1239
МЕТАЛЛИЧЕСКИЕ ПОВЕРХНОСТИ И ПЛЁНКИ
PACSnumbers:06.20.Dk,06.60.Mr,06.60.Vz,62.20.mm,62.20.Qp,62.23.Pq,68.35.Ct
Experimental
Investigation
of
B
4C
Particulate
Reinforced
Aluminium
6061
Based
Composite
Material
in
Wire-Cut
EDM
Ş.Karabulut,U.Gökmen*,H.Karakoç,Ö.K.Kalkan,andR.Çıtak**
HacettepeUniversity,
DepartmentofMechanicalProgram, 06935Ankara,Turkey
*GaziUniversity,
TechnicalSciencesVocationalSchool, 06935Ostim,Ankara,Turkey
**
GaziUniversity, FacultyofTechnology,
DepartmentofMetallurgyandMaterialsEngineering, 06500Ankara,Turkey
Inthepresentpaper,theinfluencesofcuttingparametersonsurface
rough-ness in wire electric-discharge machining of (WEDM) process of
particle-reinforcedaluminiumAA6061alloycompositeareinvestigated.The
compo-sitesareproducedusing15%wt.B4Cfractionusingpowdermetallurgy.
Ex-perimental trialsare performedbasedonTaguchiL18(2132)withamixed
orthogonalarray,andthe WEDMcuttingparametersareoptimizedforthe
bestsurfacequality.Theinvestigationresultsareevaluatedbyresponse
sur-faceplotsandmaineffectgraphs.Themachinedsurfaceofthemetalmatrix
composite isinvestigatedusingscanningelectronmicroscopy(SEM)
micro-graphs.TheeffectofWEDMmachiningvariablesaredeterminedusing
anal-ysis of variance(ANOVA).Theanalysisresult showsthat themost
signifi-cantcuttingparameterispeakcurrentforsurfaceroughness.TheSEMand
opticalmicrographsindicatethatthereinforcedB4Cparticlesare
homogene-ouslydistributedinthematrixstructure.Mathematicalmodelsarealso
gen-erated using regression analysis for the surface roughness. Confirmation
tests arecarriedout todeterminethepredictionperformanceofthe
mathe-matical models,and thesurface roughness ispredicted with an acceptable
meansquarederror.
Вційроботідослідженовпливпараметріврізаннянашерсткістьповерхні приобробленнінаелектроерозійномувирізномустанку(ЕЕВС)композита алюмінійовогостопуAA6061,армованогочастинками.Композити вироб-лялися з використанням 15% вагової фракції B4C методою порошкової Фотокопирование разрешено только в соответствии с лицензией Напечатано в Украине.
металургії. Експерименти виконувалися на базі Taguchi L18 (2132) зі змішаним ортогональним масивом; параметри ЕЕВС-оброблення різан-нямоптимізувалися,щободержатинайкращуякістьповерхні. Результа-ти досліджень оцінювалисязаграфіками поверхнівідгуку та головного ефекту.Обробленаповерхнякомпозитноїметалевоїматриці досліджува-лася з використанням мікрознімків сканівної електронної мікроскопії (СЕМ). Вплив ЕЕВС-оброблення визначався за допомогою дисперсійної аналізи(ДА).Аналізарезультатівпоказала,щонайбільшістотним пара-метромрізаннядляшерсткостиповерхнієпіковийструм.СЕМтаоптичні мікрознімкипоказали,щоармувальнічастинкиB4Cрозподіленів струк-турі матриці рівномірно. З використанням реґресійної аналізи були та-кожзґенерованіматематичнімоделідляповерхневоїшерсткости. Випро-буваннянавідповідністьтехнічнимумовамбуливиконанізметою попе-редньої оцінки математичних моделів, і поверхневу шерсткість було спрогнозованозприпустимоюсередньоквадратичноюпохибкою. Вданной работеисследовановлияниепараметров резания на шерохова-тостьповерхностиприобработкенаэлектроэрозионномвырезномстанке (ЭЭВС) композита алюминиевого сплаваAA6061, армированного части-цами.Композитыпроизводилисьсиспользованием15%весовойфракции B4Cметодомпорошковойметаллургии.