论挤压时的成形缺陷

On Forming Defect During Extrusion

本文首次将主应力法(Slab Method)应用到金属挤压变形时的宏观缺陷预测中,提出了一种金属大变形时宏观缺陷预测的新方法。并以轴对称正挤压为例,给出了近似主应力法预测中心缩孔缺陷的具体应用。本文结果与实验结果比较表明:近似主应力法预测金属变形时的宏观缺陷不仅可以弥补已有文献分析结果之不足,而且在工程上实际应用时简便、易行。实验结果也证实了本文提出的新方法是正确的。

The upper-bound approach and the rigid-plastic finite element method maybe applied to predict the forming defect during extrusion[1,3-9], but the calcu-lating processes are very complicated, so much so that the expression for thecritical condition of defect formation is sometimes impossible to be written out[6,9]. The slab method is the simplest among al1 the calculation methods formetal forming [2]. But, to the authors best knowledge, it has never beenused to solve metal forming problems other than the forming load. The authors' contributions, of which the first is the most important, areas follows: 1. The slab method is creatively applied to predict the forming defect. Forthis purpose, the existence and location of the defect are taken into accountwhen the equations of equilibrium are established, and the minimum energyprinciple and the extremum principle are introduced into the slab method. 2. Creatively applying the slab method, the authors derive the criticalcondition for central cavity formation during forward-extrusion. 3. The process data for avoiding central cavity formation during forward-extrusion are calculated and presented. They are verified by the experiments inthis paper. Applying the slab method to predict the forming defect is simple, convenientand computer time-saving as compared with the other methods. And the relatio-nship among the deforming parameters may be displayed distinctly. Moreover, the results about defect formation calculated by the slab methodsometimes may be superior to the results calculated by the other methods in manyprevious papers[6,7]. For example. the calculated result in this paper for centralcavity during forward-extrusion, verified by the authors' experiments, shows thatthe formation of the cavity is not always aggravated as friction coefficient incr-eases, whereas it is not like this in the previous papers [6,7].