Abstract
The Particle Image Velocimetry (plV) flow measu rement technique is a very efficient technique for studying the structure of various fluid flows. It provides quantitative and qualitative full-field velocity information which can be analyzed to find the flow's velocity field, vorticity components, turbulent intensities, etc. The measurement technique relies on fast and efficient methods to accurately track images of neutral density particles which have been suspended in the flow. Since a large quantity of data has to be analyzed, the. tracking process must be fast and reliable. A new tracking method, known as the Genetic Algorithm, was developed for the two-dimensional PIV technique. This method utilizes a combinatorial approach to identify corresponding particles between sequential frames. The new tracking method was compared with three other existing tracking methods to evaluate their efficiency, i.e., reliability and yield. Synthetic data from a Large Eddy Simulation (LES) computational fluid dynamic code (GUST) were generated. All four methods were used to track the synthetic data. Then the simulated which are determine vectors were compared with the reconstructed vectors, the tracking, results of four tracking techniques, to the yield and reliability.
Yoon, Churl (1996). Development of a genetic algorithm tracking technique for the particle image velocimetry and comparison with other tracking models. Master's thesis, Texas A&M University. Available electronically from
https : / /hdl .handle .net /1969 .1 /ETD -TAMU -1996 -THESIS -Y66.