Computational aeroacoustics of ducts using parallel computers

Abstract

This paper will discuss recent attempts at using computational aeroacoustics (CAA) algorithms on parallel computers to solve duct propagation problems. The algorithms are higher-order accurate methods for solving the Euler and Navier–Stokes equations. As computer speed increases, more and more of the physics can be simulated. This talk will discuss the current status and trends of parallel supercomputers. While hardware trends allow us to solve more complicated problems, algorithm developments are also essential. This talk will address some new methods for solving aeroacoustics problems. Most of these new developments involve hybrid schemes that couple CAA methods with integral methods or traditional CFD codes with CAA codes. Since aeroacoustics problems involve length and time scales that are beyond computer power at the present time, clever schemes must be developed for using appropriate algorithms for the various physical phenomena. One very promising scheme [J. Comput. Phys. 133 (1997)] is the nonlinear disturbance equation (NLDE) method, which uses a traditional CFD code to predict the steady-state portion of the problem, and then a higher-order method to solve for the disturbances. The steady-state portion is included in the coefficients and source terms in the NLDE. [Work supported by NASA Langley.]