Экспериментыпроизводилисьна базе Taguchi L18 (2132) со смешанным ортогональным массивом; пара-метры ЭЭВС-обработки резанием оптимизировались с целью получения наилучшего качества поверхности. Результаты исследований оценива-лись по графикамповерхностиотклика и главногоэффекта. Обработан-ная поверхность композитной металлической матрицы исследовалась с использованиеммикроснимковсканирующейэлектронноймикроскопии (СЭМ). ВлияниеЭЭВС-обработкиопределялосьпри помощи дисперсион-ногоанализа(ДА).Анализрезультатовпоказал,чтонаиболее существен-нымпараметромрезаниядляшероховатостиповерхностиявляется пико-выйток.СЭМиоптическиемикроснимкипоказали,чтоармирующие ча-стицыB4Cраспределенывструктурематрицыравномерно.С использова-ниемрегрессионногоанализабылитакжесгенерированыматематические модели для поверхностной шероховатости. Испытания на соответствие техническимусловиямбылипроведенысцельюпредварительнойоценки математическихмоделей,и поверхностнаяшероховатостьбыла спрогно-зированасприемлемойсреднеквадратичнойпогрешностью.
Key words: wire electric dischargemachining, surface roughness,Taguchi
method,responsesurfacemethodology.
(ReceivedJuly6,2015)
1.INTRODUCTION
Metalmatrixcomposites(MMCs)havebeenwidelyinvestigatingin
re-cent years andare now utilizedin manyengineering fields including
sporting goodsbecauseoftheirlowdensityincombinationwiththeir
excellent wear resistance, highspecific strength,hardness, and
frac-turetoughness [1—6].However,machinabilityofMMCsisconsidered
difficult in connection with hard reinforcement elements in matrix
structure[1—7].Boroncarbide (B4C)is extremelyhard reinforcement
material with the superior properties such as good wear resistance,
highhardness,lowspecificweight,corrosionresistance,highmelting
point, adequate resistance to chemical agents, and good mechanical
properties.Theseoutstanding performancesofB4Cmadeita
prefera-ble reinforced material, widely used in numerous industrial
applica-tionsrequiringhighresistance,suchasthenuclearindustry,fortank
armour, and ballistic protections. Hence, several researchers have
studied the production and machinability properties of MMCs
rein-forced with B4C in recent years [8, 9]. However, there are two main
problems hindering the superior properties of B4C, one is that very
hightemperatureisrequiredforitssinteringandtheotheristhelow
fracturetoughness[10].Wireelectricaldischargemachining(WEDM)
is ahighprecision machiningmethod widelyusedforhardmaterials,
metallic alloys, andgraphite that would bevery difficult tocutwith
traditional machine tools using the best economic cutting tools. In
wire-cutting technique, a thin single-strand metal wire is machined
the workpiecesubmerged inatank of deionized watertoutilize heat
fromelectricalsparks.WEDMusesanonstopcuttingwireelectrodeto
machine the desired shape alongside the cutting path using 0.05—
0.30mmindiameterthin copper,brassor tungstenwireandcan
ma-chineverysmallcornerradiuswithhighprecision[11].Motorcuetal.
studied theinfluenceofcuttingparametersonthesurface roughness
andmaterialremovalrate(MRR)incuttingofAl/B4C/Grhybrid
com-posites usingWEDMdependenceonthewirespeed,pulse-ontimeand
pulse-off time.Theyobservedthat themost significantparameter on
surfaceroughnessandMRRwasthewirespeedwith85.94%
contribu-tionrate[12].Yanetal.investigatedtheeffectsofmachiningprocess
on surface roughness (Ra), cutting width, and material removal rate
andwirebreakagebehaviourintheWEDMofAl6061compositeswith
different reinforcement ratios of Al2O3.The test results showed that
theAl2O3 reinforcementvolumefractioninfluencesontheRa,kerfand
MMR.Theyalsoreportedthatahighwirespeed,verylowwiretension,
and highflushing ratemust bechosentopreventwire breakage[13].
