Super-resolution PIV study of the flow behind square-shaped and triangular-shaped bluff bodies.

The paper presents the results and analyses of two cases of the flow behind the square-shaped and triangular-shaped bluff bodies under a moderate Reynolds number (Re=5000). Experimental values have been obtained with Particle Image Velocimetry (PIV) methods in an open wind tunnel. Two different PIV methods were used: traditional PIV based on cross-correlation (CC) algorithms and Artificial Intelligence PIV (AI PIV) based on Deep Learning and Convolutional Neural Networks, which offers super-resolution velocity fields with down to 1 vector per pixel. In addition, HWA measurements were conducted at a few points for verification and validation of the results.

These two selected flow cases differ in the nature of the downstream wake. Triangle bluff body wake is characterized by a higher vorticity and Strouhal Number (St) compared to that of a square body. This made it possible to assess the impact of the advanced PIV data post-processing methodology. It should be noted that AI PIV gives a more precise and a higher spatial resolution of the flow field so that we can see more details from the flow maps. Both applied methods demonstrated a predominance of Karman vortex shedding; however, with AI PIV, many small-scale vortices were also identified. Comparing successive snapshots made it possible to see, on the one hand, the process of formation and, on the other, the merging of smaller vortices, especially those at a further distance behind the object. The obtained results showed a more precise velocity field with a higher spatial resolution using AI PIV. The full paper will provide not only qualitative but also quantitative comparisons between the two methods.