1.
Vijayant Maan , Abhishake Chaudhary / International Journal of Engineering Research andApplications (IJERA) ISSN: 2248-9622 www.ijera.comVol. 3, Issue 3, May-Jun 2013, pp.206-216206 | P a g eOptimization of Wire Electric Discharge Machining Process of D-2 Steel using Response Surface MethodologyVijayant Maan , Abhishake ChaudharyLecturer,Department of Mechanical Engineering,S.C.R.I.E.T,C.C.S University Meerut,Uttar Pradesh,IndiaLecturer,Department of Mechanical Engineering,S.C.R.I.E.T,C.C.S University Meerut,Uttar Pradesh,IndiaABSTRACTIn the present study,Response surfacemethodology was used to investigate the effect offour controllable variables on the materialremoval rate(MRR).The work piece material isD-2 tool steel and the four process variables arepulse on time,pulse off time,peak current andservo voltage.These parameters are varied tostudy their effect on the MRR of D-2 steel. Theresponse surface methodology (RSM) inconjunction with central composite design hasbeen used to develop the empirical models forresponse characteristics. Desirability functionshave been used for simultaneous optimization ofperformance measures.It was found that thematerial removal rate (MRR) directly increaseswith increase in pulse on time and peak currentwhile decreases with increase in pulse off timeand servo voltage.Keywords:Pulse on time(Ton),Pulse offtime(Toff),Material removal rate(MRR),Servovoltage(SV),Peak current(IP),Central compositedesign(CCD).I. INTRODUCTIONWire Electric Discharge Machining(WEDM) is a non-traditional process of materialremoval from electrically conductive materials toproduceparts with intricate shapes and profiles. This processis done by using a series of spark erosion. Thesesparks are produced between the work piece and awire electrode (usually less than 0.30 mm diameter)separated by a dielectric fluid and erodes the workpiece to produce complex two and three dimensionalshapes according to a numerically controlled pre-programmed path. The sparks produce heating andmelt work piece surface to form debris which is thenflushed away by dielectric pressure. During thecutting process there is no direct contact betweenthe work piece and the wire electrode. The wireelectrical discharge machining (WEDM) hasbecome an important non-traditional machiningprocess because it can machine the difficult-to-machine materials like titanium alloys andzirconium which cannot be machined byconventional machining process.Though EDM process is very demanding but themechanism of process is very complex,therefore,it istroublesome to establish a model that can accuratelypredict the performance by co-relating the processparameters.The optimum processing parameters arevery essenstial to establish to boost up theproduction rate to a large extent.In this paper amodel is developed by using RSM methodology andCentral Composite Design.II. LITERATURE REVIEWAuthor Parameters Factors Material Method Remark1.Liao etal.MRR,SR andGap WidthTon, Toff, peakcurrent and tablefeedSKD11 alloysteelTaguchimethodPulse-on time have asignificant influence on themetal removal rate and gapewidth2.Jangraet al.CS, SR valueand dimensionallagTon, Toff, Peakcurrent, Wire speedand Wire tensionD3 tool steel Taguchiand GreyRelationalAnalysisThe cutting speed wasobserved 3.80mm/min andsurface finish was not good.3.Mahapatraet al.MRR and SR Discharge current,pulse duration,dielectric flow rateD2 tool steel TaguchimethodMathematical models weredeveloped for optimizationof MRR and surface finishusing non linear regressionmethod4.Kanlaya SR Ton,Toff,wire DC53 die Full Surface roughness increased
2.
