POD analysis of the unsteady behavior of a laminar separation bubble

• A laminar separation bubble has been investigated with PIV and POD analysis.
• POD identifies the vortex shedding induced by Kelvin–Helmholtz instability.
• The POD eigenvectors are used to sort and phase-average the experimental snapshots.
• The results are used to separate out the deterministic from the stochastic dynamics.

Particle Image Velocimetry (PIV) measurements have been performed in order to analyze the unsteady flow field developing along the separated flow region of a laminar separation bubble. Data have been post-processed by means of Proper Orthogonal Decomposition (POD) to improve the understanding of the physics of this complex phenomenon. The paper shows that the first two POD modes of the normal to the wall velocity component are coupled. Thus, they are representative of a vortex shedding phenomenon which is identified to be induced by Kelvin–Helmholtz instability. The POD allows the phase identification of each PIV image within the vortex shedding cycle. The computed eigenvectors are used to sort the experimental snapshots and then reconstruct a phase-averaged velocity field which highlighted the motion of vortices shed close to the bubble maximum displacement. Moreover, other sources of deterministic fluctuations characterized by frequencies which are different from the one induced by the Kelvin–Helmholtz instability are also revealed. Indeed, the most energetic POD mode of the streamwise velocity component is not related to the shedding frequency, while it describes large velocity fluctuations in the shear layer region upstream of the bubble maximum displacement, where the turbulent activity is not yet present. The POD decomposition presented here identifies the large scale structures within the flow, thus separately accounts for both coherent and stochastic contributions to the overall energy of the velocity fluctuations.