拉深模設(shè)計(jì)中拉深壁起皺的分析-外文翻譯本科學(xué)位論文

畢業(yè)設(shè)計(jì)(論文)外文資料翻譯學(xué)院(系)：專業(yè)：姓名：學(xué)號(hào)：外文出處：AnalysisofDraw-WallWrinklinginaStampingDieDesign附件：1.外文資料翻譯譯文；2.外文原文。指導(dǎo)教師評(píng)語：簽名：年月日注：請(qǐng)將該封面與附件裝訂成冊(cè)。附件1：外文資料翻譯譯文拉深模設(shè)計(jì)中拉深壁起皺的分析起皺是金屬板料成形中常見的失效形式之一。由于功能和視覺效果的原因，起皺通常是不能為零件制品所能接受的。在金屬板料成形加工中通常存在三種類型的起皺現(xiàn)象：法蘭起皺；側(cè)壁起皺和由于殘余壓應(yīng)力在未變形區(qū)產(chǎn)生的彈性變形。在沖壓復(fù)雜形狀的時(shí)候，拉深壁起皺就是在模具型腔中形成的褶皺。由于金屬板料在拉深壁區(qū)域內(nèi)相對(duì)無支撐，因此，消除拉深壁起皺比抑制法蘭起皺要難得多。我們知道在不被支撐的拉深壁區(qū)域中材料的外力拉深可以防止起皺，這可以在實(shí)踐中通過增加壓邊力而實(shí)現(xiàn)，但是運(yùn)用過大的拉深力會(huì)引起破裂失效。因此，壓邊力必須控制在一定的范圍內(nèi)，一方面可以抑制起皺，另一方面也可以防止破裂失效。合適的壓邊力范圍是很難確定的，因?yàn)槠鸢櫾诶盍慵闹行膮^(qū)域以一個(gè)復(fù)雜的形狀形成，甚至根本不存在一個(gè)合適的壓邊力范圍。為了研究起皺的原因，Yoshidaetal.發(fā)明了一個(gè)試驗(yàn)，即：一張薄板延著對(duì)角的一個(gè)方向進(jìn)行不均勻拉深。他們還提出了一個(gè)近似的理論模型，起皺的初始是由于彈性變形導(dǎo)致橫向壓力發(fā)展成為不均勻的壓力場。Yuetal.用試驗(yàn)和理論分析的方法來研究起皺問題。他們發(fā)現(xiàn)根據(jù)他們的理論分析，起皺發(fā)生在兩個(gè)環(huán)形的起伏處，而且試驗(yàn)結(jié)果指出了4—6處起皺。Narayanasamy和Sowerby通過圓錐形凸模和半球形凸模的拉深來研究金屬板料的起皺。同時(shí)，他們也試圖整理防止發(fā)生起皺的特性參數(shù)。一個(gè)有斜度的方形盒，如圖1（a）所示，盒形件的每一個(gè)傾斜的拉深壁都與圓錐盒形件相似。拉深成形過程中，在拉深壁處的金屬板料是相對(duì)無支撐的，因此，褶皺是傾斜的。在目前的研究中，各種關(guān)于起皺的成型過程參數(shù)都被研究。在帶有階梯的方形盒件的研究中，如圖1（b）所示，觀察到了另一種類型的起皺。在當(dāng)前的研究中，為了得出分析的效果，實(shí)際生產(chǎn)用階梯形結(jié)構(gòu)的零件來研究。使用有限元方法可以分析出起皺的原因，并且可以使一個(gè)最優(yōu)的模具設(shè)計(jì)消除起皺現(xiàn)象。有限元分析使得模具設(shè)計(jì)在實(shí)際生產(chǎn)中更為合理化。（a）帶有斜度的方形盒件（b）帶有階梯的方形盒件圖11有限元模型模具的幾何結(jié)構(gòu)（包括凸模、凹模、壓邊裝置等等），通過使用CAD和PRO/ENGINEER來設(shè)計(jì)。使用CAD將3個(gè)節(jié)點(diǎn)或4個(gè)節(jié)點(diǎn)形成殼形的單體，進(jìn)而在模型上形成網(wǎng)格體系。使用有限元模擬，模型被視為是剛性的，并且相對(duì)應(yīng)的網(wǎng)格僅僅可以定義模型的幾何形狀，不能對(duì)壓力進(jìn)行分析。使用CAD所建立的4個(gè)節(jié)點(diǎn)的殼形單體可以為板料創(chuàng)建網(wǎng)格體系。圖2給出了模型完全建立時(shí)的網(wǎng)格體系和用以成形帶有斜度的方形盒件的金屬板料。由于對(duì)稱的原因，僅僅分析了零件的1/4。在模擬過程中，金屬板料放在壓邊裝置上，凹模向下移動(dòng)，夾緊板料。凸模向上移動(dòng)，拉深板料至模具型腔。為了精確的完成有限元分析，金屬板料的實(shí)際壓力——拉力的關(guān)系需要輸入相關(guān)的數(shù)據(jù)。從目前的研究來看，金屬板料的深拉深的特性參數(shù)已經(jīng)用于模擬。一個(gè)拉深的實(shí)驗(yàn)已經(jīng)用于樣品的生產(chǎn)，并且沿著壓延方向和與壓延方向成45°和90°的方向切斷。平均的流動(dòng)壓力σ可以通過公式σ=（σ0+2σ45+σ90）/4，計(jì)算出來，進(jìn)而準(zhǔn)確測量出實(shí)際拉力，如圖2所示，以用于帶有斜度的方形盒件和帶有階梯的方形盒件的拉深。圖2目前研究中的所有模擬都在SGIIndigo2工作站使用有限元可調(diào)拉深程序完成。完成了用于模擬所需數(shù)據(jù)的輸入（假定凹模速度為10m/s，并且平均摩擦系數(shù)為0.1）。2帶有斜度的方形盒件的起皺2.1凸模間隙的影響為了研究凸模間隙對(duì)起皺的影響，現(xiàn)在分別用凸模間隙為20mm，30mm和50mm的帶有斜度的方形盒進(jìn)行拉深模擬。在每次模擬拉深中，凹模口部尺寸為200mm固定不變，并且拉深高度均為100mm。在3次模擬中，均使用尺寸為380mm×380mm的方形板料，且板料厚度均為0.7mm，凹模對(duì)板料的壓力——拉力關(guān)系，如圖3所示。圖.3模擬結(jié)果表明：三個(gè)有斜度的方形盒均發(fā)生了起皺現(xiàn)象，圖3給出了凸模間隙為50mm的方形盒的形狀。從圖3可以看出，起皺分布在拉深壁處，并且拉深壁鄰近的拐角處起皺現(xiàn)象尤為嚴(yán)重。