【作者】 管婧；

【导师】 赵国群；

【作者基本信息】 山东大学 ， 材料加工工程， 2008， 博士

【摘要】 金属塑性加工是国民经济的基础制造产业,被广泛应用于许多工业生产部门,在国民经济中占有极为重要的地位。随着全球经济竞争的日趋激烈,低成本、高质量和高效率成为制造业企业在竞争中取胜的关键因素,也成为制造业企业追求的目标。这种市场竞争和产品的快速更新换代,对产品开发周期与成本及产品质量的要求日益提高,对精密成形工艺和模具设计水平的要求越来越高。计算机与信息技术的发展,使现代制造业正经历着一场深刻的变革-仿真制造/虚拟制造-制造业信息化。这些新特点、新要求都要有相应的数字化和科学化的塑性成形工艺和模具设计方法及其软件的支撑。现有的常规经验设计方法已经无法适应这种状况,需要研究和开发基于数值模拟技术和计算机技术的模拟优化设计系统,以不断满足制造工艺和工装设计数字化的要求。随着计算机软硬件技术的发展及金属塑性流动理论的成熟,使得塑性成形过程的计算机数值模拟得以实现。利用数值模拟方法,可以实时跟踪描述金属的流动行为,模拟成形过程,揭示金属的流动规律,研究各种因素对金属变形行为的作用与影响,方便地确定成形过程各个阶段所需的变形功和载荷,获得工件及模具的内部应力、应变、温度分布和金属流动规律,预测工件的成形状况、残余应力、缺陷、晶粒度和晶粒取向分布等,为金属塑性成形工艺和模具设计提供了强有力的工具。材料在锻造过程中的微观组织演变及最终产品的组织结构分布直接决定着产品机械性能,因此,锻造过程微观组织模拟与优化的研究是锻造工艺优化设计中必须且非常关键的内容,具有重要的应用价值。预成形模具的形状与终成形件形状直接对应,它直接限制金属的流动情况,从而直接影响成形件的最终形状和微观组织性能。在工艺设计中,预成形设计对产品的制造成本、质量和生产效率的影响最为直接,也最为显著。因此,预成形设计便成为控制产品质量与性能、实现塑性加工生产要求的必需且非常重要的方面,而预成形的优化设计也就成为锻造工艺优化设计中最为关键也最有成效的一个方面。近年来,形状优化和各种优化理论从结构优化领域被引入到锻造过程的工艺和模具设计中。基于灵敏度分析的工艺和模具设计方法,是将工艺和模具设计问题处理为优化问题,在有限元正向模拟的基础上,结合优化算法来实现对工艺和模具形状的优化设计。它充分利用了有限元模拟过程中提供的信息来实现对锻造过程设计的自动化,是一种较新颖的设计方法。本文针对锻造过程数值模拟与优化技术及其系统开发开展研究,研究和建立基于灵敏度分析以及遗传算法的锻造过程多目标模具优化设计方法。本文以刚/粘塑性有限元为理论分析根据,采用基于灵敏度分析和遗传算法的正向模拟法,以预成形模具型腔形状为优化对象,用B样条曲线表示预成形模具的形状,B样条曲线控制点的坐标作为优化设计变量,以获得包括锻件形状和锻件微观组织分布在内的综合质量较好的终锻件为目标,建立了同时以锻件微观组织和锻件形状为优化目标的多目标优化方法,对锻造成形过程进行了多目标优化设计的研究。本文对金属体积成形中的锻造过程优化设计建模问题进行了研究。锻造过程中的优化设计建模包括选取合适的设计变量,建立目标函数的数学模型,选择合适的优化方法求解数学模型。本文详细地给出了锻造过程中各种优化目标的数学模型。阐述了锻造过程中优化设计变量的选取原则,比较了预成形件形状和预成形模具形状作为设计变量的区别,给出用三次B样条函数表示预成形模具形状的方法。比较了用于锻造过程优化设计方法的优缺点。本文确立的锻造过程优化设计建模方法也同样适用于其它体积成形过程。单目标优化是多目标优化的基础。在进行多目标优化之前首先进行单目标优化设计的研究。锻件精确的形状是锻件质量的首要指标。本文研究了基于有限元的灵敏度分析法,以获得少无飞边的锻件为优化设计目标,以锻造过程预成形模具型腔形状为优化对象,对非等温成形过程进行了灵敏度分析。以实际锻件与理想锻件之间不重合部分的面积(体积)作为优化设计的目标函数,建立了目标函数的表达形式。用三次B样条函数表示模具形状,B样条函数的控制点坐标作为设计变量。推导了目标函数相对于优化设计变量的灵敏度。详细推导了节点速度灵敏度方程,给出速度灵敏度边界条件;给出流动应力灵敏度计算公式,详细推导了节点温度灵敏度方程。锻件内部微观组织的分布直接影响着锻件的机械性能,因此,对锻件内部微观组织的控制和优化是锻造工艺优化设计的重要内容,也是提高产品质量的重要途径。本文应用现有的组织分布模型,采用有限元灵敏度分析的方法,对锻造过程中的微观组织进行了模拟和优化设计。以变形结束时的晶粒尺寸分布均匀性为目标,采用每个单元的平均晶粒尺寸与整个锻件的平均晶粒尺寸之差的平方和作为目标函数,对锻造过程进行了优化设计。详细推导了目标函数对设计变量的灵敏度及晶粒尺寸对设计变量的灵敏度公式。建立了锻造成形过程微观组织分布变化的模拟系统,给出了微观组织的优化设计步骤。为获得形状和微观组织分布均匀综合性能较好的终锻件,对锻造成形过程进行了多目标优化设计的研究。建立了适合于组成总目标函数的新的形状子目标函数和微观组织分布均匀性子目标函数,给出了其详细的数学表达形式。采用线性加权的方式将两个子目标函数组合成总目标函数,对总目标函数进行优化。推导了总目标函数相对于优化设计变量的灵敏度。给出了不同情况下加权因子的选择方法。开发了锻造预成形模具优化设计的程序,并对截面为H型的轴对称锻件的预成形模具形状进行了优化设计。给出了优化迭代中各目标函数的变化曲线。由预成形件形状和终锻件形状及其微观组织分布的比较,可以看出优化达到了同时对锻件形状和微观组织分布均匀性进行优化的目的。说明本文提出的基于有限元灵敏度分析的锻造过程微观组织优化理论的可行性。遗传算法由于思想简单易于实现,赢得了许多应用领域。本文综合微观组织与形状对锻件质量的影响,建立了同时以锻件微观组织和锻件形状为优化目标的基于遗传算法的多目标优化设计方法。重构了形状子目标函数和晶粒尺寸子目标函数,并采用加权和的方式组成总目标函数,对总目标函数进行优化。用B样条曲线表示预成形模具形状,应用所开发的基于预成形的锻造过程微观组织模拟与优化软件,对截面为H型轴对称锻件进行了面向微观组织优化的预成形优化设计。并对圆柱体镦粗进行了基于宏观工艺参数的优化。同时,对毛坯尺寸和预成形压下量对成形后微观组织的影响进行了研究,给出了优化结果。实现了提高锻件晶粒尺寸均匀细小性的目的,取得了较好的优化效果。以截面H型轴对称锻件为例,确定了锻造过程微观组织模拟与优化程序的实验验证方案,进行了相应的模具设计和实验研究,为计算机模拟优化结果的可靠性分析提供了实验依据。通过实验结果与计算机模拟优化结果的比较分析表明,计算机模拟与优化结果与实验结果具有较好的一致性。证实了本文的基于微观组织优化和预成形形状优化的多目标优化理论与系统在锻造成形优化过程中的应用具有较好的稳定性和可靠性。更多还原

【Abstract】 Metal plastic processing is the fundamental manufacturing industry for national economy. For the reason of intense market competition and rapid updating of the products, the requirements of development period, cost and quality for products are increasingly improved. The design level of processes and dies are more recognized by industry, especially for the precision forming. With the development of computer and info-technology, the modern manufacturing industries are undergoing a profound revolution about simulation, virtual manufacturing and computerized manufacturing. The new characteristic requires digitized and scientific method and according software for the plastic forming processes and dies design. The traditional and empirical method of the design could not satisfy the requirement. The study and development on the numerical simulation and optimization technique are needed to satisfy the digitized requirement of the manufacturing processes and designs.With the development of the computer technology and the reliability of metal forming theory, the numerical simulation of forging processes can be realized on the computer. With the method of numerical simulation, many information of the metal forging such as metal flow, stress or strain and so on can be obtained, the load and energy needed in plastic forming