Simulation and formability analysis of drawing process for automotive B-pillar

Joumal o[ Harbin Institute of Technology (Ner Series), VoL. 16, No. 2, 2009Simulation and formability analysis of drawing processfor automotive B-pillarXING Zhong wen, FANG Hua-song, BAO Jun邢忠文,方华松，包军(School of Mechanical and Eletrical Engincering, Harbin Instiute of Technology , Harbin 150001 , China, gchxl@ hit. edu. cn)Abstract: The numenical simulation on drawing process of autombtive B-pillar was carried out on AutoFormsoftware, and dangerous forming areas were discovered. The processing parameters, such as the layout of draw-beads, blank holding force and the shape of blank, were adjusted and optimized according to the simulation re-sults. Results indicate that the quality defects can be forecast and removed, which improves the stability offorming process. The cost of design is decreased and the research cycle is shortened. It is proved that the draw-ing process and die design of B-pillar forming are feasible in actual production.Key words: draw forming; simulation; processing parameter; automotive B-pillarCLC number: TM92Document code: AArticle ID: 1005-91 13(2009 )02-0209-04The forming process of the automobile is a plasticsame time，some appended components, such as safetyforming process with big deflection and deformationbelt in the front row, should be fixed inside the B pil-which contains great geometry nonlinearity， touchingar. In order to improve the supporting capability, thenonlinearity and material nonlinearity. There are manyB pillar must have higher strength and stiffness. Fur-influencing factors and it is difficult to make accuratethermore the shape of the B pillar should be curved indecision. Therefore, for a long time the final resultsaccord with the bodywork.of the stamping planning and die design have dependedThe structural characteristic of the B pillar is thatgreatly on the try-and error method which wastes largethe depth of the forming is large and has great undula-amount of time，manpower and material resources. It istion, the change of the cross section is very complicat-difficult to satisfy the requirement of good trealing accu-ed, and the round angle radius is small. Those broughtracy day by day, and those products have no advanlagemuch trouble to forming technology and die design. Inon cost and quality, which can not catch up with theorder to satisfy the requests of strength and rigidity ofresponse to the changes in marketplace in time owing tothe B pillar， the material with higher yield strengthartificial operation. Surely the method will be replacedshould be used. The funcetion request and the structureThe appearance and development ocharacteristic for the B pillar make the control of qualitycomputer, nonlinearity theory and finite element meth-quite difficult. So the design and production of the Bod，especially the broadly spreading out of the study a-pillar become one of the difficult issue in the automo-bout the numerical simulation technology for the large-bile manufacturing.scale complicated stamping process have renewed themethod of stamping process and die design'i. Taking 2 Model Preparationthe B pillar as an example, the special software Auto-Form was used to simulate the forming process , and theFirstly, The B pillar model, which is created byreasonable processing parameters were obtained, whichuniversal CAD software such as Pro/E,UG is trans-can guide the production and shorten the designingformed to the igs format, and then imported into thetime.AutoForm software. The finite element model for thesimulation could be obtained after taking the steps as1 Characteristic and Function Request of B Pillarfollow..1 Check and Repair the ModelThe B pillar can also be called as middle postUsuallv. the geometrv model. which created bywhich locates beltween the front door and the backthed中国煤化工CAD software, candoor. As an important holding component, it supports not s_inite element meth-both the coping and the pressure from the doors. At theod cMCN M ! Grain som dsigureReceived 2007 - 10 -22.