Error propagation in the procedure of pressure reconstruction based on PIV data

Abstract

Pressure reconstruction based on particle image velocimetry (PIV) has become a popular technique in experimental fluid mechanics. Errors in the raw velocity field significantly affect the accuracy of pressure gradient field and further reduce the quality of reconstructed pressure. Thus, error propagation deserves a serious concern. The format and magnitude of errors are investigated using a probability density function (PDF) based method. Theoretical derivation and numerical validation are operated in both Eulerian and Lagrangian descriptions. The influence of spatial and temporal resolutions on error propagation is discussed. A criterion of parameter selection for error suppression in the Lagrangian method is proposed. The results show that error propagation in the Lagrangian method has definite format and magnitude which makes this method more suitable for error control through specific treatment. Time interval during pressure reconstruction should be carefully determined since a large uncertainty appears when the time interval is employed improperly.