An adaptive multi-moment FVM approach for incompressible flows

An adaptive multi-moment FVM approach for incompressible flows
复制标题

DOI:
10.1016/j.jcp.2018.01.006
复制
发表时间:
2018-04
期刊:
J. Comput. Phys.
影响因子:
--
通讯作者:
Cheng Liu;Changhong Hu
Cheng Liu;Changhong Hu
中科院分区:
其他
文献类型:
--
作者:
Cheng Liu;Changhong Hu

文献摘要

被引文献

相似文献

在本研究中,提出了一种基于块结构自适应笛卡尔网格的多力矩有限体积法(FVM)来模拟不可压缩流动。遵循约束插值轮廓(CIP)方法的思想,提出了一种保守插值方案,用于新创建网格的延长操作。使用锐浸入边界 (IB) 方法对浸入刚体进行建模。应用移动最小二乘 (MLS) 插值方法重建固体表面周围的速度场。提出了一种自适应网格上拉普拉斯算子离散化的有效方法。对几个测试用例进行数值模拟以验证所提出的方法。对于粘性流通过脉冲启动圆柱体的情况(Re= 3000, 9500),计算出的表面涡度与贴体方法的结果一致。对于中等雷诺数 (Re= 10000, 45000) 的快速俯仰 NACA 0015 翼型的情况,预测的阻力系数 (C D) 和升力系数 (C L) 与其他数值或实验结果非常吻合。对于以规定运动 (Re= 5000, 40000) 流经俯仰板的粘性流的 2D 和 3D 模拟,预测的 C D、C L 和 C M(力矩系数)与其他数值方法获得的结果非常一致。
In this study, a multi-moment finite volume method (FVM) based on block-structured adaptive Cartesian mesh is proposed for simulating incompressible flows. A conservative interpolation scheme following the idea of the constrained interpolation profile (CIP) method is proposed for the prolongation operation of the newly created mesh. A sharp immersed boundary (IB) method is used to model the immersed rigid body. A moving least squares (MLS) interpolation approach is applied for reconstruction of the velocity field around the solid surface. An efficient method for discretization of Laplacian operators on adaptive meshes is proposed. Numerical simulations on several test cases are carried out for validation of the proposed method. For the case of viscous flow past an impulsively started cylinder (Re= 3000, 9500), the computed surface vorticity coincides with the result of the body-fitted method. For the case of a fast pitching NACA 0015 airfoil at moderate Reynolds numbers (Re= 10000, 45000), the predicted drag coefficient (C D) and lift coefficient (C L) agree well with other numerical or experimental results. For 2D and 3D simulations of viscous flow past a pitching plate with prescribed motions (Re= 5000, 40000), the predicted C D, C L and C M (moment coefficient) are in good agreement with those obtained by other numerical methods.