Validation of the pyramid tracing algorithm for sound propagation outdoors: comparison with experimental measurements and with the ISO–DIS 9613 standards

Abstract

The aim of this work is to test the accuracy of numerical previsions of the sound propagation outdoors made with the newly developed pyramid tracing algorithm (Farina A, RAMSETE—a new pyramid tracer for medium and large scale acoustic problems, Proceedings of the Euro-Noise 95, Lyon, France 21–23 March 1995, 1995;1:55.; Farina A, Pyramid tracing vs. ray tracing for the simulation of sound propagation in large rooms, in COMACO'95, Proceeding of the International Conference on Computational Acoustics and its Environmental Applications, Southampton, England, 1995, Computational Mechanics Publications:Southampton 1995: 109.) Pyramid tracing has already been proposed and validated for indoor calculation, both for the study of the sound quality in concert halls or other Sabinian spaces, and for noise reduction in factories or other non Sabinian spaces (Farina A, Verification of the accuracy of the pyramid tracing algorithm by comparison with experimental measurements of objective acoustic parameters, Proceedings of the ICA'95, Trondheim, Norway, 26–30 June 1995, 2;1995:445.; Farina A, Auralization software for the evaluation of the results obtained by a pyramid tracing code: results of subjective listening tests, Proceedings of the ICA'95, Trondheim, Norway, 26–30 June 1995, 1995;2:441.) As the algorithm allows for the evaluation of sound passing through sound insulating panels, taking also into account the edge diffraction over free boundaries of screens, it is natural that it can be used also for the simulation of sound propagation outdoors. Anyway it must be noted that the actual implementation of pyramid tracing, available in the Ramsete package, does not take into account the interference effects caused by propagation at grazing incidence over the soil, neither the ray curvature caused by temperature or air velocity vertical gradients. In this paper a comparative experiment is described, in which the pyramid tracing code was tested against experimental measurements (made with an innovative Maximum Length Sequence signal, previously used only for room acoustics). The sound source was a directive loudspeaker, which Sound Power Levels and Directivity Balloons in Octave Bands were previously measured in free field conditions. The measurement in each point was obtained through asynchronous cross-correlation of the signal coming from a standard Sound Level Meter (recorded for convenience on a DAT tape recorder) and the original MLS sequence, through a fast-Hadamard algorithm, yielding the Impulse Response between the Source and the Receiver positions. With proper synchronous averaging of the incoming signal, a great improvement in the signal-to-noise ratio was achieved, making it possible to make measurements almost not affected by background noise even in highly shielded positions. The comparison is made also with an Image Source code, built up around the computing formulas contained in the new ISO–DIS standard 9613, kindly made available by Luigi Maffei (Farina A, Maffei L, Sound propagation outdoor: comparison between numerical previsions and experimental results, in COMACO'95, Proceedings of the International Conference on Computational Acoustics and its Environmental Applications, Southampton, England, 1995, Computational Mechanics Publications: Southampton 1995:55). The test case was chosen in an area containing all the most interesting acoustic phenomena: large distance propagation over absorbing and reflecting soil, shielding by embankments and buildings, multiple reflections on buildings facades. Only adverse atmospheric conditions were not taken into account (strong wind, inverted temperature gradient). Both the experimental and numerical data were used to build graphical plots, enabling a direct comparison of the results: they show that the capability of accurately modeling the source directivity produce generally a better estimate using the pyramid tracing algorithm, but the shielding effects and excess attenuation are more accurately modeled by the ISO9613 code.