Elimination of discrete phase outliers in DT-PTV by preferential probability

A new method called elimination by preferential probability (EPP) to remove discrete phase outliers in Delaunay tessellation-particle tracking velocimetry (DT-PTV) was proposed. In the present method, the modules to eliminate redundant matches and the inverse calculation to enlarge matching information were adopted and a multiple-way in outliers elimination used in discrete phase flow field was developed on the basis of the double-layer matching pattern in DT-PTV and the combination modes of modules and inverse calculation. From the application in self-defined flow fields and comparison of accuracy rate, matching information maximizing in DT triangle level followed by most-suitable-matches selection in particle level is an optimum way of outlier elimination in discrete phase with an accuracy rate of about 92% in a low spatial rate (187.215). In addition, the DT-PTV combined with EPP was used for velocity measurement of sand phase in saltation layer of wind-sand flow, and the results of single-grain paths abstracted can meet the requirement.