INVESTIGATING TO THE EFFECT OF GAS METAL … TO THE EFFECT OF GAS METAL ARC WELD PARAMETERS ON WELD...

INVESTIGATINGTOTHEEFFECTOFGASMETALARCWELDPARAMETERSONWELDBEADHEIGHTUSINGD.O.E

TARUNKUMARJHA1,BHUVNESHBHARDWAJ2,KULBHUSHANBHAGAT3,VARUNSHARMA4

1Scholar,DepartmentofMechanicalEngineering,SureshGyanViharUniversity,Jagatpura,Jaipur,Rajasthan

2,3AssistantProfessor,DepartmentofMechanicalEngineering,SureshGyanViharUniversity,Jagatpura,Jaipur,Rajasthan

4AssistantProfessor,DepartmentofMechanicalEngineering,SLITUniversity,Punjab

Email:[email protected]

ABSTRACT

Weld surfacing isamethodofproducing surfaceswhichhavegoodcorrosionresistantpropertiesby sourceofdepositing/layingofstainlesssteelonlowcarbonsteelcomponentswithanabilityofgainingmaximumeconomyandenhancedlife.ThusstudyundertakenwastoshowofftheeffectofgivenparameterprocessinmechanizedGMAWprocessondifferentparametersof cladding.Thereforeanexperimental techniquebeadonplatehasbeenused,which includes takingofa singlebeadsofstainlesssteelusingsolidfillerwireofdia.1.2mmon8mmthicklowcarbonsteelusingvariousweldingconditions,aspertheourdesignmatrix.TogiveupaperfectresultinexperimentationastatisticalsoftwareDesignexpert9.3.0wasusedtocarryout themodel tests (using ANOVA) and the effects of different parameters investigatedwas represented in the form ofresponse surfaces and contour graphs. The finding of this studywould be beneficial in enriching the data bases for thepressurevesselindustry,chemicalindustry,petroleumandhydrogenstorageindustrywheresurfacinghavingaveryvitalrole.

Keywords‐MIGWeldingprocess,RSM,ANOVA,Parameteroptimizationprocess,304Steel,heightofbead.

1. INTRODUCTION

1.1. Introduction‐ Modern industrial products whichhavecontributedtotheprosperityofmankind,owetheirdevelopment and efficient performance to welding,which is now the universally accepted method ofpermanently joining metals and producing monolithicstructuresefficientlyandeconomically.Steelsbyfararethemostwidelyusedandweldedmaterials,chiefly,dueto the realness that they can be manufacturedcomparatively inexpensively in large amounts to veryexact stipulation and have an encompassing range ofmechanicalproperties.

1.2. SurfacingorCoating‐ Metal surfacing refers toprocess of depositing a metal on the surface of basemetal either to meet the dimensional requirements(reclaimingofwornoutparts)componentsortoimpartcertain surface properties such as immunity to wear,impact ,or corroding . Themain aimbehind both thesesituationsistosavemoneyandmaterialbycuttingcostsandreducingtheconsumptionofexpensivemetals

Reviewof availablepublished literatureonany specifictopicofresearchisthefirstvaluablesteptobetakentoget an idea about the present status of that particulartopic.Togetfamiliaritywiththeestablishedmethods&procedures of conductingwork, to avoid duplication ofwork & wastage of precious time, this stage ofinvestigationmustbecarriedoutascomprehensivelyaspossible.In this chapter, an effort has been made to present areviewoftheavailable&relevantinformationrelatedtothedifferentaspectsof the investigationcovered in thepresent dissertation work. The literature survey iscarriedonthefollowingtopics.1. Austeniticstainlesssteel.2. WeldcladdingbyGMAWofausteniticstainlesssteel.3. OptimizationusingRSM.

Due to this fact, the effect of major micro alloyingelementsonthemicrostructure&mechanicalpropertiesof theweld &HAZ has been the subject of interest formanyresearchers.Abrief reviewof theirwork isgivenbelow:

Properties ofASS clad layer is of prime importance forthe fabrication of equipments, M.R.EI‐ Hebeary & J.Rittinger , 1988 had worked on the factors affectingtoughness of austenitic stainless steel clad layer. Theexperiments were carried out to study the effect ofwelding variables on both weld bead geometry &chemistry. The test results revealed that the clad layersolidificationpattern,chemical&phasecompositioncansecure its freedom from solidification cracking. Theembrittlement of the clad layer may be overcome bydecreasing the PWHT temperature which affects theferritecontent&morphologythroughsegregationofthealloying elements & thus improving the solidificationconditions.

