Midfrequency acoustic radiation analysis.

Abstract

Boundary, finite, and infinite element methods have proven useful for solving the 3-D acoustic wave equation at low frequencies. For high frequency, localized formulations such as plane-wave and ray-tracing methods have been applied successfully. These methods are impractical, inaccurate, or both for treating midfrequency problems. Low-frequency methods make very large demands on computer processing while high-frequency methods account only for local effects. Full-spectrum methods have been developed which apply interpolation between the low and high ends of the frequency range to estimate the response in the midfrequency. While these methods have provided some insight to trends as one moves toward the midfrequency range, it is not well established that they can reliably solve problems well away from their established limits. For surfaces with features having dimensions close to midfrequency wavelengths, the physics is distinct from those that govern low and high frequencies, and is not merely a linear combination of the two. This presentation reports on research to develop a midfrequency analysis tool that directly addresses the midfrequency physics. The author will identify the challenges in a physics-based method, discuss an approach to address them, and report on simulations exercising the approach.