●209●Jounal of Harbin Insitute of Technology (New Series), Vol, 16, No.2, 2009ments, such as the overlaps of the curved surfaces,In the equation: σ, ,σz are the maximum and thegaps and some excessively slender curved surfaces. Sominimum main stress on the surface, the σ is equivalentmodel checking and repairing are needed to avoid thestress and r is the coefficient of anisotropy. Finally, thedefects. Besides, some trivial characteristics in thefinite element model obtained is shown in Fig. 1. .primitive design should be deleted lest that there aretoo many trivial elements in these areas, which will in-， PunchDrnawing Directioncrease the amount of unnecessary calculation works.2.2 The Determination of the Drawing DirectionRindeAs the three-dimensional product model of the Bpillar is created according to the installing position onthe bodywork and the direction disaccords to the stam-ping one in most cases. The direction, which the mod-el shows, should be adjusted. And as the dimension ofthe local shapes in the two ends of the B pillar is largeand there is a certain angle between them, the drawingdirection shown in Fig. 1 is adapted, which reducesthe drawing depth and availed the forming process.Fig. 1 Finite element model.3 AddendumThe design of the addendum is an important step3 Forming Simulation and Processing Optimizationduring the stamping processing design. Because the ad-dendum would influence the success or failure for the3.1 Forming Simulationforming, and also impact on the deformation conditionIn order to realize the sheet metal flowing in thefor the blank, the degree of deformation, the distribu-mold and forecast the difficult forming location rough-tion of the deformation, and some surface quality prob-ly, which is helpful to determine the layout of draw-lems , such as cracks, wrinkles and so on. The deforma-bead as well as optimize the forming technology andtion degree of B pillar at different places is quite differ-blank holding force, the first fast simulation is proent, and the distribution is also uneven. The addendumgressed using the One-Step module of the AutoFormis designed to adjust the magnitude of deformation at differ-software without the drawbead and the result is shownin Fig. 2(a). The outline of the die is used as the orig-ent places, which can improve the quality of the products.2.4 Material and Finite Element Modelinal shape of the blank, the friction cofficient is set toThe power-harden material model is used. Thethe value of 0.15 and the blank holding force is set tothickness of the blank is 1.6 mm, and the materialthe value of5 x 10° kN. Fig. 2(b) shows the approxi-performance is shown in Tab. 1.mate shape and dimension of the blank which is gainedat the base of the above result. Then the accurate sim-Tab.1 Material performanceulation is completed using the incremental module ofthe AutoForm software with the approximate shape ofE/CPa o,/MPap/(kg.m)blank obtained above and the result is shown in Fig. 3.200390.287.8x 10'0.83 0.135It is seen that, although both of the dimensions of thelocal shape at two ends and the plastic deformation areAccording to Hill Normal Anisotropy Yield Criterion:too large, obvious wrinkles are occurred on the edgesoσ +σz-i-σnσz =7.and in the comers as the insufficient drawing and therather small deformation.Minor strain中国煤化工Forming limit diagram.DHCNMH G(a) Forming simulationFig.2 Simulation results for the first time●210●Journal of Harbin Institute of Technology (New Series), Vol. 16, No.2, 2009by increasing the blank holding force and the resistance.4厂force of drawbeads4,5. So a process boss with a cer-Un:limit.2 ttain height is set up in the corner. Local resistance isincreased and the sheet flowing is improved and the de-0.8-fects are eliminated finally.As there is local bulge on the left side at the bot-tom of the part, the resistance of metal flowing is in-.4十creased, and the blank moves to right which causes the)2十insufficient of material on the left side. The phenome-non is more obvious especially after process bosses are30.8 -0.6-04-0.2 0 0.2 0.4added on the right and left corners. Partial blank is eMinor stninven out of control of drawbeads which is a great disad-Forming limit diugrumvantage for the deep drawing. In order to meet theFig. 3 Simulation results using incremental module withoutforming requirement, the shape of the blank is revisedthe layout of drawbeadagain. The shape of the blank after being changed canis shown in Fig. 4(b).3.2 Result Analysis and Processing AdjustmentDuring the period of complicated part stamping,the blank will break because of the insufficient strengthor the large deformation. According to the simulation,the reason for the breakage of the straighten border areaof the part is that the drawing depth is too large. In or-der to make the material move better and overcome thebreakage, the addendum face is adjusted, and thedrawing depth is reduced.IDuring the forming process of the automobile coverpanels, the drawbeads are needed to control the behav-(a)The layout of the( b)The ulimate shapeior of the material so as to prevent from wrinkling anddrawbeads alter revisedof the blankimprove the sifness. In finite element method simula-Fig.4 The layout of the drawbeads and the shape of blanktion, there are two ways to handle the drawbeads: 1)regarding the drawbeads as one part of the