基于多目标进化算法的模锻工艺优化及模具设计

发布时间：2018-05-14 11:06

本文选题：精密模锻 + 多目标优化　；参考：《江苏大学》2017年硕士论文

【摘要】：精密模锻是一种少无切削的塑性加工工艺。鉴于模锻件优异的机械性能和较高的材料利用率,在汽车、航空航天等领域均获得了广泛应用。本课题以一种异形汽车传动件作为研究对象,其结构复杂、尺寸精度要求高,模锻成形过程中容易出现充不足、模具磨损严重及表面质量差等问题。在实际模锻工艺设计中,预成形几何形状参数对终锻零件的质量有较大程度的影响。因此本文采用将数值模拟、多目标优化、多属性决策等理论相结合的方法对异形传动件的预成形形状参数进行了优化并针对优化结果进行了生产验证。首先,针对异形传动件的模锻成形工艺,设计了二道次成形方案与三道次成形方案。利用有限元软件Forge模拟了两种不同方案的成形过程并分析了其优劣之处。根据分析结果,认为三道次方案充型完整、模具应力及模具磨损亦更小。在确定三道次方案后,为研究预成形形状对终锻件质量的影响,选择了预成形的三个关键几何尺寸作为因素,以填充性能、模具磨损为目标设计了L_(16)(4~3)的正交试验并进行了数值模拟。其次,以数值模拟获得的数据作为样本集,采用支持向量机进行学习、泛化,进而建立了预测模型。在与BP神经网络的学习结果对比后,可以发现支持向量机在小样本、存在噪声的条件下依然具有较高的预测精度,预测模型相关系数均在85%以上。随后根据预测模型建立显式方程作为多目标进化算法的目标函数,进而通过Matlab编制寻优程序,获得了Pareto最优解集。在非劣解集内寻找满足决策偏好的工程应用解的过程属于多属性决策范畴,本文选择采用优劣解距离法(TOPSIS)进行决策,最终获得符合工程实际需求的最优应用解。经有限元与实际生产验证发现,与优化前的方案相比,模具磨损量降低约52.8%,充型亦较为饱满。最后,为进一步提高产品关键部位的尺寸精度及其表面质量,设计了一种背压式冷精整工艺并从理论上对其整形过程中的应力-应变状态进行了分析。分析结果认为,施加背压力能够使变形金属进入三向压应力状态并抑制金属的反向流动,从而避免了内腔边缘的坍塌、折叠等缺陷的产生。本课题将异形传动件的精密模锻工艺、有限元分析理论、多目标演化寻优方法相结合,达到了改进成形工艺、提高模具服役寿命、缩短设计周期的目的。研究成果为异形传动件的量产提供了可靠的理论指导,并为将来此类相似锻件的模锻工艺开发提供了理论与实践依据。
[Abstract]:Precision die forging is a less cutting plastic processing technology. Because of its excellent mechanical properties and high material utilization, die forging parts have been widely used in automotive, aerospace and other fields. In this paper, a kind of special-shaped automobile transmission is studied. Its structure is complex, the dimension precision is high, the filling is easy to appear in the process of die forging, the die wear is serious and the surface quality is poor. In the actual die forging process design, the geometric shape parameters of preform have great influence on the quality of the finished forging parts. In this paper, numerical simulation, multi-objective optimization and multi-attribute decision are used to optimize the shape parameters of special-shaped transmission and the results are verified. Firstly, two pass forming scheme and three pass forming scheme are designed for the die forging process of special-shaped transmission parts. Finite element software Forge is used to simulate the forming process of two different schemes and its advantages and disadvantages are analyzed. According to the analysis results, it is considered that the three pass scheme is complete, and the stress and wear of die are smaller. In order to study the effect of pre-formed shape on the quality of final forging, three key geometric dimensions of pre-forming were selected as the factors to determine the three-pass scheme. The orthogonal test and numerical simulation for die wear were carried out. Secondly, using the data obtained by numerical simulation as sample set, support vector machine (SVM) is used to study and generalize, and then a prediction model is established. Compared with the learning results of BP neural network, it can be found that support vector machine still has high prediction accuracy under the condition of small sample and noise, and the correlation coefficient of prediction model is above 85%. Then the explicit equation is established as the objective function of the multi-objective evolutionary algorithm according to the prediction model, and the optimal solution set of Pareto is obtained by programming the optimization program by Matlab. The process of finding the engineering application solution which satisfies the decision preference in the non-inferior solution set belongs to the category of multi-attribute decision making. In this paper, we choose to choose the method of superior and inferior solution distance to make the decision, and finally obtain the optimal application solution which accords with the actual engineering demand. It is found that compared with the optimized scheme, the wear capacity of the die is reduced by 52.8%, and the mold filling is relatively full by the finite element method and the actual production verification. Finally, in order to further improve the dimensional accuracy and surface quality of the key parts of the product, a backpressure cold finishing process is designed and the stress-strain state in the shaping process is analyzed theoretically. The results show that applying back pressure can make the deformed metal enter into the state of triaxial compressive stress and restrain the reverse flow of the metal, thus avoiding the collapse of the inner cavity edge, folding and other defects. In this paper, the precision die forging technology, the theory of finite element analysis and the multi-objective evolutionary optimization method are combined to improve the forming process, prolong the service life of die and shorten the design period. The research results provide reliable theoretical guidance for mass production of special-shaped transmission parts, and provide theoretical and practical basis for the development of die forging process of this kind of similar forgings in the future.
【学位授予单位】：江苏大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TG31

【参考文献】