枪弹穿甲过程仿真的有限元接触模型与分区并行计算误差特性

Finite element contact modeling method and error characteristics of partitioned parallel computations for bullet penetration simulations

为了提高小口径普通钢芯弹和穿甲燃烧弹侵彻均质钢靶板的仿真分析精度,分析了接触力罚函数算法的接触刚度参数和接触摩擦系数对弹道极限速度计算值的影响,分析了通过不同的并行计算系统得到的弹道极限速度计算值的差异及其统计规律。研究结果表明:多部件间的接触力罚函数的接触刚度参数分别设置与优化方法可以提高弹道极限速度值的计算精度;接触摩擦系数的不同设置值对穿甲燃烧弹侵彻较厚靶板的弹道极限速度计算值有显著影响,需探索多部件间的接触摩擦系数的准确定义方法;由于分区并行计算过程中多个区域的数据交换及参与计算的顺序的随机变化会导致计算结果的不一致,并行计算的弹道极限速度值存在离散性,需在同一入射速度下进行多次重复计算,并基于统计分析确定可靠的弹道极限速度值。

Improved ballistic simulations were developed for homogeneous steel sheets against small-caliber bullets. The simulations analyzed the influences of the contact-force penalty function parameters and contact friction coefficients on the contact interface state and ballistic limit velocity. The inconsistencies among computed ballistic limit velocities predicted by different partitioned parallel systems were investigated statistically. The interface stiffnesses within the contact-force penalty function should be defined between different contact pairs rather than assigning an equal interface stiffness to all contact regions to improve the accuracy of the ballistic limit velocity. Different friction coefficients between contact pairs can seriously affect the computed ballistic limit velocity for thicker steel targets against armor piercing bullets. Further studyis needed to determine the correct friction coefficients for multiple contact pairs. The exchange of regional data and the use of this data in different sequences make the partitioned parallel computations of repeated computation tasks inconsistent, which affects the ballistic limit velocities predicted by the parallel computations. Therefore, repeated calculations using the same model for the same impact velocity are needed to obtain reliable ballistic limit velocity statistics.