A compact windscreen has been conceived for a microphone of a type used outdoors to detect atmospheric infrasound from a variety of natural and man-made sources. Wind at the microphone site contaminates received infrasonic signals (defined here as sounds having frequencies <20 Hz), because a microphone cannot distinguish between infrasonic pressures (which propagate at the speed of sound) and convective pressure fluctuations generated by wind turbulence. Hence, success in measurement of outdoor infrasound depends on effective screening of the microphone from the wind.

To be effective, an infrasonic windscreen must fulfill four basic requirements: (1) it must attenuate noise generated by ambient wind, (2) it must transmit infrasound propagating across the microphone, (3) it must be useable in all weather, and (4) it must not be susceptible to generation of infrasound through shedding of vortices.

Figure 1. These Plots Are Results of Tests of the wind-noise-attenuation and infrasound-transmission properties of a polyurethane-foam windscreen.
Past methods of wind screening include the use of cloth or open-cell foam, and the use of an array of pipes. A windscreen made of cloth or open-cell foam is thought to break up incident airflow into very small turbulent eddies that dissipate wind energy in the form of heat. Such a windscreen is effective at audio frequencies (>20 Hz) but not at infrasonic frequencies (<20 Hz).

An array of pipes used as a windscreen consists, more specifically, of several perforated pipes, called a "spider," fanning out radially from a microphone situated in an enclosed housing. The array is vast — covering an area comparable to that of an athletic field — and its performance as a windscreen is degraded by resonances that depend on the lengths of the pipes.

The present compact windscreen is based on an entirely different principle: that infrasound at sufficiently large wavelength can penetrate any barrier of practical thickness. Thus, a windscreen having solid, non-porous walls can block convected pressure fluctuations from the wind while transmitting infrasonic acoustic waves. The transmission coefficient depends strongly upon the ratio between the acoustic impedance of the windscreen and that of air. Several materials have been found to have impedance ratios that render them suitable for use in constructing walls that have practical thicknesses and are capable of high transmission of infrasound. These materials (with their impedance ratios in parentheses) are polyurethane foam (222), space-shuttle tile material (332), balsa (323), cedar (3,151), and pine (4,713).

A small wind tunnel was built to test the acoustical properties of a variety of windscreen materials. A fan generated wind at speeds up to 21 mph (9.4 m/s) across an infrasonic microphone. Tests were conducted with and without the windscreen; the difference in the noises detected in the presence and absence of the windscreen was used as a measure of the attenuation of wind noise by the windscreen.

The windscreen that performed best in the wind-tunnel tests was a cylinder made of polyurethane foam of a type known in the industry as "eightpounder," having an inside diameter of 3 in. (7.62 cm), a wall thickness of 0.5 in. (1.27 cm), and a length of 12 in. (30.48 cm). The attenuation of wind-generated noise was quantified as the ratio between the wind noises measured without and with this windscreen. The results, plotted in the upper part of Figure 1, show that this windscreen attenuated wind noise by amounts ranging from 12 to 20 dB at frequencies ranging from 0.7 to 20 Hz. The large spikes in the spectrum represent aeolian tones generated by the wind passing over the windscreen, but these lie above the infrasonic range.

Figure 2. A Cylindrical Windscreen Covers a Microphone mounted on a pole outdoors.
For measurements of the infrasound transmission coefficient of this windscreen, a subwoofer was placed at an end of the wind tunnel and used to generate a tone that was swept over the frequency band from 10 to 200 Hz. In this case, the ratio between detected sounds, with and without the windscreen, was taken as a measure of the transmission through the windscreen. The results for the portion of the spectrum from 10 to 100 Hz, plotted in the lower part of Figure 1, show that this windscreen had a large transmission coefficient at frequencies below 25 Hz, even exhibiting a gain as high as 8 dB at 10 Hz, but then attenuated sound at higher frequencies. Finally, a soak test revealed that the water absorbed by the polyurethane windscreen material amounted to only 2.1 percent by weight.

Figure 2 shows a windscreen installed over a microphone mounted on a pole in the field. The windscreen has proved robust in weather conditions of all seasons and it survived Hurricane Isabel with wind gusts up to 67 mph (30 m/s).

This work was done by Allan J. Zuckerwar, Qamar A. Shams, Bradley S. Sealey, and Toby Comeaux of Langley Research Center. For further information, contact the Langley Innovative Partnerships Office at (757) 864-3521. LAR-16833-1