Stress jump boundary condition capturing for the lattice Boltzmann simulation methods

当两种或两种以上不同的相或类型的流体聚集在一起时，就会发生多相流。它们出现在物理和工业类型系统的广泛范围内。这类系统在食品、石油、天然气、废物和泥浆的加工、生产和运输中仅举几例；蒸发器、冷凝器、泵和涡轮机的能源生产；在自然系统中，如地球物理和地球化学流动、储层萃取/过滤、生物和生化流动。在这样的系统中，不同相相交的点被称为界面，这个界面区域由于发生的应力而引起许多复杂的流变现象。悬浮动力学、润湿、堵塞、聚并、破裂、碰撞和毛细作用等现象都是界面主导的流动，在典型的工程场景中不容易在数学上进行预测。在这些情况下，数值计算机模拟已被证明是成功理解、诊断、预测和优化系统的宝贵工具。用于工程多相流的计算机数值模拟方法是晶格玻尔兹曼方法。然而，在这种有前途的方法中，缺点是为了解决和计算必要的流细节而花费了大量的资源来平滑和扩展接口。这严重限制了模拟的物理代表性大小和工业上有用的应用范围，这些应用可以从这种类型的预测建模中受益，这通常需要避免长时间的开发延迟。本研究计划将开发数值晶格玻尔兹曼方法的全新技术，以便以精确的方式应用数学上正确的应力跳变边界条件。这将释放昂贵的计算资源，这意味着(i)可以修改现有的模拟以占用一小部分（估计最多减少4倍）的时间和内存，（ii）可以模拟新的更大的更具物理代表性的，准确的和工业相关的多相流范围。为了确保新开发的技术的正确性，将对它们进行已知数据的测试，并与现有技术进行比较，以证明通过本研究有望实现的显着增强。将使用本研究工作中开发的技术的研究类型将主要是与多相相关的，但值得注意的是，所开发的技术将适用于晶格玻尔兹曼方法中涉及应力边界的任何输运现象。例如，流入多孔介质模型的开放流体的连接处包含应力跳变。更具体地说，这项研究将继续应用于乳剂和悬浮液的显式建模。这些流动包含大量的颗粒，具有多个相互作用的界面，主导着涌现的复杂流变行为。这种流动在食品，饮料，奶油，糊状物，生物流体（血液）和其他加工行业中很普遍，这里开发的建模工具将在未来许多年里改进非牛顿流体的构成理论，知识转移和过程优化。

Multi-phase flows occur when two or more different phases or types of fluid are brought together. They are seen to occur in a vast range of both physical and industrial type systems. Such systems are, to name but only a few, in processing, production and transportation of foods, oil, gas, waste and slurries; in energy production from evaporators, condensers, pumps and turbines; in natural systems such as geophysical and geochemical flows, reservoir extraction / filtration, biological and biochemical flows. In such systems the point at which different phases meet is termed an interface and this interfacial area gives rise to a host of complex rheological phenomena due to stresses that occur. Phenomena such as suspension dynamics, wetting, jamming, coalescence, break-up, collision and capillarity are all heavily interface dominated flows and are not readily mathematically easy to predict in typical engineering scenarios.In these cases numerical computer simulations have proved an invaluable tool in successfully understanding, diagnosing, predicting and optimising systems. A growing current state of the art class of numerical computer simulation methods used for engineering multi-phase flow is called the lattice Boltzmann method. However, in this promising method, a drawback is the large amounts of resources that are spent smoothing and broadening interfaces in order to resolve and calculate the necessary flow details. This severely restricts the physical representative size of a simulation and the range of industrially useful applications that can benefit from this type of predictive modelling which is often needed to avoid long development delays.This programme of research will develop brand new techniques for the numerical lattice Boltzmann methods in order to apply the mathematically correct stress jump boundary conditions in a sharp exacting manner. This will free up expensive computational resources which means (i) that existing simulations can be modified to take a fraction (estimated at up to 4 times less) of the time and memory, (ii) that a new range of larger more physically representative, accurate and industrially relevant multi-phase flows can be modelled. To ensure the correctness of the newly developed techniques they will be tested against known data and compared against the present day techniques in order to demonstrate the significant enhancements expected to be achieved through this research. The types of research that will use the techniques developed in this research work will predominantly be multi-phase related but it is noted that the techniques developed will apply to any transport phenomena that involves stress boundaries within the lattice Boltzmann methods. For example the junction of an open fluid flowing into a porous media model contains a stress jump. More specifically this research will go on to be applied to the explicit modelling of emulsions and suspension. These are flows that contain a large number of particles with multiply interacting interfaces dominating the emergent complex rheological behaviour. Such flows are prevalent in the foods, drinks, creams, pastes, bio-fluids (blood) and other processing industries and the modelling tools developed here will lead to improved constitutional theories of non-Newtonian fluids, knowledge transfer and process optimisation for many years to come.