Shandilyaet al.studiedtheeffectoftheinputparametersonaverage
cutting speed duringWEDM ofAl6061/SiC metal matrix composite.
Servovoltageisthemoresignificantinputparameterforaverage
cut-ting speed than pulse-off time and wire feed rate [14]. The surface
roughness and material removal rate were increasedwith increase in
pulse-ontimeanddecreasedwithincreaseinpulse-offtime.MRRwas
time (Toff), pulse-on time (Ton) and peak current (IP), pulse-off time
(Toff)andpeakcurrent(IP).Pulse-ontime(Ton)and peakcurrent (IP)
affected themachinedsurfaceroughness[15].SatishKumar etal.
in-vestigatedtheeffectsofdifferentmachiningparametersonMRRand
Ra intheWEDMofAl6063/SiCMMCatdifferentreinforcement
rati-os. The researchers reported that surface quality and MRR were
de-creasedwiththeincreasingpercentagevolumefractionofSiCparticles
[16]. Surface roughness and gap width were mainly affected by the
pulse-ontimeintheWEDMofAl6061 reinforcedwithAl2O3 particle
MMC[17].Pulse-ontimeandcurrentwerethemosteffective
parame-tersformachiningspeed andsurfacequalityintheWEDMofAl—SiC
metalmatrixcomposite[18].
2.EXPERIMENTALMATERIALSANDMETHOD
Theexperimentalworkpieceswereproducedfromhigh-purity
alumin-ium 6061mixedwith15%commercial-gradeB4Cpowdersusing
pow-der metallurgy method. The median size of Aluminium 6061 powder
used in metal matrix composite (MMC) was100m and B4C powders
had average size of 10m. Aluminium alloy and B4C powders were
mixed toachieve homogeneity for 45minutes ina three-dimensional
Turbula mixer. The mixed powders were compacted by cold pressing
under 300MPa.Thespecimens were sintered inavacuum furnace at
550Cfor60minutesandextrudedusingapre-heatedextrusionmould
oftemperature500Cfor1hour.Thethicknessofproducedcomposite
sheetswas12.7mm.WorkpiecematerialswereanalysedusingaJEOL
JSM6060LWscanningelectronmicroscope(SEM)andenergy
disper-sive spectroscopy(EDS).Theoptical and SEMmicrographofthe
sur-face texture of the machined composite and B4C reinforcement
ele-mentscanbeseeninFig.1.TheopticalandSEMmicrographs
ed that the B4C particles distribution is fairly uniform in composite
specimenandachievedagoodinterfacialbondingbetweenmatrixand
B4Cparticles.Thechemicalcompositionsandmechanicalpropertiesof
Al6061alloyandreinforcedwith15%wt.B4Cmetalmatrixcomposite
arepresentedinTable1andTable2,respectively.Thehardness
meas-urements ofspecimens wereperformedbyVickers HV3 hardness
ma-chineEMCOTESTDuravision200applyingaloadof3kgforaperiod
of 5s. The average hardness value for each sample was obtained by
measuring five different areas. Impact energy of composite samples
was tested using sharply impact-testing machine Instron Wolpert
PW30 withmaximumhammerenergy of150J.Impacttestswere
ap-pliedtoV-notchedspecimensforfracturetoughnessdeterminationof
composite samples according to EN ISO148.01. Tensile and flexural
testswereperformedusingInstron3363universaltestingmachineat
a constant strain rate of 1mm/s. Everyimpact, tensile and flexural
testswereemployedatleastthreetimes,andtheaveragevalueforeach
setofthecompositessampleswascalculated.
