One of the objectives of the NEXTFLOW project is the development of novel concepts to increase the spatial resolution of PIV, going beyond the spatial “Nyquist limit” set by the average interparticle spacing. Within this line, the NEXTFLOW team recently published the work Tirelli, I., Ianiro, A., & Discetti, S. (2023). An end-to-end KNN-based PTV approach for high-resolution measurements and uncertainty quantification. Experimental Thermal and Fluid Science, 140, 110756.
In this contribution, we aim to virtually reduce the interparticle spacing by merging similar snapshots. The main concept is that particles randomly sample in space the velocity fields; if we are able to identify velocity fields that look similar, we can merge the particles from different snapshots and thus increase the density.
The topological similarity of snapshots is evaluated with a K-Nearest Neighbour (KNN), carried out on a space of significant flow features built with Proper Orthogonal Decomposition. This analysis is performed in subdomains, thus allowing to target local similarity when merging snapshots. The method is tested and validated against datasets with a progressively increasing level of complexity (virtual and real experiments) where it exhibits a strong improvement in the spatial resolution.
The final result is an end-to-end tool that is able, starting from raw images, to obtain high-resolution flow fields without the need for particular expertise from the user, or of advanced computational resources.
The work has been presented at “20th International Symposium on Application of Laser and Imaging Techniques to Fluid Mechanics” in Lisbon and published in Experimental Thermal and Fluid Science. The full paper is also accessible through Arxiv portal, the code is available on GitHub and the datasets on Zenodo platform.
The work is part of the PhD work of Iacopo Tirelli, and coauthored by Andrea Ianiro and Stefano Discetti, PI of this project.
