DOC

AXISYMMETRIC THIN-WALL TUBE OFFSET SPINNING

By Melanie Matthews,2014-06-20 05:55
11 views 0
AXISYMMETRIC THIN-WALL TUBE OFFSET SPINNING

    AXISYMMETRIC THIN-WALL TUBE

    OFFSET SPINNING

    CHINESEJOURNALOFMECHANICALENGINEERING

    Vo1.19,No.2,2006?2l7?

    KUANGWeihua

    FacultyofMaterialandEnergy,

    GuangdongUniversityofTechnology,

    Guangzhou510006,China

    XIAQinxiang

    RUANFeng

    CollegeofMechanicalEngineering,

    SouthChinaUniversityofTechnology,

    Guangzhou510640,China

    0INTRoDUCTIoN

    MECHANISMRESEARCHFOR3D

    NON.AISYMMETRICTHIN-WALL

    TUBEOFFSETSPINNING

    Abstract:Difrerenceofoffsetspinningwithconventiona1symmetricspinningisanalyzed.A

    3D

    FEMmode1foroffsettubeneck

    spinningisestablishedandthespinningprocessissimulatedby meansOfANSYSsoftware.Dynamicboundaryandcontactproblemsinsimulationaresolve

    d.1hn

    sients~essdistributionofcontactarea.transientstraindistributionofnodesintypicalsectiona

    nd

    straindistributionoftheworkpieceatlastareattained.theplaceandthecauseofcrackareanaly

    zed.

    Strainvariationcurveswithtimeofoffsetspinningandconventiona1spinningarecompared.Itshows

    themechanismdifferencebetweenoffsetspinningandconventiona1spinning.Inaddition,simulation

    resultsshowhowstraindistributionoftypicalsection,thicknessofsometypicalnodes,axia1exten

    sionin1eftsectionandforceofthreerollerschangewithtime.Accordingtothestudyofthevariation

    curve.materia1flow1awalongradial,tangentialandaxia1directionisattainedandthewholespinning

    processisstudied.Thesimulationresultsdiscoverthatoffsetdistanceisthekeytomanufactureoet

    non

    symmetrictube,andprocessparameterschangeWlthspinningangle.Experimentdatareally reflectdifferentprocessparameters'influenceonconventionalsymmetricandoffsetspinningforce.

    Experimentsaccordwellwithsimulation.

    Keywords:0ffsetSymmetricNeckspinningSimulationMechanism

    Spinning,asanimportantpartofadvancedplasticmanufac

    turingtechnologies,hasbecomeapopularprocessforthin-wall

    tubeforming,andplaysanimportantroleinmetalprecision

    processingfields,suchasaeronautics,weaponry.

    Theobviousdifierencebetweenoffsetandconventional

    symmetrictubeneck.spinningisthattubeinoffsetspinningdoes

    notrotate,butrollersrotatearoundtube.Beforeeverypass,tube

    movesalittledistanceinverticalaxis.direction.thenthreerollers

    moveinradialdirectionrespectivelyandrotatearoundtheofet

    axis.Indifferentpass.offsetdistancemaybedifferent.butoffset

    axismustbeparalle1.Non.axisymmetricoffsettube(Fig.1)can

    notbemanufactureduntilthetotaloffsetdistancereachesthe

valuefi[i-2].

    ROIl

    OffsetdistanceOffsetdistance

    Fig.1Schematicdiagramofoffsetspinning

    2

    Recentlytherearesomeresearchesonnon-axisymmetric spinning,buttheyfustfoCUSonqualitativeanalysisoftheproc. essandthedistributionofspinningforceL1-ZJ.0ffsetspinningis verycomplexandtheplasticformationissmallandchanges eve:13'time.Distributionofstressandstrainisnothomogeneous andchangesaccordingtorollertrace.Theprecisionandquality offormedtubeareinfluencedbyalotoffactors.suchasspin. ningstyle,offsetdistance,rollertrace,feedrate,rollerround. nessradius.passreductionandtubethicknesst.Bycomputer simulation,thispaperattemptstodiscovernonaxisymmetric

    deformationprincipleandtofindthelawofmetalflowand distributionofstressandstrain.

