Abstract The clustering behaviour and settling velocity of bidisperse inertial particles with sub-millimetre diameters travelling in a turbulent open channel flow are investigated using the simultaneous particle image velocimetry and particle tracking velocimetry measurement technique. Instantaneous images of flow field and inertial particles are separated based on different wavelengths of light emitted by fluorescent seeding particles and inertial particles. Four flow configurations of turbulence are generated by turbulence grids with different sizes in the open channel. The clustering behaviour of monodisperse and bidisperse inertial particles and the discrepancy between them in the different flow cases are studied. We found that bidisperse inertial particles have weaker clustering phenomenon as compared with corresponding monodisperse inertial particles and this weak clustering phenomenon is almost consistent when the Taylor microscale Reynolds number Reλ changes from 90 to 267. The relationship between settling velocity and local concentration of inertial particles is then investigated. The results support the preferential concentration as the mechanism for the enhancement of settling velocimetry when the normalized particle concentration is in the range of 0.1~10. However, the anisotropy property and large eddy effect in turbulence will greatly change the interaction between inertial particles and turbulent flows and further alter their settling velocity, such as the average settling velocity decreasing with increasing Stokes number which depends on the dissipation time scale. Accordingly, we propose a new dimensionless multiscale parameter S v η R e L 1 / 2 to take into account the effect of large eddies and turbulence anisotropy on the settling velocity. With the simultaneous measurement of particle settling velocities and fluid velocities, we also analyse the unsteady force term in the particle dynamic equation. When the turbulence intensity is small and/or the particle inertia related to particle diameter and density ratio between inertial particle and fluid is small, the unsteady term can have a large contribution and it should not be neglected in the particle dynamic equation for the point-particle method in the numerical simulations of particle-laden flows.

中文翻译：

---

双分散惯性粒子在湍流明渠流中的聚集行为和沉降速度

摘要 使用同步粒子图像测速和粒子跟踪测速测量技术研究了亚毫米直径双分散惯性粒子在湍流明渠流中的聚集行为和沉降速度。流场和惯性粒子的瞬时图像根据荧光种子粒子和惯性粒子发出的不同波长的光进行分离。湍流的四种流动配置是由明渠中不同尺寸的湍流网格产生的。研究了单分散和双分散惯性粒子在不同流动情况下的聚集行为以及它们之间的差异。我们发现双分散惯性粒子与相应的单分散惯性粒子相比具有较弱的聚集现象，并且当泰勒微尺度雷诺数 Reλ 从 90 变为 267 时，这种弱聚集现象几乎是一致的。 沉降速度与惯性粒子的局部浓度之间的关系为然后进行了调查。结果表明，当归一化粒子浓度在 0.1~10 范围内时，优先浓度是增强沉降测速的机制。然而，湍流中的各向异性和大涡效应将极大地改变惯性粒子与湍流之间的相互作用，并进一步改变它们的沉降速度，例如平均沉降速度随着斯托克斯数的增加而降低，这取决于耗散时间尺度。因此，我们提出了一个新的无量纲多尺度参数 S v η R e L 1 / 2 以考虑大涡和湍流各向异性对沉降速度的影响。通过同时测量颗粒沉降速度和流体速度，我们还分析了颗粒动力学方程中的非定常力项。当湍流强度较小和/或与粒子直径和惯性粒子与流体的密度比相关的粒子惯性较小时，非定常项可以有很大的贡献，在粒子动力学方程中不应忽略点-粒子流数值模拟中的粒子方法。我们提出了一个新的无量纲多尺度参数 S v η R e L 1 / 2 以考虑大涡和湍流各向异性对沉降速度的影响。通过同时测量颗粒沉降速度和流体速度，我们还分析了颗粒动力学方程中的非定常力项。当湍流强度较小和/或与粒子直径和惯性粒子与流体的密度比相关的粒子惯性较小时，非定常项可以有很大的贡献，在粒子动力学方程中不应忽略点-粒子流数值模拟中的粒子方法。我们提出了一个新的无量纲多尺度参数 S v η R e L 1 / 2 以考虑大涡和湍流各向异性对沉降速度的影响。通过同时测量颗粒沉降速度和流体速度，我们还分析了颗粒动力学方程中的非定常力项。当湍流强度较小和/或与粒子直径和惯性粒子与流体的密度比相关的粒子惯性较小时，非定常项可以有很大的贡献，在粒子动力学方程中不应忽略点-粒子流数值模拟中的粒子方法。通过同时测量颗粒沉降速度和流体速度，我们还分析了颗粒动力学方程中的非定常力项。当湍流强度较小和/或与粒子直径和惯性粒子与流体的密度比相关的粒子惯性较小时，非定常项可以有很大的贡献，在粒子动力学方程中不应忽略点-粒子流数值模拟中的粒子方法。通过同时测量颗粒沉降速度和流体速度，我们还分析了颗粒动力学方程中的非定常力项。当湍流强度较小和/或与粒子直径和惯性粒子与流体的密度比相关的粒子惯性较小时，非定常项可以有很大的贡献，在粒子动力学方程中不应忽略点-粒子流数值模拟中的粒子方法。