whole finiteThe ultimate simulating result ( as shown in Fig.element model, then the grids are divided, and the5) can be obtained by optimized processing parame-calculating and fllwing grid dividing are based on it;ters, such as adding drawbeads and changed blank2) using equivalent drawbeads, simplifying the draw-holding force, and the shape of the blank. It can bebeads as the force focusing on the node and not partici-seen in Fig. 6 that the dangerous area has been elimi-pating the grid dividingFor saving the CPUnated basically, and the qualifred part is obtained.time, the paper adopts the second way, using equiva-lent drawing model to replace the true drawbeads.Drawbeads were established on the blank holderalong the trimming line, the distance from which is 201.4-30 mm. And the resistance curve is uniformly in the1form of semi-circle drawbead, the height of which is 51.0mm. For easy to adjust the processing parameter ac-0.8cording to the different parts and different input speed,the drawbead is established part by part. The layout of0.6the drawbeads is shown in Fig. 4( a), and the blankholding force is adjusted. In the simulation result afterthe drawbeads are set up and the blank holding force isadjusted, it can be seen that the wrinkles of the middle-0.6 -0.4 -0.2 0 0.2 0.4part is well controlled, and the shape quality of theorming limit diagrarmpart is improved very well. But there are still obviouswrinkles in the cormer with a larger size, and the thick-.FYHC N M H Gsing pesingness subtraction of the straighten border area is very se-rious. It is not possible to improve the forming quality●211●Journal of Harbin Institue of Technology (New Series), Vol. 16, No.2, 2009[4]Zhao Xia, Fu Jian, Yu Ling. Application of numerical sim-ulation in automobile covering parts forming process. Forg-ing and Stamping Technology. 2006 (1):15- 18. .[5] Tisza M. Numerical modlelling and simulation in sheet met-al forming. Joumnal of Materials Processing Technology ,2004 (151): 58- 62.[6]Shi Lei, Wang Yongzhi, Yang Yuying. Numerical simula-tion of wing part drawing process. Materials Science andTechnology, 2004(2): 29-31.[7]Zhou Jie, Yang Desen, Luo Zhengzhi. Finite element anal-ysis and optimization of blank for fender groupware. Forg-ing and Samping Technology，2006(1):9- I2.[8] Silva M B, Baptista R MSO, MartinsP A F. Stamping ofFig.6 The utimate qualifed partautomotive components a numerical and experimental in-vestigation. Jourmal of Materials Processing Technology,2004 (155 -156): 1489 - 1496.4 Conclusions[9] Shi Xiaoxiang, Chen Jun, Peng Yinghong, et al. A newapproach of die shape optimization for sheet metal formingThe finite element model of the simulation of the Bprocesses. Joumal of Materials Processing Technology,pillar forming is obtained after the checking of model，2004(152):35 -42.the determination of the drawing direction, the adden-[ l0]Ohata T, Katayama T, Nakamachi E, et al. Improvementdum and the setting up of material model. According toof optimum process design system by numerical simulationthe result, the simulation can forecast the forming de-- discretized optimization method. Proceedings of theNumisheet'99. 1999. 299 - 304. .fects which may appear in the whole process very well.By adjusting the processing parameters and the blank[11] Yu Haiyan, Chen Guanlong, Zhang Weigang. Develop-ment of drawbead investigation in sheet metal forming tech-shape of blank，qualified part is obtained and the qual-nology. Joumal of Plasticity Engineering, 2004(3):77 -ity of the processing design is improved while the costof design decreases and the research cycle is short-. [12 ]Courvoisier L, Mariny M, Ferron C. Analytical modelingened.of drawbeads in sheet metal forming. Journal of MaterialsProcessing Technology, 2003(133): 359 - 370.References:[13 ]Samuel M. Infuence of drawbead geometry on sheet metal[1] Li Shuoben. Stamping Processing Technology. Beijing:forming. Joumnal of Materials Processing Technology, 2002Machinery Industry Press, 1982.(122): 94- 103.[2] Ffang Gang, Lei Liping, Zen Pan. Design of the dies for the[14]Guo Huaying, Jin Miao. Study of sheet deforming in theauto body metal sheet parts based on fnite elemenlanalys-semi- eircle drawbead. Forging and Stamping Technology ,is. Molding Industry, 2004(6): 6- 10.2006(6) :29 -32.[3] ParkCD, Chung WJ, Kime B M. A numerieal and exper-[15 ]Liu Yanfang, Shi Fazhong, Feng Tianfei. Flaborate imple-imental study of surface deflections in automobile exteriortion of an equivalent drawbead model in the numeri-cal simulation of sheet melal forming. Chinese Joumal ofpanels. Journal of Materials Processing Technology, 2007(187- 188): 99- 102.Mechanical Engineering, 2005(1):114-118. .中国煤化工MYHCNMHG●212●