The experiments wereperformed ontheMitsubishiMV1200 series
CNC WEDM.Rectangularparts ofsize 31.76.3512.7mm3werecut
fromtheworkpiecematerialasshowninFig.2.Abrasswireelectrode
of diameter 0.30mm wasused as the cuttingtool forconducting the
experiments anddeionizedwaterwasusedasthedielectricfluid.The
machined surface of the workpiece was measured using Mitutoyo
SurftestSJ210device.Surfacequalitywasmeasuredatfourdifferent
machinedsurfacesandtheaveragesurfaceroughnessvaluewas
calcu-lated. Machining parameters and their levels used in the WEDM of
MMCsarelistedinTable3.
3.EXPERIMENTALRESULTSANDDISCUSSION
Thepurposeofthisstudyistoinvestigatetheeffectofwire-EDM
ma-TABLE1.ChemicalcompositionofAl6061alloyelements.
Element Fe Si Cr Mn Mg Zn Cu Ti Al
Al6061 0.5 0.6—1.0 0.1 0.2—0.8 0.8—1.2 0.25 0.6—1.1 0.1 Balance
TABLE2.MechanicalpropertiesofAl6061/B4C.
Workpiece material Hardness, HV Impactenergy, J Maximumtensile stress,MPa Maximumflexure stress,MPa Al6061 68.2 26.3 201 467 15%wt.B4C 74 6.1 194 456
chiningparametersonthesurfaceroughnessduringcuttingofB4C
re-inforced metal matrix composite. The effects of spark gap voltage,
peakcurrentandwiretensiononsurfaceroughnessusingabrass
elec-trodewereinvestigated.Theexperimentswerecarriedoutbasedonthe
TaguchiL18(2132)withamixedorthogonalarrayandtheanalysisof
variance (ANOVA) has been employed using statistical software
Minitab16todeterminethesignificantcontributionofmachining
pa-rameters. Theexperimental time and costcan bedecreased using
or-thogonalarraysbyreducingthenumberoftestsandminimizesthe
ef-fectsofparametersthatcannotbecontrolled.
Furthermore, it ensures a simple, powerful, and systematic
ap-TABLE3.Machiningparametersandtheirlevels.
Factor Processparameters Level1 Level2 Level3
A Wiretension(WT) 10g 13g
B Sparkgapvoltage(SV) 30V 60V 80V
C Peakcurrent(IP) 8A 10A 13A
a b
proachtospecifyingtheoptimalmachiningfactorsduringthe
experi-ments. A number of external factors not considered in the
experi-mentaldesigncanaffecttheexperimentalresults.Theseexternal
fac-torsand theireffectontheresultsintermsofqualitycharacteristics
are named ‘thenoise’. Thesignal-to-noise ratio(S/Nratio) computes
theaccuracyofthequalitycharacteristic.TheS/Nratioiscalculated
in two processes. First, mean squared deviation (MSD) between the
experimentalresultsandoptimalvaluesarecalculatedbyequation(1).
Second,computedMSDresultsareconvertedusingequation(2)[19].
Then,thecuttingparametersareanalysedbasedontheS/N.Thereare
three differentsignal-to-noiseratiosandindividualdesirability
func-tions:largerisbetter,nominalisbest,andsmallerisbetter.S/Nratio
indications can be selecteddependingon theaim ofthe experiments.
The objective of this investigation is tominimize the surface
rough-nessvalue.Therefore,the-smaller-the-betterhasbeenchosento
calcu-latetheS/Nratiosusingthefollowingformulae:
, / ) (y12 y22 y32 y2 n MSD n (1) ), lg( 10 /N MSD S (2)
where yisthemeasuredvalueofsurfaceroughnessandnisthe
num-ber ofexperiments intheexperiments. Ahighervalueof S/Nmeans
the signal is much higher than the random effects of noise factors.
Higher values of S/N ratios are described as control factor settings
thatminimizedtheeffectsofthenoisefactor;therefore,ahigh
signal-to-noiseratioisalwayspreferred.
