Initial Investigations into the Failure Modes of a Swirl Distortion Generator Using Computational Methods

The need for more efficient and environmentally sustainable aircraft has been a rapidly increasing topic for research and development over the last few decades. Within this area of research, boundary layer ingestion (BLI) concepts have been developed which integrates the airframe and propulsion system of an aircraft. In turn, BLI increases the fuel efficiency and decreases emissions by reducing the overall drag and reenergizing the aircraft wake. However, the boundary layer flow of an airframe or duct can impose undesired flow conditions, such as swirl and pressure distortions, at the inlet of a jet engine. Therefore, efficient research and testing capabilities are essential to advance the development of these integrated systems.

The StreamVane swirl distortion generator was developed by Virginia Tech to provide cost and time efficient ground testing methods for BLI research. StreamVanes are constructed of unique vane packs that are specifically tailored to generate a desired swirl distortion profile. To maximize efficiency, StreamVanes are additive manufactured which cause geometry limitations to the overall vane design. Due to these restrictions, as well as the complexity of the vane pack, unwanted dynamic responses and unsteady flows can be generated. In order to predict both of these phenomena before testing, two different computational methodologies were developed and investigated on a StreamVane and its airfoil parameters.

First, a one-way fluid-structure interaction methodology was developed to predict flutter mconditions of the vanes within StreamVanes. The presented methodology includes steady and unsteady computational fluid dynamics (CFD) as well as linear structural and modal finite element analysis (FEA) simulations. A simplified StreamVane model was designed as a testcase for the methodology, and it was found that two unique vane shapes did not undergo flutter conditions at three different operating points. The results provided a linear analysis method to compute the aerodynamic damping, which gave insight on how different vane shapes respond dynamically.

Secondly, a parameter study was conducted to predict the vortex shedding from the modified NACA 63-series airfoil profile used within StreamVane design. The effects of the airfoil turning angle and trailing edge thickness on the vortex shedding frequency were computationally predicted using the unsteady Reynolds averaged Navier-Stokes equations (URANS) and shear stress transport (SST) turbulence model. In turn, the shedding frequencies for each parameter were recorded, and more intuition was gained on the TE flow field in correspondence to different airfoil specifications. Overall, the two sets of methodologies and results can be used to efficiently reduce failure uncertainties in future StreamVane designs.