經(jīng)分析，在拉深過程中，起皺是由于拉深壁處存在過大的無支撐區(qū)域，而且凸模頂部和凹模口部長度的不同是由于凸模間隙的存在。在凸模頂部與凹模之間的金屬板料的延伸變得不穩(wěn)定，是由于斷面壓力的存在。在壓力作用下，金屬板料的無約束拉深是在拉深壁處形成褶皺的主要原因。為了比較三個(gè)不同凸模間隙的試驗(yàn)結(jié)果，需要引入兩個(gè)主應(yīng)力的比值β，β為εmin/εmax,εmin/εmax是主應(yīng)力相對(duì)的最小值和最大值。Hosford和Cadde指出，β值比臨界值更重要，如果起皺發(fā)生，那么β值越大，起皺現(xiàn)象就可能越嚴(yán)重。如圖4和圖4的曲線所示，三次不同凸模間隙的拉深模擬，沿M——N截面的相同拉深高度處的β值。從圖4可以看出，在3次模擬中位于拉深壁的拐角處起皺比較嚴(yán)重，在拉深壁的中間起皺比較弱。還可以看出，凸模間隙越大，比值β就越大。因此，增加凸模間隙將可能增加帶有斜度的方形盒件在拉深壁處起皺的可能性。2.2壓邊力的影響眾所周知，增加壓邊力可以幫助削弱拉深過程中發(fā)生的褶皺。為了研究增加壓邊力的影響，采用凸模間隙為50mm，不同的壓邊力數(shù)值來對(duì)有斜度的方形盒進(jìn)行拉深起皺的模擬。壓邊力從100KN增加到600KN，以提供壓邊力0.33Mpa到1.98Mpa。其他模擬條件和先前的規(guī)定保持一致（在模擬當(dāng)中采用了300KN的壓邊力）。模擬結(jié)果表明：增加壓邊力并不能消除拉深壁處起皺現(xiàn)象的發(fā)生。如圖4所示，在M-N截面處的β值，和壓邊力分別為100KN、600KN的拉深相比較，模擬結(jié)果指出，在M-N截面處的β值都是相同的。為了分析兩次不同壓邊力時(shí)出現(xiàn)起皺的不同，從拉深壁頂部到直線M-N處，對(duì)5處不同高度截面進(jìn)行了分析，如圖4所示，圖5給出了所有情況的曲線。從圖5可以看出，幾種情況截面處的波度是相似的。這就證明壓邊力與有斜度的方形盒件拉深中的起皺現(xiàn)象無關(guān)，因?yàn)轳薨櫟男纬芍饕怯捎诶畋谔幋竺娣e無支撐區(qū)域存在較大的橫斷面壓力，所以壓邊力并不影響凸模頂部與凹模肩部之間的制件形狀的不穩(wěn)定狀況。圖4圖5(a)100KN.(b)600KN.3帶有階梯的方形盒件在帶有階梯的方形盒件的拉深中，即使凸模間隙不是這樣重要，而在拉深壁處仍然會(huì)發(fā)生起皺。圖1（b）所示為帶有階梯的方形盒件拉深用的凸模，圖1（b）給出了拉深壁C和階梯處D、E。目前，實(shí)際生產(chǎn)中一直在研究這種類型的幾何結(jié)構(gòu)。生產(chǎn)中，板料的厚度為0.7mm，壓力-拉力關(guān)系從應(yīng)力試驗(yàn)中獲得。這種拉深件的生產(chǎn)是通過深拉深和整形兩個(gè)工序組成的。由于凸模拐角處的小圓角半徑和復(fù)雜的幾何結(jié)構(gòu)，導(dǎo)致在盒形件的頂部邊緣發(fā)生破裂，在盒形件的拉深壁處發(fā)生褶皺，如圖6所示。從圖6中可以看出，褶皺分布在拉深壁處，尤其在階梯邊緣的拐角處更為嚴(yán)重，如圖1（b）所示的A-D和B-E處。金屬板料在凸模頂部的邊緣開裂，進(jìn)而形成破裂，如圖6所示。圖6圖7為了對(duì)拉深過程中金屬板料出現(xiàn)的變形現(xiàn)象有更進(jìn)一步的了解，生產(chǎn)中仍然采用了有限元分析方法。最初的設(shè)計(jì)已經(jīng)用有限元模擬完成。模擬的盒形件外形如圖7所示。從圖7可以看出，盒形件頂部邊緣的網(wǎng)絡(luò)拉深比較嚴(yán)重，褶皺分布在拉深壁處，這與實(shí)際生產(chǎn)中的狀況是一致的。小的凸模圓角，例如A-B邊緣的圓角和凸模拐角A處的圓角，如圖1（b）所示，是拉深壁處破裂的主要原因。然而，根據(jù)有限元分析的結(jié)果，通過加大上述兩處圓角可以避免破裂的產(chǎn)生。較大的拐角圓角這種想法通過實(shí)際生產(chǎn)加工被驗(yàn)證是可行的。在拉深工序中采用有限元分析的優(yōu)點(diǎn)之一就是可以通過拉深模擬來監(jiān)視、控制金屬板料的形狀變形，而這些在實(shí)際生產(chǎn)中是不可能做到的。在拉深過程中，仔細(xì)地看金屬板料的流動(dòng)，可以看出金屬板料首先由凸模拉深進(jìn)凹模腔內(nèi)，直到金屬板料到階梯邊緣D——E處時(shí)，褶皺才開始形成。褶皺的形狀如圖8所示。有限元分析還可以為模具設(shè)計(jì)的改進(jìn)提供相關(guān)的數(shù)據(jù)信息。、圖84簡要論點(diǎn)及結(jié)束語在拉深過程中發(fā)生的兩種類型的褶皺通過有限元分析研究以及對(duì)起皺原因做的試驗(yàn)，最終發(fā)現(xiàn)了抑制起皺的方法。第一種類型的起皺出現(xiàn)在帶有斜度的方形盒件的拉深壁處。在凹?？诓康母叨瘸叽绾屯鼓ｍ敳康母叨瘸叽绲纫蛩刂?，起皺的發(fā)生歸因于較大的凸模間隙。較大的凸模間隙會(huì)導(dǎo)致拉深到凸模頂部與凹模肩部的金屬板料處產(chǎn)生較大的無支撐區(qū)域，而金屬板料較大的無支撐區(qū)域是形成起皺的最終原因。有限元模擬表明這種類型的起皺是不能通過增加壓邊力而抑制的。另一種類型的起皺發(fā)生在實(shí)際生產(chǎn)中帶有階梯的幾何結(jié)構(gòu)的方形盒件中。研究發(fā)現(xiàn)即使凸模間隙影響不是很重要，起皺還是會(huì)發(fā)生在階梯上面的拉深壁處。根據(jù)有限元分析，起皺的原因主要是由于凸模頂部和臺(tái)階邊緣之間的不均勻拉深造成的。為了避免起皺，在模具設(shè)計(jì)中使用有限元模擬做了一些試驗(yàn)，試驗(yàn)最終確定的最優(yōu)設(shè)計(jì)就是將階梯去除。