A practical approach to PIV uncertainty analysis

While PIV is an established experimental technique for determining a velocity field, quantifying the uncertainty related with this method remains a challenging task. The purpose of the current work is to provide a procedure for estimating the uncertainty associated with time-averaged two-dimensional flow field measurements obtained by means of PIV. Four sources of uncertainty are assessed: equipment, particle lag, sampling size, and processing algorithm. An example uncertainty analysis is conducted for a transverse sonic jet injected into a supersonic crossflow. This flow was selected due to its wide range of flow field phenomena (bow shock, Mach disk, barrel shock, shear layers, boundary layers, etc.). However, the analysis is not specific to the example flow field and may be generally applied to any mean velocity field. In more completely assessing the uncertainty of PIV in measuring a complicated flow field, it is proposed that tools can be developed to estimate the uncertainty throughout the image instead of just giving a single value.