process can be decided, the distribution of the stress, strain and temperature can be obtained in both workpiece and dies, and the forming status of the workpiece, residual strain, defect, grain size and orientation of the grain in the workpiece can be predicted. The simulation gives a powerful tool for the processes and die designs.The evolvement and distribution of the microstructure in the forging process have a direct influence on the mechanical properties of the final parts. So, among the various parameters of the forging process, the simulation and optimization of the microstructure are pivotal and integrant. The preform shape is corresponding to the shape of the final parts. It directly controls the reliability of metal. And it directly influences the shape of the final part and the capability of the microstructure. So the preform design becomes the crucial aspect of controlling the quality of the product. Thus, the optimal perform design is the most efficient way in the optimization design during forging processes.In recent years, the theory of shape optimization and sensitivity analysis has been introduced into processes and die designs of bulk forming process from the fields of the structure optimization. The problem of the design is regarded as the problem of the optimization for the processes and die designs based sensitivity analysis. It is on the basis of forward FEM simulation, combined with optimal algorithm to realize the optimal design for the processes and dies. It fully utilizes the information provided by the FEM simulation and can automatically realize the optimal designs. So, it is a new method for the processes and die designs.The research and development on the numerical simulation and optimization technique for bulk forming is carried out. The optimal methods based on the sensitivity analysis and genetic algorithm for the die designs are researched and constructed. Based on the above analysis, the multiple objective preform optimal design of the metal forging is studied in this dissertation. The research work is carried based on the visco-rigid plastic finite element method, the sensitivity analysis and the genetic algorithm. The optimal object is the preform die shape of the metal forging. The preform die shape is represented by the cubic B-spline. Therefore, the coordinates of the control points of the B-spline are used as the optimal design variables. The optimal objective is to obtain such a final forging that integrity quality involved the shape of forging and the uniformity of the microstructure of the forging is high. Based on the above analysis, an optimization design method of the shape and microstructure in forging process is proposed in this dissertation. And multi-objective optimization designs for the forging processes are carried out using above two methods.The establishment of optimal model in forging process is investigated, which includes choosing proper design variables, constructing mathematical model of objective function and selecting proper optimization methods to solve the mathematical model. Mathematical models of various optimal objectives in forging process are given in detail. The selection principle of design variables in forging process is specified. The preform forging shapes and preform die shapes, as design variables, are compared. The method of using cubic B-spline to represent the preform die shapes is specified. The optimization methods used in optimal design for forging process are also compared.The single objective optimization is the foundation of the multiple-objective optimization. Therefore, the research of the single objective optimization prior to the multiple-objective optimization is very important in the optimization. The precision shape of the forging is the basic requirement of the forging’s quality. So, the optimization aimed to obtain the precision shape of the forging, net shape or near net forging, is carried out first. The different areas between the desired shape and the actual shape of the forging are used as the optimal objective. Then, the objective function is given. The sensitivity of the objective function respect to the design variables is developed. The sensitivity of the nodal