Applications of Physics Based Data
Restoration from Tomographic PIV

Fulvio Scarano, Jan F.G. Schneiders


Abstract: Tomographic Particle Image Velocimetry (Tomo-PIV) has developed over a decade into a versatile tool for the inspection and visualization of complex three dimensional flows. The principal domain of application has been the study of fundamental turbulent flow phenomena occurring in boundary layers, wakes, jets including pressure evaluation and aeroacoustics.

The availability of measurements on an increased number of dimensions in space and time has paved the way for expansion of the techniques for data restoration. In particular to move from signal-processing based (e.g. spatiotemporal filters, proper orthogonal decomposition) to physics based methods. The latter assimilate measurements with the flow governing laws to yield data reconstruction conditioned towards physical consistency.

The lecture will provide a survey of such methods as they have appeared in the literature related to PIV and in relation to relevant applications with experiments. The techniques are categorized based on their objective:
1) Increasing temporal resolution (pouring Space into Time, time-supersampling);
2) Data regularization (e.g. solenoidal filtering, Lagrangian data filtering);
3) Enhancing spatial resolution (pouring Time into Space, VIC+, Shake-the-Box);
4) Pressure from PIV.

The working principles of the methods is briefly discussed, pointing the reader to the relevant literature for a detailed discussion. The impact of each technique is then discussed first by simplified cases that illustrate their potential and limitations. Examples from real-life experiments are then brought forward to illustrate the current level of applicability of these methods.