节点文献

枝桠类零件多向加载无飞边成形理论与技术研究

Study on Forming Without Flash in the Multidirectional Loading Theory and Technology for Thepart with Branch

【作者】 于建民

【导师】 张治民;

【作者基本信息】 中北大学 , 机械设计及理论, 2012, 博士

【摘要】 多向主动加载对于枝桠类零件是一种理想的成形方法。相对于其它的成形工艺,该工艺通过控制不同加载路径,改变应力状态,提高金属的流动性,并降低成形力,能有效提高产品力学性能和材料利用率。该工艺最大的优点是只需一火即可完成传统成形需多道完成的工序,实现复杂零件的整体成形,生产效率高。但是,多向主动加载的成形最主要的缺点是:由于在主动凹模的分模面上严重出现飞边,导致不能合模,后续的动作无法继续,致使该工艺的推广遇到了瓶颈。本文研究在主动加载过程中如何实现无飞边成形的理论与方法。多向主动加载过程中飞边产生的原因主要是由于在成形过程中,具有模具不封闭的自由流动空间。据此创新地提出一种锯齿形曲面分模的主动加载凹模模具结构,并建立了多向加载无飞边控制成形原理:通过在动模的分模面上采用锯齿形结构,使坯料变形后的鼓形部位的金属形成飞边的趋势得以控制,随着齿形的啮合,使得形成飞边的流动距离延长,不封闭的自由空间减少,从而实现无飞边控制成形。结果表明:采用该方法设计的模具,有效地解决了在多向主动加载过程中产生飞边,导致成形不能继续的难题,为多向主动加载的成形工艺的进一步研究与推广提供重要参考。通过对新型齿形模具成形过程中的变形体进行变形分区,获得主动加载条件下金属的流动规律。在分析计算各分区速度分量的理论基础上,采用中心复合试验设计得出枝桠类零件多向加载时产生飞边量(目标函数)与齿形长度b,齿形厚度h和成形枝桠型腔面积s(设计变量)的回归模型。通过方差分析对该模型的拟合精度检验,并对模型进行了修正。分析方差分析中的P值获得影响飞边产生的主效应的主次顺序。利用响应曲面显示出各主效应的交互作用对于产生飞边的影响。论文针对某型典型铝合金枝桠零件首次利用响应面法和分式析因设计(fractionalfactorial design)的筛选方法对多向主动加载模具结构参数进行研究,以成形件飞边质量和主动加载成形载荷为目标函数,选取齿形曲面分模结构参数为设计变量,进行模具结构参数的优化设计,得到飞边及载荷与各模具结构参数的回归模型。借助于该模型采用半正态概率图显示,有效筛选出多向主动加载模具结构设计中对飞边量和载荷影响效应显著的设计变量及其交互作用。论文还对多向主动加载过程中的工艺参数进行优化,深入研究不同工艺参数在成形过程中对不同状态变量的影响。采用BOX-Behnken实验设计对多向主动加载时的成形工艺参数设计试验方案并进行优化,得到了该典型枝桠零件最佳的成形速度、成形温度和摩擦系数的组合及最优化的目标函数值(产生飞边量和成形载荷)。研究了典型难变形材料的特性,在自主开发的12.5MN-3.15MN-3.15MN多向压机上进行了实验验证。结果显示:采用最优化的参数组合,成形过程中产生的飞边量极少。各部位的金相组织和最终的力学性能显示:所成形零件的各部位组织晶粒尺寸差别不大,变形相对均匀,组织细化,具有优良的力学性能。本文的主要创新点有:(1)提出一种锯齿形曲面分模的主动加载凹模,实现多向主动加载过程无飞边成形。(2)用响应面法对成形具有枝桠几何特征零件的主动加载模具参数进行优化,获得了目标函数(飞边量)与设计变量(齿形长度b,齿形厚度h和成形枝桠型腔面积s)的回归模型。(3)首次在多向加载模具的结构设计中使用分式析因设计的筛选方法,在多个设计变量中筛选出影响目标函数飞边量和成形载荷的显著因素,并基于Box-Behnken的试验设计对筛选出的影响效应显著的因素进行优化,为多向主动加载凹模的设计提供科学的依据。

【Abstract】 Multidirectional active load is an ideal forming method for branch-type parts. Comparedto other forming processes, this technology can effectively improve the mechanical propertiesof products and material utilization by controlling the different loading paths, changed stressstate, metal fluidity and decreased forming load. The main advantage is complete themulti-pass operation simply by one step, through which achieve the whole forming of thecomplex parts. However, the main disadvantage is lots of flash appear on the parting surfacein multidirectional active load, which resulted in mold cannot close, follow-up actions cannotcontinue. This cause the promoting of process meets the bottleneck.The paper research theory and methods for realize no flash forming in the active loadingprocess. Study found: the flash occurred in the active loading forming process is attributing tothe unclosed space on the mold, through which the metal can free flow. Based on thecondition, creatively design an active loading mold structure with tooth profile parting, andestablish without flash control forming principle for multi-directional loading. The resultsshowed that: the new mold effective resolves the problem that the appearance of flash duringactive loading process. It provides an important reference for further research and promotionof active load forming process.Through deformation partitioning of the deformable body, tooth profile mold, obtains themetal flow rule in the conditions of active loading. Based on calculates and analysis velocitycomponent, the central compound experiment design has been used for the regression modelof flash quantity (objective function) and tooth profile length, tooth profile thickness andcavity area (design variables). Check for fitting accuracy of the model by the varianceanalysis, and revise the model. Through analysis P value in the analysis variance obtains theimportant order of main effect which influence flash produces. Finally, using the responsecurved surface demonstrated the interaction of each main effect.In view of typical aluminum branch parts, research the structure parameter ofmulti-directional active loaded die by response surface model and fractional factorial designscreening methods. Select the flash quality and active forming loading as load objectivefunction, tooth shaped surface die structure parameter as design variable, optimized design thedie structure parameter. By this, the regression model of the flash quality loading andparameters of mold structure were obtained. With the help of the model with half-normalprobability diagram, design variables and its interaction which significant affect the flashquality and loading were screened.On this basis, the process parameters during multidirectional loading process wereoptimized, study the effects of different process parameters to state variables during theforming process. Using Box-Behnken experimental design to design programme and optimize the formingprocess parameters, obtain the optimal parameters combination(forming speed, temperatureand friction coefficient) and objective function value.Finally, through experimental verification on self-development 12.5MN-3.15MN-3.15MN press,the result show that based on the optimized parameter combination,theflash are almost no produced in the forming process. The difference of crystal grain size ofmicrostructure is small,deformation is relatively uniform, has the better mechanicalproperties.The innovation is: (1) a tooth shaped die has been designed, through which achieve noflash forming.(2) Using response surface method optimized the die structure of parameters foractive load, obtained the regression model of objective function (Flash) and design variables(length b of tooth, tooth thickness and branch cavity area) .(3) The first application offractional factorial design in the multi-directional loading mold structure design, screening thesignificant factor from numbers of design variables, based on box-behnken experimentaldesign, optimized the significant effect factors and provide scientific basis to the die design ofactive load.

  • 【网络出版投稿人】 中北大学
  • 【网络出版年期】2012年 08期
节点文献中: