节点文献

扁挤压筒受力与变形分析及结构优化设计研究

Research on Stress and Deformation Analysis and Structure Optimized Design of Flat Receptacle

【作者】 李燕

【导师】 刘全坤;

【作者基本信息】 合肥工业大学 , 材料加工工程, 2005, 博士

【摘要】 铝合金整体壁板具有比强度高、耐蚀性和气密性好、造型美观等优点,在制造业中日益得到广泛的应用,其成形过程主要由用组合式扁挤压筒在大型挤压机上完成。扁挤压筒在高温、高压、高摩擦的恶劣条件下工作,经常发生开裂。由于扁挤压筒的受力和变形特征尚未被真正掌握,影响了模具的强度和使用寿命,因而限制了它的应用和推广。为了改善扁挤压筒的强度,延长其使用寿命,应对组合式扁挤压筒的受力和变形状况进行科学的分析,进而对扁挤压筒进行优化设计,才能充分发挥模具材料的潜力,确保扁挤压筒的最佳工作性能。 本文以扁挤压筒为研究对象,综合运用数值模拟和物理模拟方法,对壁板挤压成形规律和组合式扁挤压筒受力、变形状况进行了全面系统的分析,并结合神经网络和遗传算法思想,对扁挤压筒结构尺寸进行优化设计,取得了有重要意义的结论和对实际应用有指导作用的成果。 首先系统地介绍了刚(粘)塑性有限元法的基本理论,并以此作为扁挤压筒数值模拟分析的理论基础,对其实现过程中的若干关键技术问题进行了研究,提出了解决方法。 综合运用刚塑性和弹塑性分析方法研究扁挤压筒挤压成形壁板这一复杂问题,即先借助MSC.SuperForm软件平台分析成形过程,获得工作状态下挤压毛坯作用于扁挤压筒内壁的压力沿高度方向和沿型腔环线方向的分布情况;然后借助ANSYS软件平台将获得的真实内压力分布规律作为边界条件引入,用APDL语言编程,分析了三层组合扁挤压筒在热-预紧-非均匀内压力作用下的受力与变形状况。文中所做的热、预紧和非均匀内压力效应的综合分析,既保证了相互作用的关联性,又获得了全过程的完整解,使得数值模拟更客观地反映了扁挤压筒的受力状态,为后续组合式扁挤压筒的优化设计提供了必要的技术支持。 用ANSYS软件提供的接触单元法分别对采用圆形内层套和椭圆形内层套两种结构下的扁挤压筒(两层)在预紧装配时的内孔变形规律进行了研究,针对内孔型腔变形的不均匀性,提出了扁挤压筒结构改进方案——圆度设计法,即用椭圆形内层套代替圆形内层套,使内孔变形均匀化,从而提高扁挤压筒型腔的尺寸精度,简化后续的修模工作。 为了验证上述数值模拟结果的准确性,从两个方面开展了物理模拟工作。一是用光弹性实验法对扁挤压筒挤压成形壁板过程进行模拟,确定了挤压力的作用方式和扁挤压筒内部的应力分布;另一是用钢模比拟法对扁挤压筒预紧装配过程进行模拟,获得了挤压筒内孔型腔变形规律。实验结果与数值模拟结果一致,表明数值模拟结果正确可靠。用光弹性实验方法测定扁挤压筒内压力的作用方式,并以此作为有限元计算的边值条件,代替以往的均匀内压力作用的假设,提高了计算的精度,使挤压模具的优化设计更加科学合理。 最后将神经网络技术和遗传算法思想引入到扁挤压筒结构优化设计中,先按影响扁挤压筒受力的主要因素(各层直径、各层间过盈量及工作内压)设计正交试验,在Matlab7.0平台下训练一个BP神经网络,找出扁挤压筒结构尺寸、工作状态与内部各点处等效应力之间的映射关系,再结合遗传算法,按等强度设计思想优化模具结构尺寸,得出一套可行的智

