Viscosity and phase boundary predictions through coarse grained molecular modelling

通过粗粒分子模型预测粘度和相界

• 批准号: 2292575
• 金额: --
• 依托单位: Imperial College London
• 依托单位国家: 英国
• 项目类别: Studentship
• 财政年份: 2018
• 资助国家: 英国
• 起止时间: 2018 至 无数据
• 项目状态: 已结题
• 来源: https://gtr.ukri.org/projects?ref=studentship-2292575
• 关键词: Viscosity; phase; boundary; predictions; through

The output of dynamic simulation programmes is a trajectory of configurations through time of all the molecules of interest and their respective arrangements against each other. In some cases, it is possible to extract physical quantities, such as diffusivity and viscosity, directly from the trajectories. In other cases, molecules may form into segregated aggregates, repeating microstructures, or any other types of patterns. In these cases, the key step in using these trajectories is to extract pattern information in some quantitative form and link these pattern descriptors to known physical properties. Such physical properties could include phase boundaries or stability observations. This constitutes a sort fo machine learning problem to pring out the full richness of dynamic simulation results.In this project, the success of the molecular modelling framework would be demonstrated by establishing relationships between the appearance of mesophases and the changes in viscosity, attempting to find ways of reproducing experimental observations. The fundamental understanding of these relationships within complex fluid processing is vital to several P&G businesses including Fabric and Home Care and Beauty.

动态模拟程序的输出是所有感兴趣的分子的构型随时间的轨迹，以及它们各自的相互排列。在某些情况下，可以直接从轨迹中提取物理量，如扩散率和粘度。在其他情况下，分子可能形成分离的聚集体，重复的微观结构，或任何其他类型的模式。在这些情况下，使用这些轨迹的关键步骤是以某种定量形式提取模式信息，并将这些模式描述符与已知的物理特性联系起来。这些物理性质可以包括相边界或稳定性观察。这就构成了一种机器学习的问题，以充分丰富动态仿真的结果。在这个项目中，分子建模框架的成功将通过建立中间相的外观和粘度变化之间的关系来证明，试图找到再现实验观察的方法。对复杂流体处理过程中这些关系的基本理解对宝洁的几个业务至关重要，包括织物和家庭护理和美容。