    111isprojectissupportedbyNationalNatura1ScienceFoundationofChina (No.50275054).ProvincialNaturalScieneeFoundationofGuangdong.China (No.020923).ProvincialScieneeandTechnologyKeyProjectofIndustryof Guangdong,China(No.2003Cl020l3)andHighLevelConstructionProject ofSouthChinaUniversityofTechnology,China(No.73056060).Received July2l,2oo5:receivedinrevisedformNovember29.2005;acceptedFeb

    ruary7,2006

    1FEMMoDEL

    ANSYSisoneofpopularFEMsoftware.Itcandealwith

    largedeformation.contactandhighlynonlinearprobleme-

    ciently.Inthepaper.a3DFEMmodelforoffsettube

    neckspinningisestablishedandthemodelcorrespondswellwith practicalprocess.Threerollersevenlydistributeatl2O.around

theworkpiece(Fig.21.

    Fig.2FEMmodel

    Forwardspinningisadoptedinthismode1.Thegeometric dimensionandtheworkingparametersareshowninTable1.The stress.strainrelationisff=569F..

    Table1Geometricdimensionandworkingparameters ParameterValueParameterValue

    DiameterofrollerDJmml50LengthOfnlbeL/nun50 DiameteroftubeDl/mm80Thicknessoftube6dram15 .ssmdiusofroll10Material.ftube0s8

    te

    A

    e

    L

    1Rn/mm

    ModulusofelasticityE|GPa200YieldpointoftubeMPa120 Poisson'sratio0.3Frictionalcoefficient"0.1 Namelypassreduction1S

    peedofrotationn/(r?min1400P'/mm

    Feedingratio9q(mm.1"-'11Ofrsetdistancemml 1.1Elementselectionandmesh

    ThetubeismeshedbvS0LID45.whichishexahedralele. mentwitheightnodes.Threeautomaticcontactpairsbetweenthe tubeandtherollersaresetbvANSYS.Threerollersareregarded asrigidpartsandthetubeisregardedasflexiblepart,theyare meshedbyTARGEl7OandCONTAl73elements.Thereare 2079elementsandl970nodesintheFEMmode1.Themeshed elementsareshowninTable2

.218'KUANGWeihua.etal:Mechanismresearchfor3Dnonaxisymmetricthin

    walltubeoffsetspinning

    Table2ElementtypeandnumberofFEMmodel

    ElementtypeNumberElementtypeNumber

    SOLID45396CONTA173924

    TARGE170756PILOT3

    1.2Applicationofdynamicboundarycondition

    Inthemode1.therighttubeisfixedandthelenisfree.the boundaryconditioncanbeappliedbythreerollers.Theroller traceisshowninFig.3.Threerollersmove1mn)innegative direction.thenmove2mminradialdirectionrespectivelyand rotatearoundtube'soffsetaxis.

    7

    Rollertmce~,

    ._I90y\I/100ffsetaxis

    

    +

    tistance}.

    Fig.3Schematicdiagramofrollertrace

    f=(0.5DIP)cosO

    {Y=(0.5P)sinO

    lz=z0+

    where0'——Degreeinoffsetcircle

    t.____________——Time

    1.3Handlingoffrictionalcontact

    Thecontactproblembetweentherollersandthepartishan. diedbycontactsearchalgorithm.Themotionconstraintandthe nodeforcewillberegardedasloadsandautomaticallybeapplied inthecontactnodesoncetheboundarycontacthappens.Coulomb frictionalmodelisadoptedandfrictioncoefficientis0.1.The

    frictionforcebetweentherollersandthetubecausesrollersspin automaticallyaroundtheirownaxiswhentheyrotatearoundthe tube.andleadsthetubematerialtoflowmorefreelyt. 2SIMULATIoNRESUrANDANALYSiS

    2.1Analysisoftransientstressandstrain

    Thematerialflowlawinradia1.tangentialandaxialdirection canbeattainedbyanalyzingthetypicalnodes'stressandstrain. andtheanalysiscanalsodiscoverthedeformationmechanism deeply.ThetypicalnodepositiondistributionisshowninFig.4. Threerollersdistributeevenlyin3.4sectionandthesign"o"in Fig.4arepresentsthecontactpositionofroller1.3withthetube. NodeCisthefirstspinningpositionandnodeIisthelastspin- ningplace.Thedistanceofeverynodeis5mn).

    Fig.5showsthestressdistributioninoutercontactsurface. Thesign"o"representsthecontactpositionoftherollerlwith thetube.Theradialstressiscompressive.buttherearesometan- gentialandaxialtensilestressarea.especiallyinthebackandthe fightoftheroller.Thetangentialandaxialtensionisbiganditis animportantcauseofsurfacecrack.Inaddition.the0.innersur. faceinthefirstneck,spinningsectionisalsoeasytocrackbe. causelargetangentialandaxialtensionoccursthere. Fig.6ashowsthestraincurveofoutersurfacenodesin1.2 section.Fig.6bshowsthestraincurveofoutersurfacenodesin 3-4section.Thesign"?"showstheapproximatepositionof

    rollersandthenumberrepresentstherollernumber.

