基于近似模型的拉延筋几何参数反求

引入响应面方法和遗传算法建立基于近似模型的拉延筋几何参数反求方法。首先以等效拉延阻力为设计变量，通过均匀拉丁方试验设计方法提取适当的设计参数样本构造响应面近似模型，并不断优化响应面模型，获取最优等效拉延阻力；然后以最优等效拉延阻力为约束条件，结合等效拉延阻力计算和小种群遗传算法反求拉延筋几何参数。整个反求过程采用等效拉延筋有限元模型进行仿真计算，避免有限元模型的网格重划分及由真实拉延筋模型引入的计算效率问题。数值算例表明，基于近似模型的拉延筋几何参数反求方法可在设计兴趣域内快速寻优，有助于加快模具设计进程，降低生产成本。     

拉延筋计算模型在数值模拟中的应用与比较

在模拟软件Dynaform中,将文献[3]中建立的拉延筋阻力计算模型作为等效拉延筋使用,模拟分析了表盘座某部件的成形.分析结果表明,在分析起皱、破裂等成形问题时,文中建立的拉延筋阻力计算模型比使用Dynaform自带等效拉延筋更接近使用真实拉延筋模拟的结果.     

汽车覆盖件成形中拉延筋的设计与数值模拟

汽车覆盖件冲压成形中容易出现未充分变形、起皱、变薄、拉裂等缺陷。为了提高成形质量,通常在模具上设置拉延筋来控制坯料的流动,从而提高成形质量。在有限元模拟仿真中,为了简化网格划分,减少计算时间,通常采用等效拉延筋,因此给出了等效拉延筋的拉延阻力和最小压边力计算模型。使用有限元分析软件Autoform模拟了某汽车覆盖件的拉延成形,并通过改变拉延筋的几何尺寸,改善了拉延成形工艺,缩短了模具的开发周期,降低了设计成本。     

两种求拉延筋约束阻力方法的对比研究

分别用两种解析模型研究了等效拉延筋约束力的计算过程，一是基于能量原理的改进Stoughton模型，二是对李东升等提出的直接法解析模型进行改进。同时对两种模型的基本假定进行了调整和补充，使两种方法的前提条件一致，并编制程序计算了在相同变化条件下的约束力，根据两组结果数据的对比及部分数据与实验结果的比较，从实验与计算两方面证明了用解析方法模拟等效拉延筋约束力的可行性。     

确定拉延筋约束阻力的一种数值模拟方法

利用率形式的虚功原理和考虑弯曲影响的 Mindlin曲壳单元 ,根据塑性变形体积不可压缩的假设 ,将大变形弹塑性有限元应用于板料成形过程中 ,建立了接触摩擦模型。在此基础上 ,利用自行开发的软件 Sheet Form模拟板料流经拉延筋的过程 ,确定板料成形数值模拟中建立等效拉延筋模型需要的拉延筋约束阻力、塑性厚向应变和最小压边力 3个边界条件 ,并与Nine等人的实验结果进行对比 ,证明了该方法的有效性     

等效拉延筋模型在板料成形数值模拟中的具体实现

在板料成形中,为有效控制材料的流动经常需要设置拉延筋。综合权衡数值模拟的效率和精度,通常采用等效拉延筋代替实际拉延筋。论述了等效拉延筋模型的实现。首先提出了一种新的拉延筋接触搜索算法,然后具体说明了节点拉延筋约束阻力、塑性厚向应变的施加方案。利用自主开发的汽车覆盖件冲压成形模拟软件SheetForm模拟了方形件的成形过程,证明了该模型的可行性。     

基于响应面法的复杂件回弹有限元模拟优化设计

由于单元多、复杂件的回弹很难预测和控制。基于应变增加降低回弹理论,采用变拉延筋阻力方法,对复杂覆盖件的回弹进行了控制研究。拉延筋阻力的预测,采用二维平面应变有限元法。考虑到拉延筋设计的困难,提出了基于数据挖掘,进行拉延筋反向设计的思想。采用响应面方法对拉延筋阻力进行了优化设计,提出采用未充分变形区的面积作为目标函数,通过响应面模型拟合得到了一个优化的阻力。采用该阻力进行模拟,发现成型件中未充分变形区得到消除,模拟发现回弹大大降低,达到设计要求。     

拉延筋对回弹的影响机理研究

采用拉延筋影响试验、平板拉伸试验,研究了板料过拉延筋后几何参数和材料参数的变化情况,结合翻边回弹试验和回弹一维分析方法,研究了拉延筋对回弹的影响机理。研究表明,板料过拉延筋后板内的残余应力和材料的硬化是影响回弹的关键因素;等效拉延筋模型即使能提供准确的拉延阻力,一般也难以满足精确计算回弹的要求。     

一种基于反求参数的拉延筋解析模型

建立了考虑板料中性层偏移、包辛格效应等因素的拉延筋解析模型。材料本构关系采用弹塑性幂次强化，利用平面应变假设来求解拉延筋阻力。验证了此模型的有效性后，采用改进的小种群遗传算法，通过拟合Nine的试验值，反求出中性层的偏移量、包辛格效应产生的应力降低率，并应用到拉延筋解析模型中，得到一个更为准确的拉延筋解析模型。     