MIG cladding procedure is advantageous when only asmallsurfaceofanymaterialistobecladded,wheretheSAWbecomesuneconomical.

In 1988, D.K. Gautam &/M. Breazu [11] detailed &analyzed the effect of weld parameters like weldingcurrent,voltage,weldingspeed&pulsefrequencyonthebeadgeometry.Hesuggestedthat forMIGcladding, thebestresultscanbeachievedbyusingpulsearcatlowerfrequency (25Hz). The conventionalwelding procedureshouldbe avoidedwhen single layer technique isused.For special cases of cladding, when there is arequirement for higher content of Cr & Ni in the welddeposit, a wire with at least 30%more content is notavailable, thus the buffer layer technique should beapplied.

TARUN KUMAR JHA et al. DATE OF PUBLICATION: OCTOBER 07, 2014

ISSN: 2348-4098 VOLUME 2 ISSUE 7 SEP-OCT 2014

INTERNATIONAL JOURNAL OF SCIENCE, ENGINEERING AND TECHNOLOGY- www.ijset.in 1482

RichardE.Avery&DavidParsons1995hadworkedonthe practical application of cladding of ASS to thecryogenic vessels. He used ASS with 9% Ni for weldcladdingofstoragevessels&equipments&observedthatmechanical properties of ASS can be maintained to atemperature of ‐3200 F or above without thedeteriorationofmechanicalproperties&notendencyforcracktopropagate&nosignificantcorrosion.Hisresultssuggest that toughnesswasnot expected todeterioratebelow code minimums after much longer service lifethan 20 years. The application of various weldingprocesses for cladding using Duplex stainless steel hasbeenstudiedindetailabove.

Themost common problem in ASS cladding is that theweld cladding cracks during bend test & first layer ofweld cladding is found to consist of marten site. Theotherproblemencounteredisthesolidificationcrackingisthecentrelinecracking.Boththeseproblemsrelatetoexcessivedilutionofweldcladdingbythesubstrate.

In 1996, Damien J. kotecki worked on the weldingprocedure that limits the dilution & provided the ASScladding, containing small amount of ferrite for crackresistance in a single layer over mild steel. The othermajor problem associated with these steels is theirsusceptibility to inter granular corrosion (IGC) due tosensitization.

N.PARVAHAVARTHIN,hadgivendetailedinformation&showedthatwhenASSareextensivelyheatedorslowlycooled inthetemperaturerangeof1123Kto723K,Crrichcarbidesprecipitatesalongboundary, leadingtoCrdepletion in the vicinity of grain boundary. Whensensitized ASS was exposed to the corrosiveenvironment,CrdepletionzonesleadstoIGC.

K. luska , 2005 studied the influence of weldimperfections on short circuit GMAW are stability. HestudiedthattheMonitoringofweldingprocesscanplayimportant role in on‐line quality control of the GMAweldingprocess.Twosignals,momentaryarcpowerandmomentaryarcresistancewereincludedintheanalysisand the results of analysis were compared with theresults of statistic analysis of welding current andweldingvoltage. Sensitivityofmeanvalueandvarianceof signals to artificial disturbances of welding arc wastested.Theresult showedthat itwaspossible todetectimperfections related to phenomena that take place inweldingarcandaffect thegasshieldof theweldingarcorthelengthofwireextension.

Mr.se.Stefanijakarlicprof.Dr..ivansamardizicin2008by using RSM technique , MAG process withexperimental analysis method of values of weld beadwidthgivesusideathattheinfluenceofthetwoeffectsheatinputandgasflowonthevaluesofthebed.

Hyoung Jin park,Sehun ,Phee andMun jin karg in 2009usedRSMtechniquewithAcpulseMIGweldingprocesshavingvariousparameterssuchasenratio,gapbirding,Intermetallic compound layer, direct current electrodepositive&negativehavingexperimentalanalysisgivesusidea that EN ratio increased the decomposition rateincreasedandheatappliedtothebasemetaldecreased.