The3Dresponsesurfaceplotsthatobtainedresponsesurface
meth-od by RSMmodel inMinitab16 software wereutilizedtospecify the
relationshipbetweentheWEDMparametersandsurfaceroughnessas
shown inFig.3. Responsesurface method is astatistical methodand
used todetermine the relation between various independent
parame-ters and dependent parameters. Figure 3 indicates the influence of
sparkgapvoltage,peakcurrent,andwiretensiononthemeanquality
of machined surface roughness during wire-EDM of MMC. The
ma-chinedsurfacequalitywasdecreasedwithanincreaseinthepeak
cur-rentand thebest surfaceroughnesswasobserved atlowestpeak
cur-rentandwiretension.Thiswasattributedtolowcuttingspeedat
low-estpeakcurrent.Thisiscausedbyincreaseofpeakcurrentthatleadto
a higher cutting speed and resulted the decreasing surface quality.
Normally,increasingwiretensionproducesanimprovedsurface
quali-tyofmachinedpartduetoreducingwirevibrationanddeflection[20].
On thecontrary, surface quality wasdecreasedwith increase in wire
tension inthisstudy.Thismaybeattributed toincreasingforces
act-ingonthewireelectrodeandwirebreakagewithincreaseinwire
higherwiretension.ThiscanbeascribedthattheharderB4Cparticles
causedfastwearofbrasswireathigherpeakcurrentandwiretension
dependsonincreasingcuttingtemperature.
Oneofthemostsignificantaimsofthisexperimentalstudyisto
de-termine anacceptablesurfaceroughnessusingoptimalmachining
pa-rameters.Thesignal-to-noiseratiosandresponsesurfaceoptimization
methods were performedinordertospecifythe bestcutting
parame-tersintheWEDMofMMCs.TheWEDMparameters,calculated
aver-age test results, desirability values, and the S/N ratios for surface
roughnessarelistedinTable4.Theoptimalwire-EDMparametersand
their levels were determined based on the S/N ratios (Table 4). The
higherS/N ratiosand compositedesirabilityvaluesindicatethe
opti-mum machining parameters and better quality ofsurface roughness.
The bestWEDMfactors basedontheresponseTable5forS/N inthe
machining ofAl6061/B4C,theoptimalsurfaceroughnessvalueswere
defined as factor A (Level 1, S/N 10.579), factor B (Level 2,
S/N 3.485), and factor C (Level 3, S/N 3.077). In the WEDM of
Al6061/B4C,thebestmachining parametersare determinedasapeak
currentof8A,sparkgapvoltageof68.89Vandwiretensionof10;the
optimizedsurfaceroughnessvalueisRa 2.8849mandthe
desirabil-ityvalueis0.97336asshowninFig.4.
TABLE4.Experimentalparametersandmeasuredsurfaceroughnessvalues. Trials number Wiretension (WT) Sparkgap voltage(SV) Peak current(IP) Surface roughness,Ra S/N ratio 1 10 30 8 3.06 9.714 2 10 30 10 3.3 10.370 3 10 30 13 4.01 12.063 4 10 60 8 2.89 9.218 5 10 60 10 3.43 10.706 6 10 60 13 3.78 11.550 7 10 80 8 2.85 9.097 8 10 80 10 3.4 10.630 9 10 80 13 3.92 11.866 10 13 30 8 3.3 10.370 11 13 30 10 3.68 11.317 12 13 30 13 4.02 12.085 13 13 60 8 3.21 10.130 14 13 60 10 3.44 10.731 15 13 60 13 4.16 12.382 16 13 80 8 3.15 9.966 17 13 80 10 3.78 11.550 18 13 80 13 3.96 11.954
The analysis of variance(ANOVA)and maineffect plots were
per-formed toinvestigatetheinfluencesparametersonsurfaceroughness
and contributionrate ofwire-EDM parameters onthe quality of
ma-chinedsurface.Thestatisticalsignificancelevelswereanalysedbythe
machining parametersP andFvaluesatthe95%confidencelevel.If
the P valuesare smaller than 0.05, theexperimentalmodels are
con-sidered ata significant levelof 95%. TheWEDM parameters,P
val-ues,andtheircontributionlevelforsurfaceroughnessarepresentedin
Table6.FromtheresultofANOVA,thepeakcurrentisthemost
effec-tivemachiningparameterswithan84.9%contributionoftotal
varia-tion onsurfaceroughnessintheWEDMof Al6061/B4C.The next
ef-fective WEDMparameter iswiretension with apercentage
contribu-tionof8.33%forAl6061/B4C.Itwasobservedthatthesparkgap
volt-age wasnot showed a meaningful effect onsurface roughness in the
WEDMofMMCs.
