The Design and Analysis of Bluff Body Dynamic Distortion Generators

Abstract

With the increased use of serpentine diffusers in aircraft comes an increase in secondary flows and large pressure gradients which decrease engine efficiency. Therefore, in order to assess new engine technology, a way to generate these unsteady flows in ground test facilities needs to be developed. StreamVanes and ScreenVanes were developed at Virginia Tech as a way to generate real-world swirl and pressure distortion profiles for engine testing. However, they are only capable of generating steady distortion patterns, which limits the scope of inlet flow patterns that can be produced to evaluate various real-world scenarios. Therefore, this research study proposes the next generation of StreamVanes which are capable of generating dynamic distortion patterns in order to test the latest advances in aircraft engines. Serpentine diffusers create unsteady flow in the form of counter-rotating vortices at the top dead center of the inlet. The frequencies associated with these dynamic distortion features that cause the most harm to the system are typically lower, on the scale of around one engine order. Knowing that, the goal of this research was to design a passive method to replicate this type of real-world flow based on existing literature on unsteady distortion and profile generation with StreamVanes. After extensive literature search, bluff bodies were identified as well-known geometries that create unsteady flow features, namely in the form of periodic, paired vortex shedding. Cylindrical bluff bodies create vortex streets that shed at a known frequency determined by their Strouhal number. BeVERLI Hill is a geometry that has been extensively studied by Virginia Tech due to its unique, asymmetric vortex shedding that alternates chaotically off the back of the hill. Both of these types of bluff bodies were integrated with a base twin-swirl StreamVane design and were predicted to cause unsteadiness similar to the vortex pairs found in serpentine inlets. The designs underwent both computational tests and experimental tests in order to determine how they interact with the steady StreamVane flow pattern, how they scale with sizing and Mach number, and the frequency content of the unsteady flow features. RANS/URANS CFD, flow visualization, and hot film anemometry were used to test the Dynamic StreamVane designs and gave results that provided a proof-of concept for this type of passive, unsteady distortion generator. From these results, it was found that the vortices were shed similarly to how they would if the bluff bodies were on their own, but as they continued downstream, they were carried in the direction of the flow from the twin swirl StreamVane and dissipated in their intensity past a diameter or two downstream. Additionally, there was evidence of the frequencies scaling with Mach number and bluff body size that aligned with expected relationships. Contrary to predictions, the frequency content did not indicate strong peaks at any particular frequencies. However, the majority of the energy was focused in the low frequency range, as desired. Further and more extensive testing of the designs is needed to fully understand how they can be used to generate specific dynamic distortion profiles at a given AIP, but this research provides a solid baseline from which to further develop the Dynamic StreamVane technology.