A Computational and Experimental Analysis of Vortex Shedding from Complex Turning Vanes

Abstract

This paper presents a comprehensive analysis of vortex shedding from two sets of complex turning vanes. The vane packs, known as StreamVanes, are swirl distortion generating devices capable of reproducing distorted secondary flow profiles to a high degree of accuracy. Due to manufacturing and structural durability requirements, the turning vanes within StreamVanes consist of airfoil profiles with thick trailing edges. This, in turn, introduces fixed separation points at the pressure and suction sides of the trailing edge which inherently produce shear layer instabilities and vortex shedding within the wake. The objective of this research was to predict and measure the wake structures and vortex shedding frequencies produced by different vane parameters within two unique StreamVane models. A commercial, unsteady Reynolds-Averaged Navier-Stokes (URANS) code was used to acquire wake-flow visualizations and frequency data generated by both vane packs. Time-resolved particle image velocimetry (TR-PIV) was conducted to measure both aspects of vortex shedding and quantify the differences from the computational fluid dynamics (CFD) predictions. A two-point space-time correlation method was developed to accurately extract the measured shedding frequencies from high-speed wakes that exceeded the Nyquist limit. It was found that the URANS code was capable of qualitatively capturing coherent wake structures and three-dimensional eddy distortions within the vane wakes. The TR-PIV measurements revealed differences from the URANS predicted non-dimensional shedding frequencies within the range of 0.05 - 14%. The presented results provide insight into the accuracy of URANS codes in predicting vortex shedding features for a combination of three-dimensional flows, blunt trailing edge airfoils, and complex turning vanes.