物理
格子Boltzmann方法
玻耳兹曼关系
统计物理学
计算流体力学
机械
水电站模型
流体力学
Bhatnagar–Gross–Krook操作员
颗粒流
经典力学
直接模拟蒙特卡罗
两相流
流量(数学)
雷诺数
湍流
蒙特卡罗方法
统计
数学
动态蒙特卡罗方法
作者
Shi Tao,Xilin Zhang,Wenhao Wang,Hao Wu,Liang Wang,Qing He
摘要
A moving multiblock (MMB) grid refinement method is developed for lattice Boltzmann modeling of fluid–solid flows. This method addresses the need for high resolution near freely moving bodies, particularly in pore-scale simulations of porous particles. The MMB method is an adaptation of the traditional static multiblock (SMB) scheme, where adjacent subdomains overlap by one coarse mesh unit to facilitate efficient information exchange. However, the computationally intensive temporal interpolation used in the SMB method is replaced by spatial interpolation in the MMB. Additionally, each grid block begins to move collectively following a single time step evolution of the coarsest grid block, which is inspired by the moving domain method. Consequently, only the buffer layer of fine grids that migrates toward the coarse grid side needs to be rebuilt, which lowers the computational costs associated with spatial interpolation while maintaining method accuracy. The second-order accuracy of the method is verified through simulation of Poiseuille flow. The method is subsequently applied to simulate particle motion in Poiseuille and Couette flows, the sedimentation of an ellipse under gravity in a vertical channel, and harmonic oscillation of a cylinder in a stationary fluid. The flow field exhibits smoothness across boundaries, and the obtained results correlate well with established findings in the literature, demonstrating the method's feasibility and accuracy for fluid-particle flows. We examine pore-scale simulations of a permeable particle translating inside channel flow as a particular application. Results indicate that porous particles migrate toward an equilibrium position between the channel wall and centerline.
科研通智能强力驱动
Strongly Powered by AbleSci AI