A PIV-Based Algorithm for Simultaneous Determination of Multiple Velocity Fields from Stratified Crossflows in Single Field of View

Abstract

This study presents a new imaging-based algorithm for simultaneously determining multiple velocity fields from stratified crossflows optically captured in a single field of view. The concept implements an additional automatic peak finding scheme into the conventional particle image velocimetry (PIV) analysis procedure, identifying multiple prominent peak cross-correlation coefficients corresponding to the flows in various directions. To examine the validity, synthetic particle images generated by computer visions and image data acquired by PIV measurements are employed in the validation study. With both root-mean-square errors (RMSEs) in magnitude and direction being found to be temporally random, the validation results suggest that the performance of the new algorithm is ideal for steady or quasi-steady flows. This implies that the new algorithm may also work well for the flows repeatable with identical initial and boundary conditions. For transient flows, more valuable data can be obtained with the new algorithm, particularly in large-scale experiments or field measurements. Moreover, tests on synthetic images show that the RMSE in magnitude decays exponentially with increasing tracking particle density, and a density of 30% is found to be the lowest for the minimum RMSE in magnitude. Discussions on the error reduction, limitations of the new algorithm, suggestions for applications, and guidance on spurious vector removal are given as well.