Experimental uncertainties associated with particle image velocimetry (PIV) based vorticity algorithms

 We have recently used Particle Image Velocimetry (PIV) to study the dynamics of vortex propagation in reacting and non-reacting flows. In order to do so, it became necessary to assess the uncertainty in PIV-based vorticity data. A computer simulation was developed to investigate how uncertainty propagates throughout the post-processing, numerical data smoothing, and vorticity calculating algorithms commonly used in the analysis of PIV data. Results indicate that the average uncertainty in vorticity per interrogation cell (normalized to the average vorticity, and then surface averaged), for a simple vortex, can be reduced to approximately ±4% with appropriate measures. This value was obtained using PIV autocorrelation software, a local regression technique combined with a Gaussian-smoothing filter. Our best experimental results (these areas with no lost or spurious vectors) are consistent with Stoke’s theorem.