On a combined adaptive tetrahedral tracing and edge diffraction model

a combined adaptive tetrahedral tracing and edge diffraction model" (2014). A major challenge in architectural acoustics is the unification of diffraction models and geometric acoustics. For example, geometric acoustics is insufficient to quantify the scattering characteristics of acoustic diffusors. Typically the time-independent boundary element method (BEM) is the method of choice. In contrast, time-domain computations are of interest for characterizing both the spatial and temporal scattering characteristics of acoustic diffusors. Hence, a method is sought that predicts acoustic scattering in the time-domain. A prediction method, which combines an advanced image source method and an edge diffraction model, is investigated for the prediction of time-domain scattering. Adaptive tetrahedral tracing is an advanced image source method that generates image sources through an adaptive process. Propagating tetrahedral beams adapt to ensonified geometry mapping the geometric sound field in space and along boundaries. The edge diffraction model interfaces with the adaptive tetrahedral tracing process by the transfer of edge geometry and visibility information. Scattering is quantified as the contribution of secondary sources along a single or multiple interacting edges. Accounting for a finite number of diffraction permutations approximates the scattered sound field. Superposition of the geometric and scattered sound fields results in a synthesized impulse response between a source and a receiver. Evaluation of the prediction technique involves numerical verification and numerical validation. Numerical verification is based upon a comparison with analytic and numerical (BEM) solutions for scattering geometries. Good agreement is shown for the selected scattering geometries. Numerical validation is based upon experimentally determined scattered impulse responses of acoustic diffusors. Experimental data suggests that the predictive model is appropriate for high-frequency predictions. For the experimental determination of the scattered impulse response the merits of a maximum length sequence (MLS) versus a logarithmic swept-sine (LSS) are compared and contrasted. It is shown that a LSS is an appropriate stimuli for testing acoustic diffusors by comparing against scattered relative levels measured by a MLS signal. Acknowledgments I wish to express my sincere gratitude to several people that made this research a possibility. First, I thank Siu-Kit Lau for his patient mentoring and his tremendous support during my research. His guidance and research philosophy has shaped to a large extent my perspective on conducting research. The camaraderie and support fostered by the Nebraska Acoustics Group is due to Lily Wang. The group has served each member in many ways professionally. Much appreciation goes to Peter D'Antonio …