板料成形拉延筋拉延阻力和变形的有限元预测

利用有限元MSC.MARC软件,通过建立试验钢板的拉延筋模型,对其拉延过程进行了模拟,得到了板料被拉延筋夹持后移动过程的应力和板料厚度的分布。并且通过提取板料的节点,得出其节点力,把所有节点力相加即得到拉延阻力。     

Optimization of the structural and process parameters in the sheet metal forming process

The quality of the sheet metal forming product is determined by defects such as wrinkling, springback, etc. Optimization techniques can avoid such defects while the desired final shape is obtained. The design variables of the optimization process consist of the structural parameters and process parameters. The structural parameters are the initial blank shape, geometry, etc. and the process parameters are the blank holding force (BHF), the drawbead restraining force (DBRF), etc. In this paper, the two groups of parameters are separately optimized. The structural parameters are optimized by the equivalent static loads method for non linear static response structural optimization (ESLSO) and the process parameters are optimized by the response surface method (RSM). A couple of examples are solved by the iterative use of ESLSO and RSM, and the solutions are discussed.     

A FE technique to improve the accuracy of drawbead models and verification with channel drawing experiments of a high-strength steel

A drawbead modeling technique is presented to improve the accuracy of finite element simulations in terms of part draw-in and thickness predictions and validated with channel drawing experiments of a high-strength low-alloy steel. The drawing characteristics of 1.5-mm thickness blanks are obtained with strip drawing tests with a round drawbead, and drawbead model parameters are computed for three bead settings. The consequences of bending deformation cycles are determined experimentally on strip draw-in and thickness values, and model limitations of equivalent drawbead elements are also assessed for test conditions in which the drawbead restraint force is lower than the sectional yield limit. The influence of omitted drawbead geometry and overestimated drawbead-exit thickness are described using an analytical model, and a closed form expression is obtained to correct draw-in model error under sectional deformation conditions. Blank thickness and equivalent strain at the drawbead exit are additional drawbead model parameters of the proposed technique. Then, drawing simulations of a variable section, open-ended channel part are performed. The drawbead design, bead settings and tool-blank interface conditions are identical to those in strip drawing tests. Computed draw-in and thickness distributions were compared with measurements on produced channels using an experimental channel draw die. It is concluded that simulation models, based on drawbead force parameters only, overestimate blank thickness at the die entry and bring about relatively high draw-in values along part border lines. The thickness distribution predicted with proposed technique shows an enhanced correlation with on-part thickness measurements, and bead penetration effects on channel border lines are also simulated acceptably.     

真实拉延筋参数化建模及其在薄板冲压仿真中的应用

提出一种参数化的拉延筋网格模型建模方法,在自主开发的有限元前处理软件中,建立了具有真实几何尺寸的半圆形、三角形拉延筋和拉深槛的网格模型,并提供了灵活的拉延筋布置手段。提出一种改进的全四边形网格加密方法,对成形过程中将会流过拉延筋区域的板料网格,以及可能与模具上曲率变化大的区域相接触的板料网格进行加密操作,以满足网格的适应性要求。提出一系列的网格拓扑清理模版,对加密后的网格进行拓扑清理操作,有效地提高了板料网格的质量。大量的算例证明所提出方法具有较高的精度和较强的工程实用性。     

Optimization of pulsed GTA welding process parameters for the welding of AISI 304L stainless steel sheets

Optimization of pulsed gas tungsten arc welding (pulsed GTAW) process parameters was carried out to obtain optimum weld bead geometry with full penetration in welding of stainless steel (304L) sheets of 3 mm thickness. Autogenuous welding with square butt joint was employed. Design of experiments based on central composite rotatable design was employed for the development of a mathematical model correlating the important controllable pulsed GTAW process parameters like pulse current (I _p), pulse current duration (T _p), and welding speed (S) with weld bead parameters such as penetration, bead width (W), aspect ratio (AR), and weld bead area of the weld. The developed models were checked for adequacy based on ANOVA analysis and accuracy of prediction by conducting a confirmation test. Weld bead parameters predicted by the models were found to confirm observed values with high accuracy. Using these models, the main and interaction effects of pulsed GTAW process parameters on weld bead parameters were studied and discussed. Optimization of pulsed GTAW process parameters was carried out to obtain optimum bead geometry using the developed models. A quasi-Newton numerical optimization technique was used to solve the optimization problem and the results of the optimization are presented.     

