Spatial detection and hierarchy analysis of large-scale particle clusters in wall-bounded turbulence

Abstract

Large-scale inertial particle clustering in a turbulent boundary layer (TBL) was investigated experimentally by two-colour particle-image velocimetry and particle-tracking velocimetry measurements in the wall-parallel plane of the log layer. The particle clusters were detected by Voronoi tessellation and a new application of a technique based on the spatial wavelet transform. Anisotropic wavelets spanning the range of measurable, large-scale particle clusters were applied to experimentally measured particle fields, and significance testing was used to identify hierarchies of particle clusters. Unlike the Voronoi tessellation, which primarily identified small-scale clusters, the wavelet technique also found a significant amount of anisotropic, large-scale clusters on the scale of the large-scale motions (LSMs) of the momentum field. The large-scale clusters were shown to be composed of a hierarchy of smaller clusters of varying degrees of isotropy. In addition to the cluster size distribution, the particle velocities within the clusters were studied as a function of cluster size. Larger clusters exhibited faster particle convection speeds, consistent with observations of the convection speeds of LSMs, thus providing additional evidence for the hypothesis that particles in the TBL accumulate on and are transported by large-scale, attached coherent structures.