修改后的模具設(shè)計(jì)生產(chǎn)出了無缺陷的盒形零件。模具分析的結(jié)果和實(shí)際生產(chǎn)所獲得的結(jié)論證明了有限元分析的準(zhǔn)確性和使用有限元模擬的有效性。因此，可以說：有限元方法可以取代傳統(tǒng)的實(shí)際生產(chǎn)試驗(yàn)的昂貴的方法。附件2：外文原文AnAnalysisofDraw-WallWrinklinginaStampingDieDesignWrinklingisoneofthemajordefectsthatoccurinthesheetmetalformingprocess.Forbothfunctionalandvisualreasons,wrinklesareusuallynotacceptableinafinishedpart.Therearethreetypesofwrinklewhichfrequentlyoccurinthesheetmetalformingprocess:flangewrinkling,wallwrinkling,andelasticbucklingoftheundeformedareaowingtoresidualelasticcompressivestresses.Intheformingoperationofstampingacomplexshape,draw-wallwrinklingmeanstheoccurrenceofwrinklesinthediecavity.Sincethesheetmetalinthewallareaisrelativelyunsupportedbythetool,theeliminationofwallwrinklesismoredifficultthanthesuppressionofflangewrinkles.Itiswellknownthatadditionalstretchingofthematerialintheunsupportedwallareamaypreventwrinkling,andthiscanbeachievedinpracticebyincreasingtheblank-holderforce;buttheapplicationofexcessivetensilestressesleadstofailurebytearing.Hence,theblank-holderforcemustliewithinanarrowrange,abovethatnecessarytosuppresswrinklesontheonehand,andbelowthatwhichproducesfractureontheother.Thisnarrowrangeofblank-holderforceisdifficulttodetermine.Forwrinklesoccurringinthecentralareaofastampedpartwithacomplexshape,aworkablerangeofblank-holderforcedoesnotevenexist.Inordertoexaminethemechanicsoftheformationofwrinkles,Yoshidaetal.[1]developedatestinwhichathinplatewasnon-uniformlystretchedalongoneofitsdiagonals.Theyalsoproposedanapproximatetheoreticalmodelinwhichtheonsetofwrinklingisduetoelasticbucklingresultingfromthecompressivelateralstressesdevelopedinthenon-uniformstressfield.Yuetal.[2,3]investigatedthewrinklingproblembothexperimentallyandanalytically.Theyfoundthatwrinklingcouldoccurhavingtwocircumferentialwavesaccordingtotheirtheoreticalanalysis,whereastheexperimentalresultsindicatedfourtosixwrinkles.NarayanasamyandSowerby[4]examinedthewrinklingofsheetmetalwhendrawingitthroughaconicaldieusingflat-bottomedandhemispherical-endedpunches.Theyalsoattemptedtorankthepropertiesthatappearedtosuppresswrinkling.Ataperedsquarecup,asshowninFig.1(a),hasaninclineddrawwalloneachsideofthecup,similartothatexistinginaconicalcup.Duringthestampingprocess,thesheetmetalonthedrawwallisrelativelyunsupported,andisthereforepronetowrinkling.Inthepresentstudy,theeffectofvariousprocessparametersonthewrinklingwasinvestigated.Inthecaseofasteppedrectangularpart,asshowninFig.1(b),anothertypeofwrinklingisobserved.Inordertoestimatetheeffectivenessoftheanalysis,anactualproductionpartwithsteppedgeometrywasexaminedinthepresentstudy.Thecauseofthewrinklingwasdeterminedusingfinite-elementanalysis,andanoptimumdiedesignwasproposedtoeliminatethewrinkles.Thediedesignobtainedfromfinite-elementanalysiswasvalidatedbyobservationsonanactualproductionpart.Fig.11.Finite-ElementModelThetoolinggeometry,includingthepunch,dieandblank-holder,weredesignedusingtheCADprogramPRO/ENGINEER.Boththe3-nodeand4-nodeshellelementswereadoptedtogeneratethemeshsystemsfortheabovetoolingusingthesameCADprogram.Forthefinite-elementsimulation,thetoolingisconsideredtoberigid,andthecorrespondingmeshesareusedonlytodefinethetoolinggeometryandarenotforstressanalysis.ThesameCADprogramusing4-nodeshellelementswasemployedtoconstructthemeshsystemforthesheetblank.Figure2showsthemeshsystemforthecompletesetoftoolingandthesheet-blankusedinthestampingofataperedsquarecup.Owingtothesymmetricconditions,onlyaquarterofthesquarecupisanalysed.Inthesimulation,thesheetblankisputontheblank-holderandthedieismoveddowntoclampthesheetblankagainsttheblank-holder.Thepunchisthenmoveduptodrawthesheet-metalintothediecavity.Inordertoperformanaccuratefinite-elementanalysis,theactualstress–strainrelationshipofthesheetmetalisrequiredaspartoftheinputdata.Inthepresentstudy,sheetmetalwithdeep-drawingqualityisusedinthesimulations.Atensiletesthasbeenconductedforthespecimenscutalongplanescoincidingwiththerollingdirection(0°)andatanglesof45°and90°totherollingdirection.Theaverageflowstress,calculatedfromtheequationσ=（σ0+2σ45+σ90）/4,foreachmeasuredtruestrain,asshowninFig.2,isusedforthesimulationsforthestampingsofthetaperedsquarecupandalsoforthesteppedrectangularcup.Fig.2AllthesimulationsperformedinthepresentstudywererunonanSGIIndigo2workstationusingthefinite-elementprogramPAMFSTAMP.Tocompletethesetofinputdatarequiredforthesimulations,thepunchspeedissetto10m/sandacoefficientofCoulombfrictionequalto0.1isassumed.2.WrinklinginaTaperedSquareCup2.1EffectofDieGapInordertoexaminetheeffectofdiegaponthewrinkling,thestampingofataperedsquarecupwiththreedifferentdiegapsof20mm,30mm,and50mmwassimulated.Ineachsimulation,thediecavityopeningisfixedat200mm,andthecupisdrawntothesameheightof100mm.Thesheetmetalusedinallthreesimulationsisa380mm_380mmsquaresheetwiththicknessof0.7mm,thestress–straincurveforthematerialisshowninFig.3.Fig3Thesimulationresultsshowthatwrinklingoccurredinallthreetaperedsquarecups,andthesimulatedshapeofthedrawncupforadiegapof50mmisshowninFig.4.ItisseeninFig.4thatthewrinklingisdistributedonthedrawwallandisparticularlyobviousatthecornerbetweenadjacentwalls.Itissuggestedthatthewrinklingisduetothelargeunsupportedareaatthedrawwallduringthestampingprocess,also,thesidelengthofthepunchheadandthediecavityopeningaredifferentowingtothediegap.Thesheetmetalstretchedbetweenthepunchheadandthediecavityshoulderbecomesunstableowingtothepresenceofcompressivetransversestresses.Theunconstrainedstretchingofthesheetmetalundercompressionseemstobethemaincauseforthewrinklingatthedrawwall.Inordertocomparetheresultsforthethreedifferentdiegaps,theratio_ofthetwoprincipalstrainsisintroduced,_being_min/_max,where_maxand_minarethemajorandtheminorprincipalstrains,respectively.HosfordandCaddell[5]haveshownthatiftheabsolutevalueof_isgreaterthanacriticalvalue,wrinklingissupposedtooccur,andthelargertheabsolutevalueof_,thegreateristhepossibilityofwrinkling．The_valuesalongthecross-sectionM–Natthesamedrawingheightforthethreesimulatedshapeswithdifferentdiegaps,asmarkedinFig.4,areplottedinFig.5.ItisnotedfromFig.5thatseverewrinklesarelocatedclosetothecornerandfewerwrinklesoccurinthemiddleofthedrawwallforallthreedifferentdiegaps.Itisalsonotedthatthebiggerthediegap,thelargeristheabsolutevalueof_.Consequently,increasingthediegapwillincreasethepossibilityofwrinklingoccurringatthedrawwallofthetaperedsquarecup.2.2EffectoftheBlank-HolderForceItiswellknownthatincreasingtheblank-holderforcecanhelptoeliminatewrinklinginthestampingprocess.Inordertostudytheeffectivenessofincreasedblank-holderforce,thestampingofataperedsquarecupwithdiegapof50mm,whichisassociatedwithseverewrinklingasstatedabove,wassimulatedwithdifferentvaluesofblank-holderforce.Theblank-holderforcewasincreasedfrom100kNto600kN,whichyieldedablank-holderpressureof0.33MPaand1.98MPa,respectively.Theremainingsimulationconditionsaremaintainedthesameasthosespecifiedintheprevioussection.Anintermediateblank-holderforceof300kNwasalsousedinthesimulation.Thesimulationresultsshowthatanincreaseintheblankholderforcedoesnothelptoeliminatethewrinklingthatoccursatthedrawwall.The_valuesalongthecross-sectionM–N,asmarkedinFig.4,arecomparedwithoneanotherforthestampingprocesseswithblank-holderforceof100kNand600kN.Thesimulationresultsindicatethatthe_valuesalongthecross-sectionM–Narealmostidenticalinbothcases.Inordertoexaminethedifferenceofthewrinkleshapeforthetwodifferentblank-holderforces,fivecross-sectionsofthedrawwallatdifferentheightsfromthebottomtothelineM–N,asmarkedinFig.4,areplottedinFig.5forbothcases.ItisnotedfromFig.5thatthewavinessofthecross-sectionsforbothcasesissimilar.Thisindicatesthattheblank-holderforcedoesnotaffecttheoccurrenceofwrinklinginthestampingofataperedsquarecup,becausetheformationofwrinklesismainlyduetothelargeunsupportedareaatthedrawwallwherelargecompressivetransversestressesexist.Theblank-holderforcehasnoinfluenceontheinstabilitymodeofthematerialbetweenthepunchheadandthediecavityshoulder.Fig4Fig53.SteppedRectangularCupInthestampingofasteppedrectangularcup,wrinklingoccursatthedrawwalleventhoughthediegapsarenotsosignificant.Figure1(b)showsasketchofapunchshapeusedforstampingasteppedrectangularcupinwhichthedrawwallCisfollowedbyastepD–E.Anactualproductionpartthathasthistypeofgeometrywasexaminedinthepresentstudy.Thematerialusedforthisproductionpartwas0.7mmthick,andthestress–strainrelationobtainedfromtensiletests.Theprocedureinthepressshopfortheproductionofthisstampingpartconsistsofdeepdrawingfollowedbytrimming.Inthedeepdrawingprocess,nodrawbeadisemployedonthediesurfacetofacilitatethemetalflow.However,owingtothesmallpunchcornerradiusandcomplexgeometry,asplitoccurredatthetopedgeofthepunchandwrinkleswerefoundtooccuratthedrawwalloftheactualproductionpart,asshowninFig.6.ItisseenfromFig.6thatwrinklesaredistributedonthedrawwall,butaremoresevereatthecorneredgesofthestep,asmarkedbyA–DandB–EinFig.1(b).Themetalistornapartalongthewholetopedgeofthepunch,asshowninFig.6,toformasplit.Fig6Fig7Inordertoprovideafurtherunderstandingofthedeformationofthesheet-blankduringthestampingprocess,afinite-elementanalysiswasconducted.Thefinite-elementsimulationwasfirstperformedfortheoriginaldesign.ThesimulatedshapeofthepartisshownfromFig.7.ItisnotedfromFig.7thatthemeshatthetopedgeofthepartisstretchedsignificantly,andthatwrinklesaredistributedatthedrawwall,similartothoseobservedintheactualpart.Thesmallpunchradius,suchastheradiusalongtheedgeA–B,andtheradiusofthepunchcornerA,asmarkedinFig.1(b),areconsideredtobethemajorreasonsforthewallbreakage.However,accordingtotheresultsofthefinite-elementanalysis,splittingcanbeavoidedbyincreasingtheabove-mentionedradii.Thisconceptwasvalidatedbytheactualproductionpartmanufacturedwithlargercornerradii.Oneoftheadvantagesofusingfinite-elementanalysisforthestampingprocessisthatthedeformedshapeofthesheetblankcanbemonitoredthroughoutthestampingprocess,whichisnotpossibleintheactualproductionprocess.AcloselookatthemetalflowduringthestampingprocessrevealsthatthesheetblankisfirstdrawnintothediecavitybythepunchheadandthewrinklesarenotformeduntilthesheetblanktouchesthestepedgeD–EmarkedinFig.1(b).ThewrinkledshapeisshowninFig.8.Thisprovidesvaluableinformationforapossiblemodificationofdiedesign.Fig84.SummaryandConcludingRe