Experimental Validation of a Computational Tool for Assessing Sound Propagation in Closed Environments

Abstract

It is common, in the analysis of problems involving sound propagation in rooms, to use simulation programs in order to obtain predictions of the results obtained with the proposed solutions. How-ever, existing methodologies may present high computational cost or inaccurate results for small environments. In this way, it was proposed in this work the development of a computational tool capable of carrying out simulations of sound propagation in closed rooms with accurate results and lower computational cost. The methodology chosen was the Digital Waveguide Mesh. The formulation of this method was presented, in addition to the implementation of the following techniques: mesh interpolation – in order to reduce dispersion errors and computational cost, and more precise formulations of anechoic boundary conditions or those with frequency-dependent properties. Simulations and experiments were developed using a room in different configurations to validate the results provided by the developed tool. After implementing the code with the solutions for reducing the computational cost and more accurate acoustic modeling, the results were considered experimentally validated, allowing the tool to be characterized as accurate, in addition to presenting shorter calculation times and not requiring the allocation of larger amounts. of memory to store the matrices.