Predicting scattering at the mid-frequency range

Abstract

Predicting scattering characteristics of isolated surfaces is problematic in the mid-frequency range, due to high computational costs. In the mid-frequency range, techniques such as the finite element method or the boundary element method require extremely fine mesh structures, leading to the solution of very large matrices. A gap exists for predicting acoustic scattering in the mid-frequency range with low computational costs. A geometric acoustic time domain method, incorporating diffraction, is proposed to bridge the gap in current prediction methods. The geometric acoustic method utilized is adaptive beam tracing. Adaptive beam tracing is advantageous, compared to other geometric methods, since it does not generate aliasing errors and faulty image sources, which are shortcomings of ray tracing and the image source method, respectively. The scattering characteristics of a periodic rigid geometry are investigated. Relationships between geometry dimensions and frequency of excitation are studied. Furthermore, the importance of accounting for single diffraction or multiple diffractions will be discussed. The spatial, temporal, and frequency characteristics of the scattering geometry are quantified.