Againinyear2009MR.AjeetHooda,AshwniandSatpalSharma used RSM with MIG welding process applyingvarious experimental analysis method along variousparameters such as EN ratio, gas flow rate, speed topredict Maximum yield strength in AISI 1040 of bothtransverse and longitudinal at optimum values ofprocess variables, the longitudinal yield strength isgreaterthanthetransverseyieldstrength.

Inpresentwork,RSM is applied todetermine (1)underDOE he find that the value of depth of penetrationincreasedbyincreasingthevalueofweldingcurrentandgrainboundariesofthemicrostructurearevaried,Whenthe welding parameters are changed. (2)ANOVAmethod he found that Orthogonal array gives a set ofwellbalanced(minimum)experiments.

2. EXPERINENTALSETUPANDMETHODOLOGY

2.1.Methodology

To achieve the following objectives RSM methodologyhas been selected. This is relatively a new technique,related to, used for carrying out the design ofexperiments, theanalysisofvariance,andtheempiricalmodeling. The RSM criterion was developed to selectdesign points in a way that minimizes the varianceassociated with the estimates of specified modelcoefficients. Inasensethismethodismoreusefulthancentral composite design (a conventional responsesurface method) method that it demands smallernumberofexperiments tobeconductedandalso it cantackle categorical factors included in the design ofexperiments. Basic steps for achieving the desireobjectivesare:

DataCollection: Accordingtodesignmatrixbasedon RSM design matrix, The experiments will beconducted on mild steel plate using 304 stainlesssteelelectrodes.

Empirical Modeling: Development of empiricalmodel (relationshipbetweenGMAWresponsesandtheGMAWparameters)usingregressionanalysis.

Test foradequacyofdevelopmodel:Checkingofmodel significance, model terms significance usingANOVAanalysis.ThisempiricalmodelwillhelpfulinoptimalselectionofGMAWparameters.

2.2 OptimizationofGMAWParameters:‐

Design of experiments approach, by using of RSMtechnique was found to the best suitable optimizationtechniqueforpresentworkassatisfiedbytheliteraturesurvey.Ihavetakenfournumericfactorssuchasnozzleplate distance, open circuit voltage, welding speed andwire feed rate being identified for my present work.Basedupontheproblemformulatedordiscussedabovethefollowingobjectiveswereformulated

1. Parametric optimization of bead geometryparameters,beadheight

2.3.Materialsandequipment’sused

2.3.1Basematerial




Mildsteelplatesof8mmthicknessofsize300x100x10mmwhichwerecutfromtheflatsusingpowerhacksawafter that each blocks being used as the substratematerialforcladding.

2.3.2Fillerwire

Theausteniticstainlesssteelsolidwireofthetype304ofdiameter1.2mmisbeingusedinthepresentwork.ThephysicalandchemicalpropertiesoftheASStype304areasfollows:

Table2.1chemicalcompositionofthesolidwire304

2.3.3Shieldinggas

The shielding gas used in this investigation wasindustriallypureArgon.Theshieldinggaspressurewaskept constant and maintained at 22Lt/min for eachexperimentalrun.

2.3.4Powersource

Welding were carried out in a systematic andmechanizedwaybyusingmodelpowerwave‐355fromLincolnelectricco.,withaconstantvoltage,byapplyingthepowersourcehavingvariationofwirefeedrateandocvinstepsof0.05m/minand0.1vrespectively.

2.4.1 Identification, selection & establishment ofrangesofparameters

Thewholesetupcontainsidentificationoftheimportantvariables that having a heavy influence on this weldquality.Thevariablessuchasocv,wirefeedrate,nozzleplate distance and welding speed which will affect theweld bead dimensions being identified & selected.Various initial trial runs being carried out in order tofindtheoperatingrangeofthisprocessvariablesTHISisthefollowingtable,whichwasisbeingcomesoutofthetrialruns,showstherangeoftheprocessparametersattwo levelhigherand lower levelsasper thedemandofrequirementofdesignmatrix

Table2.4processparametersvariedattwolevels

Inputparameters units low(‐) High(+)

Nozzletoplatedistance(N) mm 6 14

Opencircuitvoltage(V) volts 15 27

Weldingspeed(s) Cm/min 20 40

Wirefeedrate(f) m/min 4 12

3. RESULTSANDDISCUSSION

Thecompeteresultofthe30experimentsperformedasper the experimental plan are shown inTable 3 alongwith run order selected at random these resultswereinput into the Design expert 9.3.0 software for furtheranalysis. After the examination of it summary, outputrevealed that the quadratic model significant forresponses and therefore it will be used for furtheranalysis.




