Prof. Stefano Discetti

is Associate Professor in the Aerospace Engineering Research group at Universidad Carlos III de Madrid. He received his B.S. (2007), M.Sc. (2009) and Ph.D. (2013) in Aerospace Engineering from University of Naples Federico II. During his Ph.D. he has been guest researcher at Arizona State University in 2012. His research interests include development and application of advanced flow diagnostic techniques for the experimental investigation of turbulent flows. He served as test case provider and referee in the team of the 4th International PIV Challenge. He is member of the Editorial board of Measurement Science and Technology, and of the Scientific Committee of the International Symposium on Particle Image Velocimetry.


Plenary Talk Title

Enhancing PIV via data-driven techniques



Data processing techniques for PIV and PTV have reached a mature state, with capability to achieve low uncertainty at scales comparable to the inter-particle distance, or even beyond when time resolution is available. This leads to the perception that PIV limits are now prescribed mainly by hardware limitations, i.e. resolution of camera sensors and repetition rate of illumination/recording hardware.
Owing to the increasing size of PIV datasets, data science methods applied on PIV data are expected to open exciting scenarios in the next decade. In this talk, an overview of data-driven techniques to enhance the current capabilities of PIV and PTV will be presented.
Temporal resolution limits can be overcome with a combination of low-repetition rate PIV with fast-recording probe data; an application to high-Reynolds-number turbulent pipe flow experiments in CICLOPE will be illustrated. A technique to push the spatial resolution of PIV/PTV beyond the inter-particle distance limit will also be presented. The method leverages on Proper Orthogonal Decomposition to extend to instantaneous flow fields the ensemble-averaging principle widely used to achieve high-resolution turbulent statistics.