The effect of wind turbulence on noise barrier performance
An investigation has been carried out into the effect of wind turbulence on the propagation of sound over open grassland both with, and without, a 2.4m high barrier present. Sound bursts, derived from a variety of original signals, were generated using a horn loudspeaker and the resulting levels were measured using two variably positioned microphones, the further of which was placed at distances of up to 24m from the source. Simultaneous measurements were made of a variety of meteorological parameters. A microcomputer-based system was developed to control the experiments-and store the measured data for subsequent retrieval and analysis. As one of the approaches adopted in a search for visible evidence of correlation between various acoustical and meteorological parameters, the measured level difference for each sound burst was displayed graphically against the corresponding value of a particular measure of wind turbulence. For many combinations of measurement geometry and acoustic signal type, the data points fell, with a noticeable degree of consistency, within an envelope of characteristic shape. The shape implied that when the instant of transmission of a sound burst coincided with a low value of the measure of local turbulence, the apparent level difference was subject to considerable statistical fluctuation. This was the case both with and without the barrier present. As the turbulence increased, however, the propagating medium appeared to behave as a progressively more uniform and stable one and the observed scatter correspondingly reduced. The origin of this behaviour remains unclear; however the presence of the barrier, it was concluded, did not appear to modify this effect. 1 to 16 laboratory scale model experiments have been carried out to investigate the replication of the above effects and the results are reported.