Vijayant Maan , Abhishake Chaudhary / International Journal of Engineering Research andApplications (IJERA) ISSN: 2248-9622 www.ijera.comVol. 3, Issue 3, May-Jun 2013, pp.206-216207 | P a g esiri et al. tension steel factorialmethodwhenTon peak currentincreased.5.Tosunet al.Wire electrodewear,MRR andSRPulse duration andopen circuitvoltageAISI 4140steelRegressi-onanalysistechniqueIt was observed that, thewear increases with theincrease in pulse durationand open circuit voltage6.Sanchez etal.Accuracy ofWEDM cornercuttingInfluence of workthickness,cornerradius and numberof trim cutsAISI D2tool steelFuzzy logic It was found that wire lag isresponsible for the back-wheel effect in cornercutting7.Haddad etal.MRR Power, voltage,Toff, and spindlerotational speedAISI D3 toolsteelDOE andRSMIn order to obtain high MRRthe Ton and voltage shouldbe fixed high.8.Guo et al. Cutting Speed Voltage, themachining currentand the pulseinterval.Reinforc-edmatrial(Al2O3 in6061 Alalloy)Orthogon-al designA large pulse duration, ahigh voltage,and a largemachining current and aproper pulse intervalprovides high machiningefficiency9.Kuriakoseet al.Cutting speed pulse-on time,pulse-off time,voltageTitaniumalloysSortinggeneticalgorithmVelocity and the surfacefinish was no single optimalcombination of cuttingparameters, as theirinfluences on the cutting arequite opposite10.Ozdemer et al.Cutting Speed Machining voltage,current, wire speedand peak currentNodular castironRegressionanalysismethodResults indicates thatincrease in SR and CRclearly follows the trendindicated with increasingdischarge energy11.S.Sarkar et al.Cutting speed Pulse-on time,pulse-off time,peak currentγ-TiAl Responsesurfacemethodol-ogyThe residual analysis andexperimental resultindicated that the proposedmodels could adequatelydescribe the performanceindicators12.Hewidyet al.MRR peak current andwater pressureInconel-601 ResponsesurfacemethodologyThe volumetric metalremoval rate generallyincreased with the increaseof the peak current valueand water pressure13.Ramakrishnan et al.Material removalrateTon,wiretension,delay time, wirefeed speed andignition currentintensityD3 tool steel Taguchi’srobustIt was indentified that thepulse on time and ignitioncurrent had influenced morethan the other parameters.14.Kozak etal.Material removalrateClamp positionsilver coatingSi3N4 Design ofexperim-entA reduction in MRR occurswhen the wire moves awayfrom the clamp Asignificant increase in MRRwas observed due to silvercoating15.Kanlayasiri et al.SurfaceroughnessPulse-on time andpulse-peak currentDC53 diesteelFullfactorialIt was reported that the SRincreases when Ton, and
3.
Vijayant Maan , Abhishake Chaudhary / International Journal of Engineering Research andApplications (IJERA) ISSN: 2248-9622 www.ijera.comVol. 3, Issue 3, May-Jun 2013, pp.206-216208 | P a g emethod pulse-peak current areincreased.16.Rao et al. Cutting Speed Discharge current,voltage, wire speedBrass Design ofexperim-entIt was observed that CSdecreases as thickness ofwork piece inreases.17.Romlayet al.Cutting speed Wire speed,wiretensionBrass DOE cutting speed of the wire-EDM has been affected bychanging cuttingparameters18.Rozeneket al.Cutting speed Discharge current,pulse-on time,pulse-off time,voltageMetal matrixcompositeDOEtechniqueThe machining feed rate ofWEDM cutting compositessignificantly depends on thekind of reinforsementmaterial19.Saha etal.Cutting speed pulse-on time,pulse-off time,peak current andcapacitanceTungstencarbidecobaltcompositeNeuralnetwork modelIt was observed that neuralnetwork architectureprovide the best resultprediction20.Bamberge et al.SR Different wirediameterp-typegallium-dopedgermaniumDesign ofexperimentIt was found that 50μmmolybdenum wire achievedthe fastest machining time21.Shunmugam et al.Cutting velocityand surfacefinishPulse-on time,pulse-off time,peak currentTitaniumalloySortinggeneticalgorithmThere was no single optimalcombination of cuttingparameters22.Ramaswamy et al.MRR and SR Ton,Toff ,wiretension, wire feedspeed and ignitioncurrent intensityDie steel Taguchi’srobustdesignapproachIt was indentified that thepulse on time and ignitioncurrent had influenced morethan the other parameters23.Kumar etal.MRR and SR Gap voltage, Ton,Toff and Wire feedIncoloy800super alloyTaguchi’sL9OrthogonalArrayIt is concluded that grey-tagauchi method is mostsuitable for parametricoptimization24.Lee etal. MRR and SR Ton,Toff, peakcurrentCeramicmaterialDesign ofexperimentOptimal combination ofprocess parameters areobtained25.Palanikumarinet al.SurfaceroughnessTon,Toff, peakcurrentGlass fiberreinforcedplasticResponsesurfacemethodologyA model is developed andtested using anovaIII. MATERIAL AND METHODThe w/p material is a high carbon,highchromium D-2 steel,with exceelent resistance towear and abrasion.D-2 steel is choosed due to it’sincresing use in the making of press tools,formingrolls,blanking dies,bushes,punchesetc.The chemical composition of the material isshown in the table given below-
4.
