遗传算法结合有限元模拟对液压胀形加载路径的优化

加载路径对液压胀形件的最终质量有着至关重要的影响。在优化加载路径的过程中,设计变量的取值区间对模拟时间的减少及最优解的收敛有重要的意义。文章根据塑性理论分析,给出了管形件液压胀形过程中弹性变形阶段、塑性变形阶段及贴模阶段的轴向喂入量及内压力的工程计算表达式。选取应力比为优化变量,依据屈服准则,得出了胀形过程中合理的变量区间。利用有限元软件包、优化算法及工程计算表达式,得到了液压胀形最优的加载路径。实验结果和有限元模拟结果一致性较好,该方法为轴对称件的加载路径优化提供了新的思路。

Optimization for loading paths in hydroforming using genetic algorithm in conjunction with FE codes

The loading paths in the tube hydroforming have the critical effects on the final part quality. During optimization for the loading paths, the design variable hounds are very important to reduce the computational time and obtain the optimum solutions. According to theoretical analysis, the engineering computational expressions of internal pressure and axial feed are proposed in the yield stage, the free forming stage and calibration stage. Taking stress ratio as the design variables, the feasible variables bonds are given on the basis of yielding criterion, The optimal loading paths are obtained by using FE code packages in conjunction with genetic algorithm and the engineering computational expressions. The experiment results have good agreements with the theoretical results, and it indicates this novel optimization method is a applied to the axisymmetrical part.