TABLE(3)DESIGNMATRIXFORBEADHEIGHT

Factor1 Factor2 Factor3 Factor4 Response1

Std Run A:NPD B:OCVC:weldingspeed

D:wire feedrate Beadheight

mm volts cm/min m/min mm

20 1 10.00 33.00 30.00 9.00 2.21

15 2 6.00 27.00 40.00 12.00 3.26

24 3 10.00 21.00 30.00 15.00 3.98

7 4 6.00 27.00 40.00 6.00 2.37

9 5 6.00 15.00 20.00 12.00 4.32

1 6 6.00 15.00 20.00 6.00 3.23

30 7 10.00 21.00 30.00 9.00 2.42

14 8 14.00 15.00 40.00 12.00 2.82

18 9 18.00 21.00 30.00 9.00 2.33

19 10 10.00 9.00 30.00 9.00 3.28

17 11 2.00 21.00 30.00 9.00 3.79

13 12 6.00 15.00 40.00 12.00 3.26

4 13 14.00 27.00 20.00 6.00 1.98

21 14 10.00 21.00 10.00 9.00 2.76

29 15 10.00 21.00 30.00 9.00 2.69

27 16 10.00 21.00 30.00 9.00 2.39

2 17 14.00 15.00 20.00 6.00 2.93

3 18 6.00 27.00 20.00 6.00 2.69

5 19 6.00 15.00 40.00 6.00 2.69

12 20 14.00 27.00 20.00 12.00 2.87

10 21 14.00 15.00 20.00 12.00 3.46

23 22 10.00 21.00 30.00 3.00 2.34

25 23 10.00 21.00 30.00 9.00 2.8

22 24 10.00 21.00 50.00 9.00 1.8

16 25 14.00 27.00 40.00 12.00 2.24

28 26 10.00 21.00 30.00 9.00 2.98

8 27 14.00 27.00 40.00 6.00 1.9

26 28 10.00 21.00 30.00 9.00 2.45

6 29 14.00 15.00 40.00 6.00 2.54

11 30 6.00 27.00 20.00 12.00 3.82

3.1ANOVAanalysis

As mentioned above ANOVA is commonly used toperform test for significance of the regression model,testforsignificanceonindividualmodelcoefficientsandtestforlack–of‐fitmodel.Thisanalysiswascarriedoutforsignificance levelofα=0.05, i.e. forconfidence levelof 95%. ANOVA test for response surface quadraticmodel for surface roughness is summarized in Table 4.Table4showsthevalueof“prob.>F”formodelis0.0001which is less than 0.05, that indicates that model issignificant, which is desirable as it indicates that theterms in the model have a significant effect on theresponse. In the samemanner thevalueof “prob.>F”for lack‐of‐fit formaineffectofnozzletoplatedistance,open circuit voltage, wire feed rate and two levelinteractionofNPDandOCV,npdandwirefeedrate,ocvandwirefeedratearelessthan0.05sothesetermsaresignificantmodel terms.othermodel termscanbe saidto be not significant. These significantmode terms canbe removedandmayresult inan improvedmodel.Thevalue of “prob.>F for lack ‐of‐ fit is0.9877, which isgreater than0.05which indicates that lackof fit is stillinsignificant.

Table4

ANOVAforResponseSurfaceQuadraticmodel

Analysisofvariancetable[Partialsumofsquares‐ TypeIII]

Sumof

Mean F p‐value

Source Squares df Square Value Prob>F

Model 10.94 14 0.78 29.50 <0.0001 significant

A‐NPD 2.55 1 2.55 96.14 <0.0001

B‐OCV 1.63 1 1.63 61.61 <0.0001

C‐weldingspeed 1.57 1 1.57 59.27 <0.0001

D‐wirefeedrate 3.37 1 3.37 127.35 <0.0001

AB 0.12 1 0.12 4.62 0.0483

AC 0.034 1 0.034 1.29 0.2736

AD 0.17 1 0.17 6.34 0.0236

BC 0.068 1 0.068 2.55 0.1311

BD 0.038 1 0.038 1.43 0.2496

CD 0.15 1 0.15 5.74 0.0301

A^2 0.42 1 0.42 15.96 0.0012

B^2 0.057 1 0.057 2.13 0.1646

C^2 0.14 1 0.14 5.19 0.0377

D^2 0.61 1 0.61 23.03 0.0002

Residual 0.40 15 0.027

LackofFit 0.11 10 0.011 0.19 0.9877 notsignificant




PureError 0.29 5 0.058

CorTotal 11.34 29

Table5

Analysisofvariancetable[Partialsumofsquares‐TypeIII]

