The paper analyzes the unsteady interaction between an impinging wake and the suction side boundary layer of a low pressure turbine cascade operating under unsteady inflow conditions. The instantaneous flow fields are obtained with Particle Image Velocimetry (PIV) for two different free-stream turbulence intensities (FSTI) and two orthogonal measurement planes (a blade-to-blade and a wall-parallel plane, respectively). The large data set has been post-processed by means of Proper Orthogonal Decomposition (POD) to provide a solid statistical interpretation of the physics of this complex unsteady interaction.
For the analysis of the blade-to-blade plane measurements, the paper takes advantage of the POD properties, which consist in splitting spatial (POD modes) and temporal (POD eigenvector) information. The phase identification of each PIV image within the wake passing period is then extracted from the eigenvectors of properly selected POD modes in order to reconstruct a phase-averaged velocity field. This post-processing allow us to separately account for the contribution to deterministic and stochastic unsteadiness as well as to have a look at the most significative scales of turbulence generating during the different phase of the wake-boundary layer interaction process.
The detailed analysis is completed by the measurements in a wall-parallel plane (aligned with the blade span). Instantaneous images and POD modes in this plane allow recognizing the presence of three-dimensional disturbances, induced by low and high speed travelling streaks (Klebanoff mode). These coherent structures have been shown to play a significant role in the unsteady transition process of the boundary layer at both low and elevated FSTI, and results reported in the paper further characterize their role in the transition process.