关键词: 特征点匹配, 表面流场, 移动式, 流场测量, 图像处理, 相干结构

Abstract: A flume experiment of flows around a bridge pier and bed load transport is conducted to analyze the turbulent structures using an innovative method of surface flow field measurement. In this experiment, move shooting is carried out with a high-speed camera moving along the flume, and the images are clipped out through intelligent identification of the coordinates of ground control points. The technologies of particle tracking velocimetry (PTV) and feature matching velocimetry (FMV) are both used to calculate the instantaneous flows, and then a large area of surface flow field is obtained through meshing and interpolation in the measurement region. The results show (1) this method improves the accuracy and dimensions of surface flow field measurement and broadens its application condition. (2) The FMV method can identify more accurately the flow details in the zones of high rotation intensity beside and behind the cylinder, and its measurements of the average velocity profiles are more consistent with the flow characteristics. (3) We have observed that during bedload movement, the time-average velocity on the turbulent surface still features a high-low alternating streaky structure, with the high and low velocity streaks corresponding to the sand ridge and sand trough respectively. This proves the turbulent flow and the bed form interact to form a stable time-average structure, and also verifies the rationality of the large-scale streamwise vortex model.

Key words: feature point matching, surface flow field, camera mobile shooting, flow field measurement, image processing, coherent structure