论文部分内容阅读
通过全原子分子动力学(MD)与等温耗散粒子动力学(DPD)的串行耦合,提出了面心立方金属粗粒化模型的建立方法。该方法将一定数量的原子粗粒化为单个介观DPD粒子,假设DPD粒子间作用势的表达式为Sutton-Chen势函数形式,利用遗传算法,以MD和DPD计算的单晶金属常温(298K)等温线相一致为目标,确定了DPD粒子间作用势函数的参数。对单晶铜纳米棒的轴向拉伸开展MD和DPD对比模拟,发现在纳米棒弹性响应阶段,两者计算结果吻合较好,而屈服应力和屈服应变存在一定差距。建议在优化DPD势函数参数时,引入更多的材料力学响应信息,进一步提高介观DPD模型的准确性。
Through the serial coupling of all-atom molecular dynamics (MD) and isothermal dissipative particle dynamics (DPD), a method for establishing the face-centered cubic metal coarse-grained model was proposed. In this method, a certain number of atoms are coarse-grained into single mesoscopic DPD particles. Assuming that the potential of the DPD particles is expressed by the Sutton-Chen potential function, using the genetic algorithm, the single crystal metal at room temperature (298K ) Isotherms. The parameters of the potential function of DPD particles were determined. The MD and DPD simulation of the axial stretching of single crystal copper nanorods was carried out. It was found that the calculated results agree well with each other in the elastic response phase of nanorods. However, there is a certain gap between yield stress and yield strain. It is suggested that more material mechanics response information should be introduced to optimize DPD potential function parameters to further improve the accuracy of mesoscopic DPD model.