coordinate and the nodal velocity with respect to the design variables are developed too. The velocity sensitivity boundary conditions are specified. The formulations of flow stress sensitivity are given. The equations of nodal temperature sensitivity are deduced in detail.The distribution of the microstructure in the forging has direct influence on the mechanical properties of the forging. So, it is important to control and optimize the microstructure to improve the quality of the forgings. Based on known model of microstructure evolution and FEM sensitivity analysis method, the simulation and optimization to microstructure in forging process are carried out. The goal is to achieve uniform recrystallized grain size after deformation. The least-square deviation between the average grain size and the actual grain size is used as the objective function. The optimal design is carried out for the forging process. The sensitivity equations of the objective function and grain size with respect to the design variables are deduced in detail. The procedure of the optimal microstructure design in forging process is given. The simulation system for microstructure distribution is established.The optimal preform design aimed to obtain the net shape or near net shape and more uniformity microstructure distribution of forging at the same time is carried out in this dissertation. The objective function for the multiple-objective optimization is put forward. The sensitivity of the objective function with respect to the design variables is developed. The sensitivity of the shape sub-objective and the microstructure distribution sub-objective with respect to the design variables is given. The selection of the weighting factor is presented. The software for multiple objective optimization of the preform design of the metal forging is developed. An example to optimizing the preform design of an H-shaped forging in axisymmetric deformation is done. The total objective, the shape design sub-objective and the microstructure distribution sub-objective versus the optimization iteration number is given. The different shapes of the preform dies and the final forgings at the various optimization iteration steps are given. The different microstructure distributions of the final forgings at the various optimization iteration steps are given, too. The optimization results show that the function of multiple-objective optimization is feasible.Genetic algorithm is widely used in many fields according to its simple ideas and easy realizability. The multiple-objective optimal method whose optimal objective is to involve the net shape of forgings and the microstructure of forgings is put forward. The multiple objective optimization function which is linearly combined by two sub-objective functions is given. The formula of the shape sub-objective function and the microstructure distribution sub-objective function are described in detail. The perform die shape is represented using the cubic B-spline. The software for optimal design method of the microstructure in forging process based on preform design using genetic algorithm is developed. The macro process parameters optimal design of the cylinder upsetting process and the microstructure optimal design based on perform shapes of an H-shaped forging in plane strain deformation realize the optimization of the final microstructure distribution and fine grain size. The initial billet dimension and deformation amount in performing process are also optimized. The optimal arithmetic is proved greatly credible.Taking H-shaped forging component as an example, an experiment is contrived for validating the microstructure simulation and optimization software developed in this dissertation. Dies used in the experiment and the experimental scheme are designed in detail. The experiment is carried out for applying the data testifications for the dependability of the computer results. The comparison of the experimental results and the optimization results shows good consistency. And thus, is testifies the stability and realizability of the multi-objective optimization design theory and system based on the microstructure and perform shapes in forging processes.更多还原