Experimental Analysis of Multiscale Vortex Shedding in Turbulent Turbomachine Blade Wakes

Abstract

Wakes behind turbomachine blades are generally highly turbulent and seemingly chaotic. Within the apparent chaos, coherent structures of various strength and length scales are present that are challenging to identify within individual snapshots of the wake flowfield. We propose a strategy to identify and separate multiscale vortices in the turbulent wake of a turbomachine blade from experimental data. Uncorrelated velocity field snapshots were obtained in the wake of a compressor blade in a linear cascade wind tunnel at Reynolds using particle image velocimetry. The snapshots were analyzed using proper orthogonal decomposition to identify the most energetic mode pairs representing vortex shedding patterns of different length scales with their own shedding frequency. The phase angles associated with the individual shedding patterns are extracted for each snapshot based on the time coefficients of the mode pairs. By averaging snapshots with the same phase angle for selected mode pairs, we were able to calculate mode- or scale-dependent phase average flowfields. These scale-dependent phase averages visually highlight the multiscale vortex character of the turbulent wake and allow for a quantitative analysis of the size, emerging location, and shedding frequency of the different structures.