Acoustic Voxels: Computational Optimization of Modular Acoustic Filters

摘要

Acoustic filters have a wide range of applications, yet customizingrnthem with desired properties is difficult. Motivated by recentrnprogress in additive manufacturing that allows for fast prototypingrnof complex shapes, we present a computational approach that automatesrnthe design of acoustic filters with complex geometries. Inrnour approach, we construct an acoustic filter comprised of a set ofrnparameterized shape primitives, whose transmission matrices can bernprecomputed. Using an efficient method of simulating the transmissionrnmatrix of an assembly built from these underlying primitives,rnour method is able to optimize both the arrangement and the parametersrnof the acoustic shape primitives in order to satisfy target acousticrnproperties of the filter. We validate our results against industrialrnlaboratory measurements and high-quality off-line simulations. Werndemonstrate that our method enables a wide range of applicationsrnincluding muffler design, musical wind instrument prototyping, andrnencoding imperceptible acoustic information into everyday objects.