论文标题
双向耦合点粒子模拟的一般速度校正方案
A general velocity correction scheme for two-way coupled point-particle simulations
论文作者
论文摘要
载有颗粒流体流量模拟中使用的Euler-Lagrange点粒模型的准确性取决于通过闭合模型对粒子力的准确估计。典型的力闭合模型需要计算粒子位置的滑动速度，这又需要准确估算不受干扰的流体速度。但是，当流体和粒子相是双向耦合时，流体速度场就会受到粒子的存在的干扰。一种常见的做法是使用受干扰的速度计算粒子力，这可能导致误差在预测粒子动力学时的误差。在这项工作中，开发了一种一般速度校正方案，以促进具有和不带无滑动壁的颗粒流体流中不受干扰的流体速度的准确估计。该模型可以处理不同大小，任意插值功能，具有较大纵横比的各向异性网格以及壁挂式流动的颗粒。目前的校正方案是由Esmaily＆Horwitz（JCP，2018）在无限粒子载粒子流动的最新工作中进行的。开发了壁挂式流动所需的修改，以使任何壁距离处的不受干扰的流体速度准确地恢复，渐近地接近远离墙壁的颗粒的无界方案。对粒子的沉降速度并行和垂直于无滑动壁的固定速度进行了详细的验证测试。一系列流参数和网格配置；详细考虑了在充满粒子的湍流通道流中通常遇到的各向异性网格的纵横比。当考虑到壁效应时，与无界校正方案相比，当前的校正方案减少了一个数量级的近粒粒子运动时的误差。
The accuracy of Euler-Lagrange point-particle models employed in particle-laden fluid flow simulations depends on accurate estimation of the particle force through closure models. Typical force closure models require computation of the slip velocity at the particle location, which in turn requires accurate estimation of the undisturbed fluid velocity. However, when the fluid and particle phases are two-way coupled the fluid velocity field is disturbed by the presence of the particle. A common practice is to use the disturbed velocity to compute the particle force which can result in errors as much as 100% in predicting the particle dynamics. In this work, a general velocity correction scheme is developed that facilitates accurate estimation of the undisturbed fluid velocity in particle-laden fluid flows with and without no-slip walls. The model can handle particles of different size, arbitrary interpolation functions, anisotropic grids with large aspect ratios, and wall-bounded flows. The present correction scheme is motivated by the recent work of Esmaily & Horwitz (JCP, 2018) on unbounded particle-laden flows. Modifications necessary for wall-bounded flows are developed such that the undisturbed fluid velocity at any wall distance is accurately recovered, asymptotically approaching the unbounded scheme for particles far away from walls. A detailed series of verification tests were conducted on settling velocity of a particle in parallel and perpendicular motions to a no-slip wall. A range of flow parameters and grid configurations; involving anisotropic grids with aspect ratios typically encountered in particle-laden turbulent channel flows, were considered in detail. When the wall effects are accounted for, the present correction scheme reduces the errors in predicting the near-wall particle motion by one order of magnitude smaller values compared to the unbounded correction schemes.
