Comparison of sound field measurements and predictions in coupled volumes between numerical methods and scale model measurements

Abstract

Prediction of sound fields in closed spaces can be achieved by various methods, either physical or numerical, based on different theoretical features. While the benefits and limitations of many methods have been examined for single volume spaces, there has been little effort in examining these effects for coupled volume situations. The present study presents a case study comparing theoretical, experimentally physical measurements on a scale model, and various numerical methods, namely boundary element method (BEM), finite-difference time-domain (FDTD) and ray-tracing through the commercial software CATT-Acoustic and ODEON. Although these numerical methods all use 3D numerical models of the architecture, each is different. Ray-tracing is more suitable to geometries with larger planes; BEM requires a more regular finer surface mesh; and FDTD requires a volumetric mesh of the propagation medium. A simple common geometry based on the scale model is used as a basis to compare these different approaches. Application to coupled spaces raises issues linked to later parts in the decay due to multi-slope decay rates, as well as diffraction phenomenon due to acoustic energy travelling between coupling surfaces from one volume to another. The ability of these numerical methods to adequately model these effects is the question under study.