Sumof

Mean F p‐value

Source Squares df SquareValue Prob>F

Model 10.75 10 1.07 34.38 <0.0001 significant

A‐NPD 2.55 1 2.55 81.51 <0.0001

B‐OCV 1.63 1 1.63 52.23 <0.0001

C‐weldingspeed 1.57 1 1.57 50.25 <0.0001

D‐wirefeedrate 3.37 1 3.37 107.96 <0.0001

AB 0.12 1 0.12 3.92 0.0624

AD 0.17 1 0.17 5.38 0.0317

CD 0.15 1 0.15 4.87 0.0399

A^2 0.39 1 0.39 12.40 0.0023

C^2 0.17 1 0.17 5.35 0.0320

D^2 0.57 1 0.57 18.23 0.0004

Residual 0.59 19 0.031

LackofFit 0.31 14 0.022 0.38 0.9309 notsignificant

PureError 0.29 5 0.058

CorTotal 11.34 29

The 5 represents the ANOVA table for the reducedquadratic model surface roughness by selecting thebackwardeliminationproceduretoautomaticallyreducethe terms that are not significant. Table 5 shows Thatthe model is still significant after the backwardelimination.However,maineffectofnpd,ocv,wirefeedrateandtwolevelinteractionofnpdandocv,npdandwire feed rate and second ‐order effect of npd and ocvare the significant model terms. The main effect ofweldingspeedandsecondordereffectofweldingspeedare added to support hierarchy. The results show thatthenpd is themostdominating factorassociatedwithsurfaceroughnessandmaineffectofocvandwire feedrate provides secondary contribution to the surfaceroughness. The value of “prob.>F” for lack –of‐ fit is0.9309, whichisgreaterthan0.05whichindicatesthatlackoffitisstillinsignificant.

Table6

Std.Dev. 0.18

R‐Squared 0.9476

Mean 2.82

AdjR‐Squared 0.9201

C.V.% 6.27

PredR‐Squared 0.8809

PRESS 1.35

AdeqPrecision 22.744

TheR2valuewhichisthemeasureofproportionoftotalvariabilityexplainedbythemode, isequal to0.9476orclose to 1which is desirable .The adjusted R2 value isequalto0.9201itisparticularlyusefulwhencomparingmodels with different numbers of terms. The resultshows that theadjustedR22.744value isvalue isveryclose to ordinary R2 value Adequate precision value isequal to which is a ratio of signal to noise. A ratiogreater than 4 is desirable which indicates adequatemodel discrimination. Adequate precision valuecompares the range of predicted values at the designpoints to the average prediction error. The finalregression model for surface roughness in terms ofcodedfactorisrepresentedinequation(1)andthefinalempiricalmodels in termsofactual factorsaregiven inequation(2)forBeadHeightisgivenbelow

Equation (1 ) Final Equation in Terms of CodedFactors:

Height=+2.67 ̶0.33×A̶0.26×B̶0.26×C+0.37×D̶0.087×AB̶0.10×AD̶0.097×CD+0.12×C2+0.14×D2

Equation(2)finalequationintermsofactualfactors

Height = +2.95540 ̶ 0.075119 ×NPD ̶ 7.0.1389E̶ 0.003 ×OCV + 0.050060 ×

Welding speed + 0.022560 ×Wire feed rate ̶ 3.64583E̶ 0.03× NPD × OCV̶

8.54167E̶0.03×NPD×Wirefeedrate ̶3.2500E̶0.003NPD2+7.35491E̶004×

Weldingspeed2+0.015853×Wirefeedrate2.

