Application of Radiosity Methods to the Prediction of Sound Fields in Diffusive Environments

University Of Colorado At Boulder, Boulder CO

Investigators

Abstract

Most architectural acoustic prediction programs utilize ray-tracing techniques that work best for environments with large reflecting surfaces such as auditoria and concert halls. This project will apply the radiosity method that has been developed for illumination and visual computer rendering to the prediction of sound fields in diffuse acoustic environments. The new prediction method will allow for prediction of sound levels in small complicated environments such as open plan offices and factories where ray tracing techniques fail. The decreased computational requirement of the method may allow acoustics to be added to the visual architectural "walk-through" of office and factory designs. In the radiosity method a set of energy balance equations between the bounding surfaces of the room is developed. Most of these equations are independent of the source and receiver positions and thus acoustic prediction on multiple source and receiver locations can be significantly faster than with ray tracing techniques. In the first phase of this project, the acoustic radiosity method is developed assuming perfectly diffuse surfaces and radiators and ignoring diffraction effects. In the second phase the method is extended to include non-diffuse surfaces. In the third phase diffraction effects are added to the model. Students working on the project will present their results at national conferences and through publications in peer reviewed acoustic journals. Results of the work will be incorporated into existing classes on building noise control and room acoustics where students will develop their own radiosity programs for acoustic prediction.