Vijayant Maan , Abhishake Chaudhary / International Journal of Engineering Research andApplications (IJERA) ISSN: 2248-9622 www.ijera.comVol. 3, Issue 3, May-Jun 2013, pp.206-216209 | P a g eThe experiments were performed on the SprintcutWEDM machine from Electronica India PvtLtd.A brass wire of 0.25 mm was used as a cuttingtool. Work pieces are cut into specimens of size20mmX10mmX10mm.Apart from these the following parameters are keptconstant during experimentation-Wire Tension- 4-7-10 unitWire feed - 4-7-10m/minServo Feed- 2050 unitTable 1: Composition of work material.Fig 1: Machine ToolDuring machining,on the basis of literature review,the following process parameters have beenselected for study in the range shown in table 2Table 2 :Process parameters with their RangeIV. RESPONSE SURFACEMETHODOLOGYResponse surface methodology (RSM) isdefined as a collection of mathematical andstatistical methods that are used to develop,improve, or optimize a product or process. Theindependent variables are controlled by theexperimenter, in a designed experiment, while theresponse variable is an observed output of theexperiment. Fig. 4.1 illustrates the estimatedrelationship between a response variable and thetwo independent variables x1 and x2.Fig 2 : An example of a response surface.The field of response surface methodology consistsof the experimental strategy for exploring the spaceof the process or independent variables, empiricalstatistical modelling to develop an appropriateapproximating relationship between the yield andthe process variables, and optimization methods forfinding the values of the process variables thatproduce desirable values of the response.In general,the relationship between the response y andindependent variables, ξ1, ξ2,…,ξky=f(ξ1,ξ2,…,ξk.)+ε (1)Usually ε is treated as a statistical error, oftenassuming it to have a normal distribution withmean zero and variance σ2. ThenE(y) = η = E [f (ξ1, ξ2,…,ξk.)] + E (ε) = f (ξ1,ξ2,…,ξk.) (2)The variables ξ1, ξ2,…,ξk in Equation (2) areusually called the natural variables.In much RSMwork it is convenient to transform the naturalvariables to coded variables X1, X2…Xk, which areusually defined to be dimensionless with mean zeroand the same standard deviation. In terms of thecoded variables, the response function (3) will bewritten asη=f(X1,X2…Xk) (3)In many cases, either a first-order or a secondorder model is used.For the case of twoindependent variables, the first-order model interms of the coded variables isS.NO MATERIAL %1. CARBON 1.50%2. SILICON 0.30%3. MAGANESE 0.30%4. SULPHUR 0.027%5. PHOSPHORUS 0.026%6. VANADIUM 0.90%7. CHROMIUM 11.50%8. MOLYBDENUM 0.78%9. COPPER 0.009%10. IRON RestS.No Input Parameters Range1. Pulse on time (Ton) 112-127 machineunits2. Pulse off time (Toff) 42-55 machine units3. Peak current (IP) 170-215 Amp4. Servo voltage (SV) 40-60Volts
5.
