Improvement in the independence of relaxation method-based particle tracking velocimetry

Abstract

New techniques are developed to improve the independence of relaxation method-based particle tracking velocimetry (RM-PTV). Firstly, Delaunay tessellation (DT) is employed to form clusters of neighboring particles with similar motion in the same frame; and then a bidirectional calculation concept is adopted to improve the way of particle pairing. These new techniques are tested with both self-defined particle images and the particle image velocimetry standard synthetic particle images. The results indicate that the DT method performs well and efficiently in determining the particle clusters, and the particle pairing process is well optimized by the bidirectional calculation concept. With these methods, three computation parameters are eliminated, which makes RM-PTV more autonomous in applications.