As showninmean effectplots inFigure 5, theeffect ofspark gap
voltageonsurfaceroughnesswasalmostconstant.Itcanbeseenfrom
themeaneffectplotsthatsurfacequalitywasdecreasedwith
increas-ingpeakcurrentfrom8Ato13Aandwiretensionfrom10gto13g.
At the base of the RSM and Taguchi methods, a regression analysis
equation forsurfaceroughness wasalso developed.Thefollowing
re-TABLE5.Responsetableforsignal-to-noiseratios(smallerisbetter).
Level Wiretension(WT) Sparkgapvoltage(SV) Peakcurrent(IP)
1 10.579 10.987 9.749
2 11.165 10.786 10.884
3 10.844 11.983
Delta 0.586 0.200 2.234
Rank 2 3 1
TABLE6.AnalysisofVariance(SS–sumsofsquares,MS–meansquare).
Source DF Sequential SS Adjusted SS Adjusted MS F P Significance level,% Regression 3 2.64982 2.64982 0.88327 68.562 0.000000 WT 1 0.23576 0.23576 0.23576 18.300 0.000766 8.33 V 1 0.01011 0.01011 0.01011 0.785 0.390594 0.35 IP 1 2.40395 2.40395 2.40395 186.602 0.000000 84.9 Error 14 0.18036 0.18036 0.01288 6.37 Total 17 2.83018
gressionequationswereobtainedforAl6061/B4Cmetalmatrix
compo-siteusingtheleast-squaremethodintheregressionanalysis.R2values
of the equations obtained from the regression for surface roughness
werecomputedas93.63%.
a 0.869055 0.0762963 0.00115351 0.177851 ,
R WT V IP (3)
q 93.63%.
RS
In ordertoverify the experimentalprocess, sixconfirmation
experi-ments were carried out within the limits of predetermined WEDM
conditions.Themeasuredsurfaceroughnesseswerecontrolledforthe
precision of the predicted values calculated from models.
Experi-mental values andpredictedvalues withthe percentageofprediction
errorratesarepresentedinTable7.AsseeninTable7,theestimated
Fig.5.Effectofmachiningparametersonsurfaceroughness.
TABLE7.Confirmationexperimentsandresults.
Wire tension Sparkgap voltage Peakcurrent (IP) Surfaceroughness Ra Predictedsurface roughness Prediction error 10 8 42 3.22 3.006 6.63% 13 8 42 3.24 3.235 0.15% 13 10 42 3.51 3.591 2.31% 10 10 42 3.72 3.362 9.62% 10 13 42 4.17 3.896 6.58% 13 13 42 4.24 4.125 2.72%
valuesbasedontheregressionmodelwiththeleastresidualerrorsare
very closetotheexperimentalresultsandpredictionerrorsareinthe
acceptablerange.
4.CONCLUSIONS
In this experimentalstudy, Aluminium 6061/B4C metal matrix
com-posite wassuccessfullyproduced byapowdermetallurgy methodand
investigatedfortheeffectofwire-EDMparametersonsurface
rough-ness using brass wire electrode. The WEDM experiments were
per-formed basedontheTaguchiL18orthogonalarray.Theinvestigation
results were examined using 3D surface plots, S/N ratio results,
ANOVA, and main effect graphs. The signal-to-noise ratios, central
composite desirability of response surface method, and regression
model were used to specify the ideal WEDM parameters for surface
roughness.
The following conclusions can be drawn from the experimental
study.
The optical and SEM micrographs indicatedthat theB4C particles
distribution is fairly homogenized in all composite specimens and
achievedagoodinterfacialbondingbetweenmatrixandB4Cparticles.
