Lattice Boltzmann Methods for Real Time Simulations of Complex Physics

在许多工业、生物、环境和地球物理过程中，可以遇到复杂的物理和有/没有传热和传质的多相流；例如，血液在可变形的通道中流动。血液由红细胞、白细胞和血小板组成。这些不相容的粒子是可变形的，漂浮在等离子体中，等离子体在弹性通道中流动，这进一步增加了问题的复杂性。在工业中，大多数流动是湍流的，通常与多物理场有关。例如在水和油分离过程中气泡和液滴的传输。了解这些问题背后的物理和不同参数的影响对工业界和学术界都至关重要。近年来，人们对将晶格玻尔兹曼方法（LBM）应用于离散力学的兴趣日益浓厚，它已成为解决此类物理问题的一种强有力的技术。LBM易于编码，并且具有允许包含多物理场来解决手头问题的灵活性，从而增加了一系列进一步可能的应用程序。然而，该方法相对较新，特别是在多物理场领域仍有很大的扩展空间。基于LBM的湍流模型以及多尺度问题也需要进一步的探索。本文的主要目标是：了解具有热质传递的层流和湍流的可变形边界的复杂流动的物理特性，并利用基于LBM的离散方法建立这些系统的数学模型。开发的模型将应用于具体问题。此外，该模型将利用LBM的局部特性，利用多处理器图形卡来加速实时仿真。长期目标是开发一个统一的平台来模拟流体流动和固体相互作用的广泛控制参数。此外，将进行实验室规模的实验来验证所开发的模型。实验和流程可视化将加强我们对问题的物理理解。该项目将培养几名解决多学科、具有挑战性的工业/生物问题的研究生。

Complex physics and multi-phase flows with/without heat and mass transfer can be encountered in many industrial, biological, environmental, and geophysical processes; for instance, blood flow in deformable channels. The blood consists of red and white blood cells and platelets. These immiscible particles are deformable and float in a plasma which flows in elastic channels, adding further complexity to the problem. In industry, most flows are turbulent and usually associated with multi-physics. Examples are the transport of bubbles and droplets in water and oil separation processes. Understanding the physics and the effect of the varying parameters underlying such problems are of vital interest to industry and academia. Recently, there has been a growing interest in applying the Lattice Boltzmann method (LBM) to discrete mechanics which has emerged as a powerful technique to tackle such physical problems. The LBM is easy to code, and enjoys the flexibility of allowing the inclusion of multi-physics to solve the problems at hand, adding an array of further possible applications. However, the method is relatively new, much room remains for expanding it especially in the realm of multi-physics. The turbulent flow model based on the LBM is also in great need of further exploration, as are multi-scale problems. The main objectives of this proposal are: to understand the physics of complex flows with deformable boundaries for laminar and turbulent flows with heat and mass transfer, and to develop mathematical models for such systems using discrete method based on the LBM. The developed models will be applied to specific problems. Also, the model will take advantage of the local nature of LBM to utilize multi-processor graphic card to speed real time simulations. The long term objective is to develop a unified platform to model fluid flow and solid interaction for a wide range of the controlling parameters. Moreover, lab scale experiments will be performed to validate the developed models. The experiments and flow visualizations will add to reinforce our understanding of the physics of the problem. The project will train several graduate students for solving multi-disciplinary, challenging industrial/biological problems.