Rotorcraft Computational Methods
In the spring of 2009 a UH-60 rotor equipped with a unique control system was tested in the 40- by 80-foot wind tunnel at the NASA Ames Research Center1. This control system is called Individual Blade Control (IBC). IBC is unique in its ability to inject high-frequency pitch inputs (>1 pitch oscillation per revolution) and to tailor these pitch inputs for each blade in a rotor. IBC has the potential to improve performance and to reduce vibration, structural loads, and radiated noise.
Rotorcraft designers need high-accuracy methods for predicting far-field noise. They also need new techniques for analyzing the computed results. This project seeks to develop new parallel computer algorithms for rotor noise prediction, plus audio and visual rendering of the computed results.
Civilian helicopters and tiltrotors that operate in urban areas must have low noise if they are to meet with public approval. In order to achieve these low noise designs, rotorcraft designers need high-accuracy methods for predicting far-field noise. Euler and Navier-Stokes CFD methods provide excellent models for the near-field nonlinear acoustic propagation, however, near-field CFD grids cannot be extended to the far-field without excessive computational cost. Linear propagation methods such as Kirchhoff integrals are much more economical in the far-field, but fail to accurately model the near-field nonlinearities. An excellent compromise is to use the nonlinear CFD methods in the near-field and Kirchhoff methods in the far-field. This research aims to combine these two approaches into an accurate and efficient acoustics prediction scheme.
Future rotorcraft must have low noise if they are to operate near heavily populated areas. This is particularly true for tiltrotors as they descend for landings. This project aims to predict the aerodynamics and acoustics of rotorcraft by solving the Navier-Stokes equations on a series of overset computational grids that surround the rotor blades.
Unstructured grid for solving problems in computational fluid dynamics have two major advantages over their structured-grid counterparts. First, the unstructured mesh allows for fast and efficient grid generation around highly complex geometries. Second, appropriate unstructured-grid data structures facilitate the insertion and deletion of points and enable the computational mesh to locally adapt to the flowfield solution.
Two approaches are typically taken to solve computational fluid dynamics problems with complicated 3-D geometries and flow features. The first uses unstructured solution-adaptive grids. These methods are well suited for solution-adaption to resolve flow features, but grid generation difficulties and large computer memory requirements tend to complicate applications for large 3-D viscous flowfields. The second approach uses a series of overset structured grids to model the flowfield. Each structured grid is usually well formed and implicit time-marching techniques result in computationally-efficient solutions. A problem with this approach, however, is that structured grid methods are not well suited for unsteady grid adaption in three dimensions.