Vijayant Maan , Abhishake Chaudhary / International Journal of Engineering Research andApplications (IJERA) ISSN: 2248-9622 www.ijera.comVol. 3, Issue 3, May-Jun 2013, pp.206-216210 | P a g eη=β0+β1X1+β2X2 (4)The form of the first-order model in Equation (4) issometimes called main effects model, because itincludes only the main effects of the two variablesX1 and X2. If there is an interaction between thesevariables, it can be added to the model easily asfollows:η=β0+β1X1+β2X2+β12X1X2 (5)This is the first-order model with interaction.Oftenthe curvature in the true response surface is strongenough that the first-order model (even with theinteraction term included) is inadequate. A second-order model will likely be required in thesesituations. For the case of two variables, thesecond-order model isη = β0 + β1X1 + β2X2 + β11 X12+ β22 X22+ β12 X1X2(6)This model would likely be useful as anapproximation to the true response surface in arelatively small region. The second-order model iswidely used in response surface methodology.5. DESIGN OF EXPERIMENTThe experiments were designed by usingDesign expert software.Response surfacemethodology was used and central compositedesign was applied.Total 30 runs were obtained byapplying given procees parameters. The value ofinput parameters were set according to this designfor each run.The value of MRR was calculated foreach run.The design is shown in table 5.1 withresponse MRR and SR.Std RunFactor 1A: TonMachineunitsFactor 2B: ToffMachineunitsFactor 3C: SVVoltsFactor 4D: IPAmpResponseMRRMm2/minResponseSRµm14 1 123 44 55 215 22.1 2.612 2 124 44 45 185 23.6 2.929 3 116 44 45 200 12.5 1.85 4 116 44 55 215 12.4 1.713 5 116 51 45 215 12.1 1.6217 6 120 47 50 185 18.3 2.098 7 123 51 55 200 21.7 2.5015 8 123 51 55 185 21.5 2.5112 9 120 51 45 215 18.2 2.0218 10 123 47 50 185 21.7 2.586 11 123 43 55 200 22.1 2.6216 12 120 51 55 185 17.9 2.0120 13 116 47 50 200 12.2 1.6611 14 123 51 45 215 21.9 2.5210 15 123 43 45 185 22.3 2.644 16 116 51 45 185 12.1 1.617 17 116 51 55 215 11.9 1.611 18 116 43 45 215 12.8 1.8513 19 120 43 55 215 18.8 2.3519 20 120 47 50 185 18.8 2.1023 21 119 40 50 200 18.1 2.1829 22 120 47 50 200 18.2 2.0930 23 120 47 50 200 18.6 2.0926 24 119 47 60 200 17.1 1.9824 25 120 55 50 230 17.8 1.9828 26 120 47 50 200 18.3 2.1027 27 119 47 50 170 16.9 2.1125 28 127 47 40 200 25 3.0822 29 120 47 50 200 18.2 2.1021 30 112 47 50 200 6.8 1.40Table 3 : Experimental design with response data(MRR and SR)
6.
Vijayant Maan , Abhishake Chaudhary / International Journal of Engineering Research andApplications (IJERA) ISSN: 2248-9622 www.ijera.comVol. 3, Issue 3, May-Jun 2013, pp.206-216211 | P a g e6. RESULT AND DISCUSSIONIn this study models as well asexperimental results of the responses have beenanalyzed.Model analysis was made by usingDesign-Expert version 8.0.3 while the analysis ofMRR is done in line with the behaviour ofmachining parameters on the responses.6.1 Material Removal Rate(MRR) and SurfaceRoughness(SR) model-The regression equation for MRR and SR as afunction of four input process variables- Pulse ontime (Ton), Pulse off time (Toff), Servo voltage(SV), peak current (IP) was developed usingexperimental data and is given below.Final Equation in Terms of Coded Factors forMRR-MRR=+17.70+5.00*A-0.33*B+0.055*C-0.41*A2-0.14* C2Final Equation in Terms of Actual Factors forMRR-MRR=-571.92416 +8.36169* ton -0.087745 *toff+0.5708* sv-0.029410* ton2-5.59881E-003*sv2Final Equation in Terms of Coded Factors for SR-SR=+2.03+0.46*A-0.089*B-0.011*D+0.056*A2+0.018* B2 +0.021* D2Final Equation in Terms of Actual Factors for SR-SR=+52.76687-0.83824* ton -0.14747 * toff*-0.037761* peak current+4.01619E-003*ton2+1.30198E-003* toff2+9.26250E-005*peak current26.2 Effect of process parameters on MRR andSR-The figure 6.1-6.3 shows the effect ofinput parameters Pulse on time (Ton), Pulse offtime (Toff), Servo voltage (SV) on the MRR andfigure 6.4-6.6 shows the effect of input parametersPulse on time (Ton), Pulse off time (Toff),Peakcurrent on the SR.The MRR and SR increased withPulse on time,and decreased with Pulse off timeand servo voltage.The peak current has a verysmall influence on the MRR and servo voltage hasvery less influence on SR,But the Ton is the mostinfluential parameter among them.In this processspark energy affects the MRR.And it is a functionof Pulse on time and peak current.In The viewpoint of industrial economy it is desirable to obtainhigher value of MRR and low value of SR.Design-Expert® Sof