge machiig process parameters usig Taguchi method It J Adv Mauf Techol., 2006, Vol.34, pp [5] Ozdemir N, Cebli Ozek, A ivestigatio o machiability of odular cast iro by WEDM, It J Advaced Mfg Tech., 2005, Vol 149, pp

6 [6] Puri A B, Bhattacharya.B, Modelig ad aalysis of white layer depth i wire cut EDM process through respose surface methodlogy, It J of Advaced Mfg Techology, 2005, Vol. 25, pp [7] Ramakrisha R, Karuamoorthy L, Multi respose optimizatio of wire EDM operatios usig robust desig of experimet, It Joural of Adv Mfg. Tech., 2005, Vol 29, pp [8] Ross J Phillip, Taguchi Techique for Quality Egieerig, 1998, McGraw Hill Ic. [9] Sarkar S, Mitra S, Bhatacharyya B, Parametric optimizatio of wire electrical discharge machiig of titaium alumiide alloy through a artificial eural etwork model, It J of Adv Mfg Tech., 2005,Vol 27, pp [10] Speedig T.A, Wag Z.G. Parametric optimizatio ad surface characteristics of WEDM process, 1997, Precisio Egg., Vol 20, pp [11] Shaja Kuriakose, Shamugam M.S, Multi-objective optimizatio of wire electro discharge machiig process by No-domiated sortig geetic algorithm, J of Material process techology, 2005, Vol 170, pp [12] Shamugam M S, P Harihara, Gowri S, Optimizatio of wire electrical discharge machiig based o Taguchi method, 2006, proceedigs of AMPC. [13] Tarag Y S ad Ma S.C, Determiatio of optimal cuttig parameters i WEDM, It j of Mach Tools ad Mfg., 1995, Vol 35, pp [14] G.Ugrasea, H.V.Ravidra, G.V.NaveePrakash, Y.N.Theertha Prasad, Optimizatio of Process Parameters i Wire EDM of HCHCr Material Usig Taguchi's Techique, Materials Today: Proceedigs, 2015, Vol.2, pp [15] Ravidraadh Bobbili, V.Madhu, A.K.Gogia, Multi respose optimizatio of wire-edm process parameters of ballistic grade alumiium alloy, Egieerig Sciece ad Techology, a Iteratioal Joural, 2015, Vol. 18, pp [16] Sameh Habib, Optimizatio of machiig parameters ad wire vibratio i wire electrical discharge machiig process, Mechaics of Advaced Materials ad Moder Processes, 2017, 3, pp. 1-9.