Themachinedsurfacequalitywasworseningwithanincreaseinthe
peak current and the best surface roughness was observed at lowest
peakcurrentandwiretension.Surfacequalitywasdecreasedwith
in-creaseinwiretension.
Brass wire electrode was broken at higher wire tension. It can be
suggested a very low wire tension to avoid wire breakage during
WEDMofMMCswithreinforcedB4C.
TheoptimalWEDMparametersaredeterminedasapeakcurrentof
8A, spark gap voltage of 68.89V, and wire tension of 10; the
opti-mized surface roughness value is Ra 2.8849m and the desirability
valueis0.97336.
FromtheresultofANOVA,thepeakcurrentisthemostsignificant
wire-EDM parameters with an 84.9% contribution of total variation
onsurfaceroughnessintheWEDMofAl6061/B4C.
The sparkgapvoltagedidnotshowameaningfuleffect onsurface
roughnessintheWEDMofMMCs.
The estimated values basedon theregression model with the least
residual errors are very close totheexperimental results and
predic-tionerrorsareintheacceptablerange.
TheauthorswishtothankHacettepeUniversityScientificResearch
Projects Coordination Unit for the financial support of this
experi-mental research supported by the Scientific Research Projects Grant
funding number #1743 and POYRAZCNC Company forWEDM
REFERENCES
1. H. Mindivan, M. Baydogan, E. S. Kayali, and H. T. Cimenoglu, Materials
Characterization,54:263(2005).
2. A. B. Sadat, MachiningofMetalMatrixComposites (Ed. P. J. Davim) (London: Springer:2012),p.51.
3. J. Wang, W. Lin, Z. Jiang, L. Duan, and G. Yang, CeramicsInternational,40:
6793(2014).
4. A. Manna and B. Bhattacharayya, J.MaterialsProcessingTechnology,140:
711(2003).
5. U. Soy, F. Ficici, and A. Demir, J.CompositeMaterials,46: 851 (2011). 6. M.El-GallabandM.Sklad,J.MaterialsProcessingTechnology,83:151(1998). 7. Ş. Karabulut, Measurement,66: 139 (2015).
8. A.Thuault,S.Marinel,E.Savary,R.Heuguet,D.Goeuriot,andD.Agrawal,
CeramicsInternational,39: 1215 (2013).
9. A.K.SahooandS.Pradhan,Measurement,46:306(2013).
10. J. Wang, W. Lin, Z. Jiang, L. Duan, and G. Yang, CeramicsInternational,40:
6793(2013).
11. G. Amitesh and K. Jatinder, InternationalJournalofAdvancedEngineering
Technology,3:170(2012).
12. A. R. Motorcu, E. Ekici, and A. Kuş, ScienceandEngineeringofComposite
Materials,doi:10.1515/secm-2014(2015).
13. B. H. Yan, H. C. Tsai, F. Y. Huang, and L. C. Lee, InternationalJournalof
MachineToolsandManufacture,45,Iss.3:251(2005).
14. P. Shandilya, P. K. Jain, and N. K. Jain, ProcediaEngineering,64: 767 (2013). 15. A.GoswamiandJ.Kumar,EngineeringScienceandTechnology,
anInternationalJournal,17, Iss. 4: 236 (2014).
16. D.Satishkumar,M.Kanthababu,V.Vajjiravelu,R.Anburaj,
N. T. Sunda-Rajan, and H. Arul, Int.J.Adv.Manuf.Tech.,56: 975 (2011). 17. K.ChiangandF.Chang,J.Mater.Process.Technol.,180:96(2006). 18. R. K. Fard, R. A. Afza, and R. Teimouri, J.Mater.Process,15: 483 (2013). 19. R.K.Roy,SocietyofManufacturingEngineers(Dearborn,Michigan:1990). 20. F. Nourbakhsh, K. P. Rajurkar, A. P. Malshe, and J. Cao, ProcediaCIRP,5: 13
