Fluid Dynamics - understanding fluid dynamics to improve our lives

在含颗粒流体中加入小浓度的高分子量聚合物，将不沉降的细固体从水悬浮液中分离出来，是一种很有前途的促进沉降的方法。在核废料处理中，多相污泥废物的管理对正在进行的旨在将聚集的废物运送到临时储存设施的作业至关重要。它也与矿物工业和一般水处理应用有关。本项目旨在利用数值模拟研究聚合物颗粒吸附和后续絮凝动力学的基础。非平衡朗之万动力学和有限可扩展非线性弹性（FENE）哑铃模型将用于预测聚合物与模拟核废料的相互作用，模拟为完全分解的粒子。将采用高精度的直接数值模拟技术预测剪切和各向同性湍流等相关流动条件。机器学习还将用于增强预测能力，并限制所需的高计算成本模拟的数量。对颗粒-聚合物体系的全面机理理解将有助于为正在进行的核操作中的安全工业实践提供信息，并提高过滤和沉淀过程的效率。

The addition of small concentrations of high-molecular-weight polymers to particle-laden flows to separate non-settling fine solids from aqueous suspensions is a promising method to instigate settling. This is of invaluable use in nuclear waste processing, where management of multiphase sludge waste is critical to on-going operations which aim to transport aggregated waste to interim storage facilities. It is also of relevance to the minerals industry and in general water treatment applications. This project aims to investigate the underpinning polymer-particle adsorption and subsequent flocculation dynamics using numerical simulation. Nonequilibrium Langevin dynamics and the finitely-extensible nonlinear elastic (FENE) dumbbell model will be used to predict the interaction of polymers with simulated nuclear waste material, modelled as fully-resolved particles. Relevant flow conditions such as shear and isotropic turbulence will be predicted using the high-accuracy technique, direct numerical simulation. Machine learning will also be used to augment the predictive capabilities and to limit the number of high computational cost simulations required. A full mechanistic understanding of the particle-polymer systems will help inform safe industrial practice in ongoing nuclear operations and improve the efficiency of filtration and settling processes.