    Fig.6ashowsthereisdramaticplasticdeformationalongthe radia1.tangentialandaxialdirectioninthecontactarea.Because ofneck.spinning.thecontactmaterialiscompressedalongtan. gentialdirection.anditflowsalongaxialandradialdirection.In

    addition.materialflowincontactareaisresictedbysurround materia1.Sothetangentialstrainincontactzoneiscompressive.

    buttheradialstrainandtheaxialstrainaretensile.Thebigger strainvaluelocatesindeformedzoneneartherollers.Fig.6b

    showsthatradialstrainandaxialstrainaretensilein3.4sec

    tion,andisbiggerthansa.Thetangentialstrainstiscompres?

    siveandisbiggerthanotherstrain.In3.4section.0.strainis

    biggerthanotheranglestrain.but180.sainissmaller. (a112section

    0.directic

    .

    M1

    ..,

    '

    1'n.

    70.

    KOter

    rrota

    

    0odirec

    ..,

    .

    (b)34section(c)Leftsection

    Fig.4Positionoftypicalnodes B=131026

    C=-90343

    D--49657

    E=-8972

    F=31714

    G=72400

70.

    A=296550

    B=233753

    C=170957

    D-1O816O

    E=45364

    F=17433

    G=80229

    H143026

    (b)Tangentialstress

    A=226500

    B171642

    C1l6784

    D61925

    E=-7067

    F=47791

    G=1O2649

    =157507

    Y

    :

    1:212.365

    (c)Axialstress Fig.5Stressdistributionofoutercontactarea

    Ja=oJJoco?.J1cl^o ,Ja=oJJocoJ1cl^o

    ,

    Ja=oJJocou.J1cI^o

CHINESEJOURNALOFMECHANICALENGINEERING?2l9?

C

    e

    AxiallengthL/mm

    fa1StraincurveIn12section

    er

    10n

    Angle0/(.)

    (b)Straincurvein3-4section

    Fig.6Straincurveoftypicalnodes

    +-?一?Ca'

    Fig.7showsthestraindistributionofthewholepartafterone pass.Alltangentialstrainiscompressivebecauseofneckspin

    ning.Thewholepartextendsalongaxialdirection,sotheaxial strainisalmosttensile.Thebiggesttangentialandaxialstrain locatesin0.zonenearthefirstnecksection.

    (a)Tangentialstrains

    003382

    Ca

    .

    6

    007331O

    O1l2793

    (b)Axialstrain

    Fig.7Straindistributionofwholetube

    2.2Strainvariationwithtime

    Fig.8showstypicalnodes'radia1.tangentialandaxialstrain variationwiththetimein12section.Fig.8ashowsalmostwhole

    radialstrainin12sectionincreasesfirstandgraduallydecreases later.Itindicatesthatthepart'sthicknessin0.areaincreasesfirst anddecreasesgraduallylater,butthethicknessincreasesatlast.

    ThereiscompressivestraininnodeB.itshowsthatthethickness decreasesheavilynearfirstneck-section.Inpracticalprocess. crackeasilyoccursinthisplaceandtheareaisadangeroussec

    tion.ThestrainofnodeIisbigger,whichindicatesthethick

    nessinthosesectionsincreasesmoreheavily.

    Fig.8bshowstangentialstrainofnodeB.CDdecreases sharplyfirstandkeepsconstantlater.Thestrainofothernodes decreasesgradually.ThetangentialstrainofnodeIistensilefirst andtheniscompressive.andtheabsolutevalueissmallerthan thatofnodeE.Fatlast.Itindicatesthatthediameteroftheleft sectionincreasesfirstandthendecreaseswithrollerfeed.butit reboundsalittleatlast.

    Fig.8cshowsthewholepartextendsandtheextensionof nodeD.Eisbigger.TheaxialstrainofnodeIiscompressivefirst andtensilelater.whichshowsthattheleftsectionshrinksfirstand extendsalongaxialdirectionlaterbecauseofaxialtension.The valueofnodeIisthesmallestbecausetheleftsectionisfreeand thediametereasilyreboundsintheend

    

    

    

    C

    

    

    

    fa1RadiaIstrain

    Timet/s

    (b)Tangentialstrain

    Timet/s

    (c)Axialstrain

Report this document

For any questions or suggestions please email
cust-service@docsford.com