Molecular dynamics (MD) is appearing in increasing applications in materials science, nanotechnologies, condensed matter physics, computational physics, biochemistry, and biophysics. Finding mechanically static equilibrium configurations of molecular systems is one of the most practical tasks in MD. Most existing potential energy optimization algorithms do not permit searching equilibrium configurations through longer MD trajectories. We introduce a simple method of utilizing a microcanonical (NVE) ensemble to obtain static equilibriums of molecular systems, that is significantly faster than the standard implementations of quick-min (QM) and fast inertial relaxation engine (FIRE) optimization algorithms. The new method is based on the capability of NVE to convert potential energy to kinetic energy. The surprising efficiency of the method is illustrated using an indentation test on monolayer graphene and, in particular, the versatility of the method is illustrated using relaxation of a polystyrene chain through longer MD trajectories and large deformation. The capability of the new method in finding more stable equilibrium configurations than common optimization algorithms is demonstrated in relaxation of a pressured lubricating oil layer and a warped monolayer graphene cantilever.

中文翻译：

---

通过长轨迹使结构松弛，从而保持稳定

分子动力学（MD）出现在材料科学，纳米技术，凝聚态物理，计算物理，生物化学和生物物理领域的越来越多的应用中。寻找分子系统的机械静态平衡构型是MD中最实际的任务之一。大多数现有的势能优化算法都不允许通过更长的MD轨迹搜索平衡构型。我们介绍了一种利用微规范（NVE）集成获得分子系统静态平衡的简单方法，该方法明显快于快速最小（QM）和快速惯性松弛引擎（FIRE）优化算法的标准实现。新方法基于NVE的功能将势能转换为动能。通过在单层石墨烯上进行压痕测试说明了该方法的惊人效率，尤其是通过较长的MD轨迹和较大的变形使聚苯乙烯链松弛，从而说明了该方法的多功能性。与普通的优化算法相比，该新方法能够找到更稳定的平衡构型，这是通过放宽压力润滑油层和翘曲的单层石墨烯悬臂来证明的。