Institute of Thermodynamics and Thermal Process Engineering, University of Stuttgart, Stuttgart 70569, Germany.
J Chem Phys. 2013 Jun 21;138(23):234106. doi: 10.1063/1.4808032.
This article introduces an efficient technique for the calculation of vapor-liquid equilibria of fluids. Umbrella Sampling Monte Carlo simulations in the grand canonical ensemble were conducted for various types of molecules. In Umbrella Sampling, a weight function is used for allowing the simulation to reach unlikely states in the phase space. In the present case this weight function, that allows the system to overcome the energetic barrier between a vapor and liquid phase, was determined by a trivialized Density Functional Theory (DFT) using the PC-SAFT equation of state. The implementation presented here makes use of a multicanonical ensemble approach to divide the space of fluctuating particle number N into various subsystems. The a priori estimate of the weight function from the analytic DFT allows the parallelization of the calculation, which significantly reduces the computation time. In addition, it is shown that the analytic equation of state can be used to substitute sampling the dense liquid phase, where the sampling of insertion and deletion moves become demanding.
本文介绍了一种计算流体汽液平衡的有效技术。我们对各种类型的分子进行了巨正则系综的 umbrella 抽样蒙特卡罗模拟。在 umbrella 抽样中，使用权重函数来允许模拟达到相空间中不太可能的状态。在目前的情况下，这个权重函数，允许系统克服汽相和液相之间的能量障碍，是通过一个简化的密度泛函理论（DFT）使用 PC-SAFT 状态方程来确定的。这里提出的实现方法利用多正则系综方法将波动粒子数 N 的空间划分为各个子系统。从分析 DFT 得到的权重函数的先验估计允许计算的并行化，这大大减少了计算时间。此外，还表明可以使用解析状态方程来替代密集液相的采样，在密集液相中，插入和删除移动的采样变得很困难。
