摘要
采用第一原理分子动力学(QMD)方法模拟液体钚的输运性质.计算的粘性和扩散系数在较低温度时与文献有明显差异,在实验测量范围内,模拟结果与实验一致,温度升高时数值模拟结果趋于一致.利用QMD的模拟结果计算了应力自相关函数和速度自相关函数,结果表明:在温度较低时,液体钚呈现明显的强关联特性.对于具有强关联特性的液体,利用较短时间的QMD模拟结果,通过简单e指数拟合外推到t→∞得到的扩散系数和粘性具有较大偏差,这是造成本文模拟结果与文献结果出现差异的主要原因.通过增加QMD模拟时间步数,获得了更为准确的输运性质.
We studied transport properties of liquid Pu with quantum molecular dynamics( QMD) simulations. We focused on calculating viscosity and diffusion properties from autocorrelation function. Our results are in good agreement with experiment in the range of measurement. There is obvious difference between our results and reference at lower temperatures and they show good agreement at relatively higher temperatures. Structures of autocorrelation function of off-diagonal components of stress tensor( STACF) and velocity autocorrelation function( VACF) at lower temperatures are observed and they exhibit correlated liquidlike behavior. Extrapolating autocorrelation function in t→∞ limit leads to larger error for viscosity and diffusion coefficient of correlated liquid. We obtain accurate viscosity and diffusion coefficient of liquid Pu from long simulation trajectories.
作者
王帅创
张弓木
孙博
宋海峰
田明锋
方俊
刘海风
WANG Shuaichuang;ZHANG Gongmu;SUN Bo;SONG Haifeng;TIAN Mingfeng;FANG Jun;LIU Haifeng(Institute of Applied Physics and Computational Mathematics,Beijing 100094,China)
出处
《计算物理》
EI
CSCD
北大核心
2019年第3期253-258,共6页
Chinese Journal of Computational Physics
基金
Supported by the NSFC(51671033)and the science challenge project(TZ2016001)
关键词
量子分子动力学
液体钚
粘性
扩散系数
quantum molecular dynamics simulation
liquid Pu
viscosity
diffusion coefficient