工业气固流化床的快速准确模拟是化学工程学科的挑战之一。大量研究表明CFD-DEM方法是研究气固两相流的有效手段之一，但是其在工业规模流化床的模拟应用还鲜有报导。这主要是因为：（1）工业流化床内固体颗粒数目巨大导致计算量非常巨大；（2）采用非结构化网格导致计算效率和精度低、收敛性差。申请人所在的EMMS团队已经通过发展粗粒化CFD-DEM模型和基于GPU的大规模并行计算基本上解决了第一个关键问题，本申请聚焦解决第二个关键问题。.本申请完全抛弃使用非结构化贴体网格的基本思想，在结构化笛卡尔网格(Cartisian grid)下采用浸没边界法（IBM）处理复杂边界问题，并提出一种通过壁面流场重构且无需内迭代的CFD-DEM-IBM算法。新算法在计算效率上比OpenFOAM平台现有算法提高两个数量级，从而突破CFD-DEM方法从基础研究到工业应用的瓶颈，实现工业流化床的快速准确模拟。

Accurate and efficient simulation of industrial-scale gas-solid fluidized beds is one of the challenges of chemical engineering field. Extensive research has concluded that CFD-DEM method is an effective method to explore the hydrodynamics of gas-solid flow, however, its industrial application has seldom reported. This is mainly because: (1) the difficulty in tracking the massive numbers of solid particles inside industrial scale fluidized beds, and (2) the use of unstructured computational grid to fit the complex geometry of fluidized beds results in low computational efficiency and accuracy, even causes convergence problem constantly. The EMMS group at IPE, CAS has largely solved the first challenge via developing coarse-grained CFD-DEM method and GPU-based high performance computing. The proposal focus on solving the second challenge. .We use immersed boundary method with Cartesian grid to meet the challenge of complex geometry of industrial fluidized beds and propose a new CFD-DEM-IBM algorithm without the need of iteration at the implementation of wall boundary condition. With the newly developed CFD-DEM-IBM method, it was able to speed up the computational efficiency by two orders of magnitude as compared to the method available in OpenFOAM, thus achieving the accurate and efficient simulation of the hydrodynamics of gas-solid flow in industrial fluidized beds.