twareFactor Coding: ActualmrrCI BandsX1 = A: tonActual FactorsB: tof f = 47.50C: sv = 50.00D: peak current = 200.00115.75 117.25 118.75 120.25 121.75 123.25A: tonmrr12141618202224One FactorFig 2:Showing the effect of Ton on M.R.RDesign-Expert® Sof twareFactor Coding: ActualmrrCI BandsX1 = B: tof fActual FactorsA: ton = 119.50C: sv = 50.00D: peak current = 200.0043.75 45.25 46.75 48.25 49.75 51.25B: toffmrr1717.217.417.617.81818.218.4One FactorFig3 :Showing the effect of Toff on M.R.RDesign-Expert® Sof twareFactor Coding: ActualmrrCI BandsX1 = C: svActual FactorsA: ton = 119.50B: tof f = 47.50D: peak current = 200.0045.00 47.00 49.00 51.00 53.00 55.00C: svmrr17.217.317.417.517.617.717.817.9One FactorFig4:Showing the effect of S.V onM.R.RDesign-Expert® Sof twareFactor Coding: ActualsrCI BandsX1 = A: tonActual FactorsB: tof f = 47.50C: sv = 50.00D: peak current = 200.00115.75 117.25 118.75 120.25 121.75 123.25A: tonsr1.41.61.822.22.42.6One FactorFig 5 : Showing the effect of Ton S.RDesign-Expert® Sof twareFactor Coding: ActualsrCI BandsX1 = B: tof fActual FactorsA: ton = 119.50C: sv = 50.00D: peak current = 200.0043.75 45.25 46.75 48.25 49.75 51.25B: toffsr1.91.9522.052.12.152.2One FactorFig 6 : Showing the effect of Toff on S.RDesign-Expert® Sof twareFactor Coding: ActualsrCI BandsX1 = D: peak currentActual FactorsA: ton = 119.50B: tof f = 47.50C: sv = 50.00185.00 191.00 197.00 203.00 209.00 215.00D: peak currentsr1.9822.022.042.062.082.12.12One FactorFig 7 : Showing the effect of Peak Currenton S.R
7.
Vijayant Maan , Abhishake Chaudhary / International Journal of Engineering Research andApplications (IJERA) ISSN: 2248-9622 www.ijera.comVol. 3, Issue 3, May-Jun 2013, pp.206-216212 | P a g eThe effects of process parameters are taken two at atime on MRR is shown in the Fig 6.7-6.8.Fig 6.7shows the combined effect of Ton and Toff onMRR.Fig 6.8 shows the combined effect of peakcurrent and SV on MRR.Design-Expert® Sof twareFactor Coding: ActualmrrDesign points abov e predicted v alueDesign points below predicted v alue256.8X1 = A: tonX2 = B: tof fActual FactorsC: sv = 50.00D: peak current = 200.0043.7545.2546.7548.2549.7551.25115.75117.25118.75120.25121.75123.251012141618202224mrrA: tonB: toffFig 8 : Showing the combined effect of Ton andToff on MRRDesign-Expert® Sof twareFactor Coding: Actualmrr256.8X1 = C: svX2 = D: peak currentActual FactorsA: ton = 119.50B: tof f = 47.50185.00191.00197.00203.00209.00215.0045.0047.0049.0051.0053.0055.0017.517.5517.617.6517.717.75mrrC: svD: peak currentFig 9 : Showing the combined effect of SV andPeak Current on MRRDesign-Expert® Sof twareFactor Coding: ActualsrDesign points abov e predicted v alueDesign points below predicted v alue3.081.4X1 = A: tonX2 = B: tof fActual FactorsC: sv = 50.00D: peak current = 200.0043.7545.2546.7548.2549.7551.25115.75117.25118.75120.25121.75123.251.41.61.822.22.42.62.8srA: tonB: toffFig 10: Showing the combined effect of Ton andToff on SRDesign-Expert® Sof twareFactor Coding: Actualsr3.081.4X1 = D: peak currentX2 = A: tonActual FactorsB: tof f = 47.50C: sv = 50.00115.75117.25118.75120.25121.75123.25185.00191.00197.00203.00209.00215.001.61.822.22.42.6srD: peak currentA: tonFig 11: Showing the combined effect of Ton andPeak Current on SR6.3 Residual Analysis-The residual analysis as a primarydiagnostic tool is also done. Normal probabilityplot of residuals has been drawn (Fig 611,6.12).All the data points are following the straight line.Thus the data is normally distributed. It can beseen from Figure 6.13 and 6.14 that all the actualvalues are following the predicted values and thusdeclaring model assumptions are correct.Design-Expert® Sof twaremrrColor points by v alue ofmrr:256.8Internally Studentized ResidualsNormal%ProbabilityNormal Plot of Residuals-3.00 -2.00 -1.00 0.00 1.00 2.00 3.00 4.0015102030507080909599Fig 12 :Normal plot of residulals for MRRDesign-Expert® Sof twaresrColor points by v alue ofsr:3.081.4Internally Studentized ResidualsNormal%ProbabilityNormal Plot of Residuals-3.00 -2.00 -1.00 0.00 1.00 2.00 3.0015102030507080909599Fig 13: Normal plot of residulals for SRDesign-Expert® Sof twaremrrColor points by v alue ofmrr:256.82ActualPredictedPredicted vs. Actual5.0010.0015.0020.0025.0030.005.00 10.00 15.00 20.00 25.00Fig 14 : Predicted v/s Actual for MRRDesign-Expert® Sof twaresrColor points by v alue ofsr:3.081.422ActualPredictedPredicted vs. Actual1.001.502.002.503.003.501.00 1.50 2.00 2.50 3.00 3.50Fig 15 : Predicted v/s Actual for SR
8.