【Abstract】 Monolithic aluminum alloy wallboard has got extensive use in manufacturing because of its advantages of high strength/density ratio, fine airtightness, unfailing performance and attractive appearance. It is mainly formed through extruding by combined flat receptacle, which is easy to crack under severe condition of high temperature, high pressure and high fricition. Nowadays, as a result of knowing the stress and deformation distribution of flat receptacle incompletely, poor intensity and short working life of the receptacle limit the application and generalization of this extruding process. In order to improve the strength and prolong the working life of the flat receptacle, we should analyze its stress and deformation distribution with accuracy, and then optimize its structural dimensions. It is advantageous to make full use of the material potential of the receptacle and make sure of getting its best working performance.Taking the combined flat receptacle as subject investigated, the extruding forming law of the wallboard and the stress and deformation distribution of the flat receptacle are analyzed completely with numerical simulation and physical simulation. Then through introducing the thoughtway of neural network (NN) and genetic algorithms (GA), the structural dimensions of three-layer combined flat receptacle are optimized. Some important and practical conclusions and achievements are obtained.Firstly, the basic theoretics of rigid-plastic finite element method (FEM) is given. Taking it as theoretical basis of numerical simulation for the flat receptacle, some key technologies during the implementation are discussed and the corresponding methods of resolution are put forward.Rigid-plastic and elastic-plastic FEM are applied into complex problem analysis of wallboard forming by flat receptacle. At first, by right of software MSC.SuperForm, wallboard extruding forming process is simulated and the inner pressure distribution of the receptacle acted by the blank is obtained along the axial and circumferential directions of the receptacle respectively. Then, by the aid of software ANSYS, introducing the forementioned inner pressure distribution law as one of the boundary conditions, the heat-structure coupling analysis of combined flat receptacle under the nonuniformly distributed pressure is realized by APDL language programming. So, the stress and deformation distribution of three-layer flat receptacle influenced by heat, interference fit and nonuniformly distributed pressure are acquired. The analysis involving all aspects not only guarantees the relativity, but also gains the full data of the whole process. Since all those influence factors are taken into account at the same time, the numerical simulation reflects the receptacle status objectively, which offers the necessary technique support to subsequent optimum design of flat receptacle structural dimension.Research on deformation law of mould cavity is done for two kinds of structures of combined flat receptacle with circular shaped lining and elliptic shaped lining respectively with contactelement method offered by ANSYS. Aiming at nonuniform deformation of mould cavity, a structure improvement project, roundness design method, is brought forward. That is, adopting flat receptacle with elliptic shaped lining instead of circular shaped lining uniformizes the deformation distribution, which will enhance the dimensional accuracy of the flat receptacle and simplify the workload of the mould repair in manufacturing.In order to verify the accuracy of the above numerical simulation, physical simulation is done from two aspects. On the one hand, simulate the wallboard extruding process to find the action mode of extrusion pressure and the distribution of inner stress of flat receptacle by photoelastic experiment. On the other hand, simulate the interference fitting process to get the deformation distribution of the mould cavity by steel mould correlation method. The experimental results accord with the numerical simulation results, which indicates that numerical simulation exactitude is satisfying. Furthermore, using the extrusion pressure distribution by photoelastic experiment as one of the boundary conditions in FEM analysis instead of foregoing uniform extrusion pressure hypothesis to solve the stress field of flat receptacle will enhance the computational accuracy, which also makes the optimum design of extrusion die more reasonable.Finally, the ideas of NN and GA are introduced into structural dimensions optimization for flat receptacle. Orthogonal experiment is designed according to the main influencing factors such as the diameters of each layer, shrink range and working pressure. Then, a BP neural net is trained on the flat of Matlab7. 0 to find the mapping relationship from structural dimensions and extrusion pressure to the maximum of equivalent stress in the receptacle. Afterwards, the structural dimensions of the receptacle are optimized using GA in terms of constant strength design philosophy. A feasible intelligent design method is attained, through which the reasonable diameter of each layer and the best value of each fitting allowance for the three-layer flat receptacle are decided. It offers a new approach to optimum design for complex nonlinear problems.The research results can guide the optimum design of flat receptacle. Furthermore, it improves the extruding die design theory for wallboard.

节点文献中: 

本文链接的文献网络图示:

本文的引文网络