Massively parallelized models of fluid-solid multiphase flow

环境和工业流体动力学中的一个关键问题是能够有效地对颗粒材料在流内的传输进行数值模拟，该流与来自边界（例如沉积物床）的静止颗粒材料的沉积和/或侵蚀耦合。这一挑战在很大程度上仍然没有得到解决，因为颗粒传输是由湍流的小尺度过程控制的，而感兴趣的系统尺度可能要大很多个数量级。示例包括：沉积物在环境水流中的迁移，从山区河流到深海;废水在管道中的迁移;燃烧动力学;以及化学和制药生产。在环境系统中，特别是流动的流体动力学（例如河流，河口，海岸等）。取决于河床的形态动力学了解水文动力学和地貌动力学之间的耦合对于土地和海洋利用管理以及地质灾害预测和减灾至关重要。在这个研究项目中，一个内部的格子玻尔兹曼方法（LBM）代码，这是写在大规模并行图形处理单元（GPU）上运行，将进一步开发，以研究多相流体-固体流动与环境背景。

A key problem in environmental and industrial fluid dynamics is the ability to efficiently numerically model the transport of particulate material within a flow coupled with aggradation and/or erosion of stationary particulate material from boundaries (e.g. sediment bed). This challenge has remained largely unresolved as particle transport is controlled by the small-scale processes of turbulent flow whilst the system scales of interest may be many orders of magnitude larger. Examples include: sediment transport in environmental flows, from mountain rivers to deep oceans; wastewater transport in pipes; combustion dynamics; and chemical and pharmaceutical production. In environmental systems, in particular, hydrodynamics of the flow (e.g. rivers, estuaries, coasts etc.) are dependent on the morphodynamics of the bed. Understanding of the coupling between hydro- and morphodynamics is essential in land and marine use management and geohazard prediction and mitigation. In this research project, an in-house Lattice Boltzmann Method (LBM) code, which was written to run on massively parallel Graphics Processing Units (GPUs), will be further developed to study multiphase fluid-solid flows with environmental context.