Vijayant Maan , Abhishake Chaudhary / International Journal of Engineering Research andApplications (IJERA) ISSN: 2248-9622 www.ijera.comVol. 3, Issue 3, May-Jun 2013, pp.206-216213 | P a g e6.4 Analysis of variance-In order to statistically analyze the results,ANOVA was performed. Process variables havingp-value<0.05 are considered significantterms for the requisite response characteristics.The insignificant parameters were pooled usingbackward elimination.Table 4 : Anova table for MRR-Source Sum OfSquaresDf MeanSquareF-Valuep-ValueProb>fBlock 9.13 2 4.56Model 531.58 5 106.32 873.12 <0.0001 SignificantA-ton 467.42 1 467.42 3838.71 <0.0001B-toff 2.60 1 2.60 21.31 0.0001C-sv 0.060 1 0.060 0.49 0.4895A2 3.48 1 3.48 28.62 <0.0001C2 0.42 1 0.42 3.47 0.0759Residual 2.68 22 0.12Lack of Fit 2.57 20 0.13 2.41 0.3339 Not-significantPure Error 0.11 2 0.053Cor Total 543.49 29Table 5 : Anova table for SR-Source Sum OfSquaresDf MeanSquareF-Valuep-ValueProb>fBlock 0.12 2 0.059Model 4.75 6 0.79 253.30 <0.0001 SignificantA-ton 4.16 1 4.16 1332.02 <0.0001B-toff 0.19 1 0.19 60.86 0.0001D-Peakcurrent2.532E-003 1 2.532E-003 0.81 0.4895A2 0.082 1 0.082 26.25 <0.0001B2 9.192E-003 1 9.192E-003 2.94 0.0759D2 0.012 1 0.012 3.77Residual 0.066 21 3.125E-003Lack of Fit 0.066 19 3.450E-003 2.40 0.1692 Not-significantPure Error 6.667E-005 2 3.333E-005Cor Total 4.93 296.5 Multi Response Optimization Using DesirabilityFunction-The goal of optimization is to find a good setof conditions that will meet all the goals. It is notnecessary that the desirability value is 1.0 as thevalue is completely dependent on how closely thelower and upper limits are set relative to the actualoptimum.The constraints for the optimization ofindividual response chracteristicsare given in table 6.1.Goals and limits areestablished for the response in order to accuratelydetermine their impact on it’s desirability.A set of30 optimal solutions is derived with specific designconstraints for MRR and SR.The set of conditionpossessing highest desirability is selectrd asoptimium condition and is given in table6.2.Desirability Plots are drawn keeping Inputparameters in range,MRR maximum and SRminimum.The plots showing the effect of inputparameters are shown in fig. 6.14.
9.
