节点文献
汽车覆盖件冲压成形CAE技术及其工业应用研究
【作者】 雷正保；
【作者基本信息】 中南大学 ， 车辆工程， 2003， 博士
【摘要】 本文结合上汽通用五菱汽车股份有限公司新产品N1的研制,开展了汽车覆盖件冲压成形CAE技术及其工业应用研究,所得主要结论如下:⑴最优模具间隙值为坯料厚度的15%;⑵坯料形状的局部改善不会导致冲压结果明显改善;⑶坯料与凹模之间的摩擦系数变化对冲压结果的影响程度,远不如坯料与凸模及压边圈之间的摩擦系数变化所带来的影响大(即所谓的贴模性将不利于复杂冲压件的冲压成形);⑷如不进行回弹分析,选用法向方向3点积分的Belytschko-Tsay壳单元公式就足够了;如冲压分析后还需进行回弹分析,则宜采用法向方向5点积分的单元公式,其中又以全阶积分壳单元公式最理想;⑸对于冲压条件较恶劣,冲压件可能或即将产生开裂而开裂程度不是特别严重的情况,采用3参数BARLET材料模型能够获得可靠的结果,且在计算费用上与采用厚向异性材料模型时没有明显差异;⑹只有对那些拉延筋以长直线形状布置为主的冲压情况,才可以力函数模拟拉延筋的作用,否则,宜采用网格模型来模拟拉延筋;⑺采用拉延筋网格模型将大幅增加单元数目,但增加的计算费用却很低,且此时,虽然在拉延筋处的坯料壳单元存在扭曲严重的问题,对计算结果会带来一定的影响,但采用力函数模型带来的计算误差更大,因此,对复杂冲压件,建议不要采用力函数模型。⑻模具上单元尺寸的大小并不参与确定由Courant稳定性条件决定的临界时间步长,模具网格的划分只要能够合适模拟模具形面的形状即可;⑼如果其它冲压条件不太理想,那么仅仅依靠压边力的调整是难以获得理想结果的;⑽坯料不采用自适应技术,不仅影响接触界面力的传递,影响计算结果的可靠性,而且会大幅度增加计算成本;⑾采用自适应技术时,坯料初始网格形状对计算结果有很大的影响,这一点与模具上的网格划分情况是完全不同的;⑿对复杂形状的坯料,通用网格划分软件不能获得可用的坯料网格,此时将出现模具一与坯料接触坯料上就开裂的错误现象,而且,自适应网格技术将不能得到正确应用,会出现计算中途“死机”现象。此时,采用裁剪网格能够保证自适应技术的正常工作,且可获得可靠的计算结果。⒀冲压与回弹分析重要参数的选择合适与否,不仅影响计算结果的可靠性,而且涉及到计算过程能否顺利进行,回弹计算更是如此。
【Abstract】 Using the opportunity of the new product N1 in the SAIC GM Wu Ling Automotive LTD. Corporation,the CAE technique and its industry application researches of sheet metal forming for automotive body panels have been investigated in this paper, the main conclusions are as follows:1 The optimal gap value between the die and the punch is 15% of the thickness for the blank.2 The stamping results can’t be improved evidently by the partial improvement of the blank shape.3 The influence of the stamping results from the variety of friction coefficient between the blank and the die is much less than that from between blank and punch or between blank and binder. ( i.e., the so-called stick mold property won’t benefit for stamping for the forming of complicated automotive body panels)4 If springback analysis is not necessary, the Belytschko- Tsay shell formula with 3-ponts integral in normal direction can be used. Otherwise, the shell formula with 5-ponts integral in normal direction is preference, and the full rank integral shell formula is the best selection.5 For the stamping condition is worse, the blank possible to or be about to be cracked but the crack degree is not special serious, using the Barlat’s 3-parameter plasticity material model can acquire the dependable results, and the compute expenses have no obvious difference compared with that of using the transversely anisotropic elastic-plastic material model.6 Only the forming situation that drawbeads arrange mainly with long straight line, the force-function can be used to simulate the function of drawbeads correctly, otherwise, the finite element mesh model of drawbeads is preference.7 Using the finite element mesh model of drawbeads, the number of elements will increase significantly, but the augmentative calculation expense is very low. At the same time, although there are serious distortion problem in blank elements under the drawbeads that will increase the error of the calculation results, the calculation error from adopting force-function model of drawbeads is much more. Therefore, it is suggest that we shouldn’t adopt the force-function model in the CAE for the forming of complicated automotive body panels.8 Owing to the element size in the tool do not participate in deciding the critical time-step that decided from the Courant stability condition, so that the demand for the mesh of the tool is only need to fit to imitate the shape of the tool surface.