Logo of the Physikalisch-Technische Bundesanstalt

Directivity pattern of airborne ultrasonic sources

05.05.2014

Airborne ultrasound can be emitted with a very strong directivity by sound sources such as ultrasonic cleaning baths or welding equipment. Due to this focusing, very high and hearing-damaging sound pressure levels may occur locally. A new measuring arrangement now allows the emissivity of various sound sources to be determined with high resolution and with level accuracy.

The noise load due to airborne ultrasound is constantly increasing, both in the private and in work environments. Devices such as ultrasonic cleaning baths and ultrasonic welding facilities are economically sensible and widespread standard procedures in industry and trade. Also the risks due to noise pollution are increasingly receiving more attention – although the actual knowledge with regard to risks is neither well-founded nor complete yet. Hence, it is deemed certain that ultrasound can lead to irreversible hearing loss in the normal audible hearing range, however, it is not known yet how this high-frequency sound affects the hearing and how much ultrasound is too much for a person (see also [1]). The powers implemented by typical devices are usually quite high, and the emitted airborne sound – even though it is most of the time a side-effect – can be very loud.

The sound pressure level is amplified by the strong directivity of ultrasound. The wavelengths are only a few millimetres long and the reflecting surfaces are often by far larger. This leads to interferences and strongly directed reflection. The acoustic power of a source is focused in a certain direction, and the resulting sound pressure reaches high values, even at some distance from the source. Hence, even sources as small as a commercially available ultrasonic cleaning bath for glasses and jewellery may exceed the limit values (see VDI 3766 [2]) of what is considered safe for the hearing (see also [3]).

To be able to investigate the directivity of various sources more precisely, a measuring arrangement was set up which allows the sound pressure level occurring around a certain source of interest to be recorded, quantized and assessed with high resolution and with level accuracy.
The sound sources should be characterized independently of the influence of the surrounding space. For this purpose, they are placed in an anechoic chamber where reflections distorting the sound level are suppressed. Fig. 1 shows the scanner inside PTB's large acoustic anechoic chamber. Two controllable axes are used to record and assess the sound pressure signal along a fixed, adjustable radius and at several points that are distributed around the sources. The principle of this arrangement is, thus, similar to that used to determine the sound power on machines. Here, however, not a semi-infinite space with a reflecting floor, but a full-anechoic room is used to be able to determine the real directivity pattern. The recordings are carried out with a ¼"‘ measuring technique, which allows the actual levels to be determined with a specified measurement uncertainty. For the assessment, different quantities can be used such as, e.g., the unweighted peak sound pressure level LZpeak suggested in VDI 3766. Thanks to the offline analysis, any assessment quantity can be used.

Figure 1: Picture of the scanner in the anechoic room. A source can be placed on the variable rotating stage above which a microphone is positioned on a gibbet.

Fig. 2 shows the sound field around a small cleaning bath, once with the lid closed, once with the lid open. Closing the lid reduces the peak sound pressure LZpeak directly above the source from approx. 141 dB to approx. 114 dB. In the picture with the lid closed, the sound beams increased by 6 dB are clearly visible; they emphasize the opening groove of the lid (facing forwards) and the untight hinges (facing backwards).

Figure 2: Emissivity of a cleaning bath (see photo): with the lid open (left) and with the lid closed (right).

With 12430 dots, Fig. 3 shows a high-resolution scan over an array of piezo-loudspeakers. Strong level variations are clearly visible in the dihedral angle range of up to ±45°. In addition, a small pattern at the zenith of the image can be seen which, even in the far field, reproduces the size and shape of the loudspeaker array quite accurately.

Figure 3: Interference patterns of an array of piezo-loudspeakers: view from above (left), 3D lateral view (right). At the zenith, one can still clearly recognize the pattern of the four switched-on loudspeakers (highlighted in the photo) even at a measurement distance of 44 cm.

Directivity patterns detected in such a way can, for example, be used to predict how certain sources will behave in a reflecting environment. But in particular, it makes it possible now to characterize and to modify fields of nearly any ultrasonic sources.

References:

[1] Website of the EARS project dealing with the perception of non-audible sound: www.ears-project.eu

[2] VDI 3766:2012-09: Ultraschall - Arbeitsplatz - Messung, Bewertung, Beurteilung und Minderung (in German)

[3] Scientific news Opens external link in new windowSmall sources, loud noise (2013)

Contact person:

Christoph Kling, Dept 1.6, WG 1.63, e-mail: christoph.kling@ptb.de