Vijayant Maan , Abhishake Chaudhary / International Journal of Engineering Research andApplications (IJERA) ISSN: 2248-9622 www.ijera.comVol. 3, Issue 3, May-Jun 2013, pp.206-216214 | P a g eTable 6 :Range of Input Parameters,MRR andSR for Desirability-ProcessParameterGoal LowerLimitUpperLimitLowerWeightUpperWeightImportanceTon is in range 116 123 1 1 3Toff is in range 44 51 1 1 3Servo Voltage is in range 45 55 1 1 3Peak Current is in range 185 215 1 1 3MRR maximize 6.8 25 1 1 5SR minimize 1.4 3.08 1 1 3Table 7 : Optimal Set of Condition with highestdesirability-S.No.TonToffServo-VoltagePeak-CurrentMRRSRDesirability1 1235145 215 21 2.320.662 selectedDesign-Expert® SoftwareFactor Coding: ActualDesirability1.0000.000X1 = A: tonX2 = B: toffActual FactorsC: sv= 45.00D: peakcurrent = 215.0043.7545.2546.7548.2549.7551.25115.75117.25118.75120.25121.75123.250.4500.5000.5500.6000.650DesirabilityA: tonB: toffWarning! Surface truncated by selected response (Y) range0.6620.662vFig 16 : 3D Surface Graph of Desirability forMRR and SR(Ton, Toff)6.6 Ramp function and bar function graphs-The ramp function graphs and bar graphsdrawn using Design expert shows the desirabilityfor each factor and each response. The dot on eachramp reflects the factor setting or responseprediction for that response characteristic.Bar graphshows the individual Desirability of each objectivefunction.A:ton = 1.160.90 1.25B:toff = 28.0016.00 28.00C:sv = 50.9945.00 55.00D:peak current = 203.82185.00 215.00Material Removal Rate = 19.78286.8 25Surface Roughness = 2.206571.4 3.08Desirability = 0.634Fig 17: Ramp Graph showing desirability forMRR and SR11110.7136940.5194670.633548Desirability0.000 0.250 0.500 0.750 1.000A:tonB:toffC:svD:peak currentmrrsrCombinedFig 18 : Bar Graph showing desirability forMRRand SR7. CONCLUSIONIn this research an experimental investigation wasperformed to consider the machining chracteristicsof D-2 steel and following results are obtained-1.Results shows that the Central composite designis a powerful tool for providingexperimental diagrams and statstical-mathematicalmodels,to perform the experimentsappropriately and economically.2.The ANOVA shows that Ton, Toff,SV hasmaximum influence on MRR,and Ton, Toff,3.The peak current has a very less influence on theMRR and servo voltage has very lessInfluence on SR.4.The methodology adopted establishes theoptimization of D-2 steel machining inWEDM.And facilitates the effective use of D-2steel in industrial applications by reducing the costof machining.References-1. Liao Y.S., Huang J.T., Su H.C. (1997), “Astudy on the machining-parametersoptimization of wire electrical dischargemachining”, Journal of MaterialsProcessing Technology, Vol. 71, pp: 487–493.2. Jangra Kamal, Jain Ajai, Grover Sandeep(2010), “ Optimization of multiple-machining characteristics in wire electricaldischarge machining of punching dieusing grey relational analysis”, Journal ofScientific and Industrial Research,Vol.69,pp. 606-612.
10.
Vijayant Maan , Abhishake Chaudhary / International Journal of Engineering Research andApplications (IJERA) ISSN: 2248-9622 www.ijera.comVol. 3, Issue 3, May-Jun 2013, pp.206-216215 | P a g e3. Mohapatra S.S., Patnaik Amar (2006),“Parametric optimization of wire electricaldischarge machining (WEDM) processusing taguchi method”, Journal of theBrazilian Society of Mechanical Scienceand Engineering, Vol. 28, pp: 422-429.4. Kanlayasiri K., Boonmung S. (2007),“Effects of wire-EDM machiningvariables on surface roughness of newlydeveloped DC 53 die steel: design ofexperiments and regression model”,Journal of Materials ProcessingTechnology, Vol. 192-193, pp: 459–464.5. Tosun, Nihat, Cogun, Can and Gul Tosun(2004), “A study on kerf and materialremoval rate in wire electrical dischargemachining based on Taguchi method”,Journal of Material ProcessingTechnology, 152, pp.316-322..6. Sanchez J.A., Rodil J.L., Herrero A.,Lopez de Lacalle L.N., Lamikiz A.(2007), “On the influence of cutting speedlimitation on the accuracy of wire-EDMcorner-cutting”, Journal of MaterialsProcessing Technology, Vol. 182, pp:574–5797. Haddad M.J., Tehrani Fadaei Alireza(2008), “Material removal rate (MRR)study in the cylindrical wire electricaldischarge turning (CWEDT) process”,Journal of Materials ProcessingTechnology, 199,pp. 369–378.8. Guo, Z. N., Wang, X., Huang, Z. G., Yue,T. M. (2002), “Experimental investigationinto shaping particle-reinforced material byWEDM”, Journal of Materials processingTechnology, Vol.129, pp.56-59.9. Kuriakose, Shajan and Shunmugam, M. S.(2005), “Multi-objective optimization ofwire-electro-discharge machining processby non-dominated sorting GeneticAlgorithm”, Journal of MaterialsProcessing Technology, Vol.170, pp: 133-14110. Ozdemir, N. and Ozek, Cebeli. (2006),“An investigation on machinability ofnodular cast iron by WEDM”,International Journal of Advancemanufacturing Technology,Vol. 28,pp.869-872.11. Sarkar, S., Sekh, M. Mitra, S.,Bhattacharyya, B. (2008), “Modeling andoptimization of wire electrical dischargemachining of γ-TiAl in trim cuttingoperation”, Journal of MaterialProcessing Technology205, pp: 376-387.12. Hewidy, M.S., El-Taweel, T.A., El-Safty,M.F. (2005), “Modeling the machiningparameters of wire electrical dischargemachining of Inconel 601 using RSM”,Journal of Materials ProcessingTechnology, Vol. 169, pp: 328-336.13. Ramakrishnan, R. and Karunamoorthy, L.(2006), “Multi response optimization ofwire EDM operations using robust designof experiments”, International Journal ofAdvance Manufacturing Technology, Vol.29, pp: 105-112.14. Kozak, Jerzy.,Rajurkar,K.P., Chandrana,Niraj (2004), “Machining of low electricalconductive materials by wire electricaldischarge machining”, Journal ofMaterials Processing Technology,Vol.149, pp.266-27115. Kanlayasiri K., Boonmung S. (2007),“Effects of wire-EDM machiningvariables on surface roughness of newlydeveloped DC 53 die steel: design ofexperiments and regression model”,Journal of Materials ProcessingTechnology, Vol. 192-193, pp: 459–464.16. Rao, Parameswra, C.V.S, Sarcar, M.M.M.,(2009), “Evaluation of optimal parametersfor machining brass with wire cut EDM”,Journal of Scientific and IndustrialResearch, Vol 68, pp: 32-35.17. Romlay F.R, Mokhtar.A (2008), “MachineParameter Optimization for wire-ElectricDischarge machining”, InternationalConference on Computational &Experimental Engineering and Sciences,Vol. 8, No. 4, pp. 139-144.18. Rozenek. M., Kozak,J., Dabrowski, L.,Lubkowski, K. (2001), “Electricaldischarge machining characteristics ofmetal matrix composites”, Journal ofMaterials ProcessingTechnology,Vol.109, pp: 367-37019. Saha,P., Singha,A., Pal,S.K., Saha,P.(2009), “Soft computing models based onprediction of cutting speed and surfaceroughness in wire electro-dischargemachining of tungsten carbide cobaltcomposite”, International Journal ofAdvance Manufacturing technology.20. Bamberg Eberhard, Rakwal Dinesh(2008), “Experimental investigation ofwire electrical discharge machining ofgallium-doped germanium”, Journal ofMaterials Processing Technology, Vol.197, pp: 419–42721. Scott, D., Boyina, S., Rajurkar, K. P.(1991), “Analysis and optimization ofparameter combinations in wire electricaldischarge machining”, InternationalJournal of Production Resources, Vol.29,No.8, pp.2189-22022. Ramasawmy, H. and Blunt, L. (2004),“Effect of EDM process parameters on 3D
11.
Vijayant Maan , Abhishake Chaudhary / International Journal of Engineering Research andApplications (IJERA) ISSN: 2248-9622 www.ijera.comVol. 3, Issue 3, May-Jun 2013, pp.206-216216 | P a g esurface topography”, Journal of MaterialsProcessing Technology, 148, pp: 155–164.23. Kumar Muthu , Babu Suresh,Venkatasamy R, Raajenthiren M (2010) ,“Optimization of the WEDM Parameterson Machining Incoloy800 Super alloywith Multiple Quality Characteristics”,International Journal of EngineeringScience and Technology, Vol. 2(6),pp.1538-1547,.24. Lee, S.H. and Li, X.P. (2001), “Study ofthe Effect of Machining Parameters on theMachining Characteristics in ElectricalDischarge Machining of TungstenCarbide”, Journal of Materials ProcessingTechnology, 115:pp. 344-35825. K.Palanikumar,”Modelling and analysisfor surface rouhness in machining glassfibre reinforced plastics using responsesurface methodology,”Materials AndDesign28 2611-2618,2007
Be the first to comment