Optimization of drawbead design in sheet forming using one step finite element method coupled with response surface methodology

In a sheet forming process, drawbead plays an important role on the control of the material flow. In this paper, a numerical procedure for the design of forming processes is described. It is based on the coupling of an optimization technique and the simplified one step finite element method (also called inverse approach). The optimization technique allows adjustment of the process parameters so that specified criteria are fulfilled. Response surface methodology (RSM) is a global approximation method, which is ideally suited for solving highly nonlinear optimization problems. The finite element method, in addition to predicting the response of the process to certain parameters, allows assessment of the effect of a variation in these parameters on this response. The authors utilize the one step method at the preliminary design stage to supply stress or strain information for the following optimization using RSM. The procedure for this optimization process is fully described. The front fender for Numisheet 2002 is presented and the real defect free workpiece is produced to demonstrate the usefulness of the proposed optimization procedure. A comparison between the two forming limit curves (FLC) before and after optimization and results obtained using the precise incremental commercial software DYNAFORM based on the explicit dynamic approach verify that the optimization design method of drawbead could be successfully applied in designing actual tools of auto body cover panels.     

考虑接触演变的板材回弹过程建模与优化

回弹是由工件在卸载后的弹性变形引起的。板料成形过程中为了控制成形件的最终形状,必须进行回弹设计优化。准确预测回弹对于板料成形过程的模具设计非常重要。降低回弹模拟结果与试验结果的偏差是设计过程中的难题。基于NUMISHEET’02的自由弯曲标准考题考虑板材与模具间的接触演变过程,建立了一个有限元模型来预测回弹。采用一个常规的优化方法对有限元分析中的材料和单元模型进行了分析,研究发现不同模型对回弹结果有较大影响。模拟结果与参考文献中的试验结果比较表明了模型的正确性和可行性。     

Optimization techniques applied to single point incremental forming process for biomedical application

Single point incremental forming (SPIF) process has the potential to replace conventional sheet forming process in industrial applications. For this, its major defects, especially poor geometrical accuracy, should be overcome. This process is influenced by many factors such as step size, tool diameter, and friction coefficient. The optimum selection of these process parameters plays a significant role to ensure the quality of the product. This paper presents the optimization aspects of SPIF parameters for titanium denture plate. The optimization strategy is determined by numerical simulation based on Box–Behnken design of experiments and response surface methodology. The Multi-Objective Genetic Algorithm and the Global Optimum Determination by Linking and Interchanging Kindred Evaluators algorithm have been proposed for application to find the optimum solutions. Minimizing the sheet thickness, the final achieved depth and the maximum forming force were considered as objectives. For results evaluation, the denture plate was manufactured using SPIF with the optimum process parameters. The comparison of the final geometry with the target geometry was conducted using an optical measurement system. It is shown that the applied method provides a robust way for the selection of optimum parameters in SPIF.     

覆盖件冲压仿真参数化建模方法

针对在覆盖件冲压成形领域对快速、自动化的有限元网格建模方法的迫切需求,提出一种快速的汽车覆盖件冲压仿真建模的思路与方法.将覆盖件冲压工艺设计与冲压成形仿真前处理集成,使用散乱三角面片模型,在自主开发的CAE前处理软件中,进行参数化的工艺补充面和压料面设计.模型的网格剖分与冲压工艺设计同时进行,自动生成整套模具的网格模型供冲压仿真计算.为了真实地模拟板料网格的流动,提出了参数化的真实拉深筋的模型建模方法和板料网格的预细分方法.完成了相关软件的开发.多个汽车覆盖件冲压工艺设计和冲压仿真计算实例说明了该方法的有效性.     

基于有限元逆算法的拉深筋工艺设计和优化

汽车覆盖件拉深成形中，一般通过设置适当的拉深筋控制成形过程中的板料塑性流动规律来提高覆盖件成形质量。针对覆盖件工艺设计需求，提出一种基于有限元逆算法的拉深筋工艺优化算法。该算法以灵敏度优化方法为基础，考虑了板料的成形度、破裂和起皱等成形缺陷。在板料成形模拟FASTAMP系统中，开发了拉深筋优化模块，并以实际覆盖件为例，验证了该算法能快速准确地模拟等效拉深筋力的布置情况以及优化板料的成形性。     

Analysis of drawbead process by static-explicit finite element method

The problem analyzed here is a sheet metal forming process which requires a drawbead. The drawbead provides the sheet metal enough tension to be deformed plastically along the punch face and consequently, ensures a proper shape of final products by fixing the sheet to the die. Therefore, the optimum design of drawbead is indispensable in obtaining the desired formability. A static-explicit Finite element analysis is carried out to provide a perspective tool for designing the drawbead. The finite element formulation is constructed from static equilibrium equation and takes into account the boundary condition that involves a proper contact condition. The deformation behavior of sheet material is formulated by the elastic-plastic constitutive equation. The finite element formulation has been solved based on an existing method that is called the static-explicit method. The main features of the static-explicit method are first that there is no convergence problem. Second, the problem of contact and friction is easily solved by application of very small time interval. During the analysis of drawbead processes, the strain distribution and the drawing force on drawbead can be analyzed. And the effects of bead shape and number of beads on sheet forming processes were investigated. The results of the static explicit analysis of drawbead processes show no convergence problem and comparatively accurate results even though severe high geometric and contact-friction nonlinearity. Moreover, the computational results of a static-explicit finite element analysis can supply very valuable information for designing the drawbead process in which the defects of final sheet product can be removed.