4.2Optimizationofcoatingcondition

In the present study, the aim was to find the optimalvalues of coating condition parameters in order tominimize the value of surface roughnessRa during thecoating process. The optimal solutions are reported inTable7

Table7

4.3Confirmationexperiments

Statistically developed mathematical model for Ra , isgivenbyequationwasalreadyvalidatedthroughF‐testsand lack – of‐ fit test. Fitted model found to besignificant, and their lack of fit is insignificant. Thecoefficientofvariation(R2)formodelwas0.9505,whichindicatesthemodel;abilityformakingpredictions.Thisconclusion is further supported through theconfirmationruns.Asetofthreeconfirmationrunsbeenperformed to verify the prediction ability of thedeveloped surface roughness model. The details of theconfirmationaregiveninTable8.ThepercentageerrorbetweentheexperimentalandthepredictedvaluesofRais found tobe less than5%. In theotherwords, all theexperimental values within 95% prediction interval,which clearly demonstrates the accuracy of themodelsdevelopedinthestudy.




Table8

ConfirmationReport

Two‐sided

Confidence=95% n= 1

Factor Name LevelLowLevel

HighLevel

Std.Dev.

Coding

A NPD 10.00 6.00 14.00 0.000 Actual

B OCV 21.00 15.00 27.00 0.000 Actual

C weldingspeed

30.00 20.00 40.00 0.000 Actual

Dwire feedrate

9.00 6.00 12.00 0.000 Actual

Predicted

Predicted

Response

Mean Median Observed

StdDev nSEPred

95%PIlow

DataMean

95%PIhigh

Untitled3

2.67357 2.67357 ‐ 0.176809

10.19 2.28

3.07

Graphsforsatisfactionofrequiredresult

4. RESULTANDCONCLUSION

Thefollowingconclusionsareasfollows:

1. Withincreaseinwirefedrate,depthofpenetrationwillincrease.

2. Whenfirefeedrateincreases,itsweldbeadratealsoincreases.Inopencircuitvoltageandnozzletoplatedistancebutdecreasesinweldingspeed.

3. A five level four factor full factorial design matrixbased on the central composite rotatable designtechnique can be used for the development ofmathematicalmodelstopredict.

4. Out of the four process variables considered wirefeed rate was the most significant and influentialfactorhavingthepositiveeffect.

REFERENCES

[1]. Gautam,D.K.&Breazu,M.1988,“Someaspectsof MIG stainless steel cladding”, Int. conference onwelding technology,University of Roorkee, India, PP 1‐145toI‐151.

[2]. Puja,M.G.,Dayal,R.K.,Gill,T.P.S.,GnanamoorthyJ.B. 1988, “Corrosion of thermally aged Type 316stainless steel weld metals in an acidic chloridemedium”, Int. conference on welding technology,UniversityofRoorkee,PPIndia,II‐33toII‐38.

[3]. Ei‐Hebeary, M.R. Rittinger, J. 1988, “Factorsaffecting toughness of austenitic clad layers”, Int.conference on welding technology, University ofRoorkee,India,PPII‐87toII‐92.

[4] Mallya,U.D.andSrinivas,H.S.1993,“Effectsofmagnetic steering of the arc on clad quality in thesubmerged arc strip cladding”, Welding Journal, PP289to293s.

[5] Avery, R.E. & Parsons, D. 1995, “ Weldingstainless and 9% Nickel steel cryogenic vessels”,WeldingJournal,PP45to49.

[6] Kotecki, D. 1996, “Dilution control in singlewire stainless submerged arc cladding”, weldingjournal,PP45to49.




[7] Murugan,N.andParmar,R.S.May1997,“Effectof welding condition on microstructure andpropertiesofType316Lstainlesssteelsubmergedarccladding”,weldingresearchsupplement,AWS,pp210sto220s.

[8]. Murugan, N. and Parmar, R.S. Oct 1997,“Stainless steel claddingdepositedbyAutomaticGasMetalArcWelding”,Welding Journal,AWS,pp391s to403s.

[9] Sun, Z., Kuo, M. and Pan, D. 1999, “Twinwiregastungstenarccladdingoffersincreaseddepositionrates”,WeldingJournal,pp61to64.

[10] Zumelzu, E., Sepulveda, J., Ibarra, M. 1999,“Influence of microstructure on the mechanicalbehaviorofwelded316LSSjoints”,JournalofMaterialsProcessingTechnology94,pp36‐40.




INVESTIGATING TO THE EFFECT OF GAS METAL … TO THE EFFECT OF GAS METAL ARC WELD PARAMETERS ON WELD...

Documents

Transcript of INVESTIGATING TO THE EFFECT OF GAS METAL … TO THE EFFECT OF GAS METAL ARC WELD PARAMETERS ON WELD...