Ihle T, Kroll D M
Supercomputing Institute, University of Minnesota, Minneapolis, Minnesota 55455, USA.
Phys Rev E Stat Nonlin Soft Matter Phys. 2003 Jun;67(6 Pt 2):066706. doi: 10.1103/PhysRevE.67.066706. Epub 2003 Jun 12.
A discrete-time projection operation technique was used to derive the Green-Kubo relations for the transport coefficients of a recently introduced stochastic model for fluid dynamics in a previous paper (Part 1). The most important feature of the analysis was the incorporation of a new grid shifting procedure which was shown to guarantee Galilean invariance for arbitrary Mach number and temperature. This paper (Part 2) contains a detailed analysis of the transport coefficients of this model. An exact calculation of the first terms in the stress correlation function in the limit of infinite particle density is presented, which explicitly accounts for the cell structure introduced to define the collision environment. It is also shown that this cell structure can lead to additional contributions to the transport coefficients even at large mean free paths. Explicit expressions for all transport coefficients are derived and compared with simulation results. Long-time tails in the velocity, stress, and heat-flux autocorrelation functions are measured and shown to be in excellent agreement with the predictions of mode-coupling theory.
在前一篇论文(第1部分)中,采用离散时间投影运算技术推导了最近引入的流体动力学随机模型的输运系数的格林-久保关系。该分析的最重要特征是引入了一种新的网格移动程序,结果表明该程序可确保任意马赫数和温度下的伽利略不变性。本文(第2部分)对该模型的输运系数进行了详细分析。给出了在无限粒子密度极限下应力关联函数中首项的精确计算,该计算明确考虑了为定义碰撞环境而引入的单元结构。研究还表明,即使在大平均自由程情况下,这种单元结构也会对输运系数产生额外贡献。推导了所有输运系数的显式表达式,并与模拟结果进行了比较。测量了速度、应力和热流自相关函数中的长时间尾项,结果表明与模式耦合理论的预测非常吻合。