Flow regimes detection in a quasi-2D granular chute flow based on PTV technique

An improved PTV technique for understanding the mechanical behaviors of each flow regime in granular chute flows at particle scale has been proposed. The PTV technique mainly includes two parts, particle recognition and particle matching. For particle recognition, the light intensity detection plate (LIDP) was introduced to detect the minimum light intensity of particles which was used to separate the sidewall and internal particles of granular flows; these particle centroids were initially estimated by the light spot method (LSM) in granular flows and then corrected by combining the image processing method (IPM) and the inverse distance weighting method (IDWM). For particle matching, the quasi-rigidity threshold (Tq) in particle matching step of PTV algorithm were determined by layer by layer incremental-iteration method. An example analysis of three physical model experiments was used to demonstrate the capabilities of the improved PTV. The estimated particle centroids are very close to the geometric centroids of the particles; the total recovery ratio are up to 97.3% for 28°, 96.4% for 32° and 94.3% for 36° flows. Moreover, the averaged multiple collision ratio, the granular temperature and the linear strain of flow layers were used to determine which flow regime each flow layer belongs to. Finally, these detection results of flow regime were verified by the critical state soil mechanics (for solid-like flow regime) and kinetic theory (for gas-like flow regime).