Clustering, rotation, and swirl of inertial particles in turbulent channel flow

Clustering dynamics of inertial particles in turbulent channel flow are studied via tessellation-based analysis of high-fidelity simulation data at Re-tau approximate to 230 with various values of mass loading (10% - 100%) and the Stokes number (St(+) = [1 - 60]). We then characterise the solenoidal, rotational, and swirling motions of clusters by computing the probability density functions (PDFs) of the divergence, curl, and helicity of the particle velocity, as well as their dependence on wall-normal distance, using the methods of Oujia et al. (2020); Maurel-Oujia et al. (2023). Particle inertia gives heavier tails to the PDFs of divergence and curl, suggesting enhanced intermittency in the convergence/divergence of clusters, and in their rotational motions. The fluctuations of the divergence and curl are most intense in the buffer layer, due to the stronger fluctuations of fluid velocity there. Similarities are identified between the cluster dynamics in the logarithmic region and those in homogeneous isotropic turbulence, including the dependence of divergence, curl, and helicity on Stokes number. The effect of increasing mass loading on cluster dynamics is relatively small except in the viscous sublayer, where attenuation of clustering, rotation, and swirling motions are observed. The effect of increasing Stokes number on the viscous sublayer is different, resulting in more intense convergence/divergence and rotation of particle clusters, as the particles become more independent of the carrier fluid.