Traceable sound power measurements in essentially diffuse or free fields

Full text

(1)INTER-NOISE 2016. Traceable sound power measurements in essentially diffuse or free fields Håkan Andersson1; Volker Wittstock2 1. SP Technical Research Institute of Sweden, Sweden 2. Physikalisch-Technische Bundesanstalt, Germany. ABSTRACT Current methods for the determination of sound power such as the ISO 3740 series do not have a direct traceability to a primary standard. The different methods are regarded as equal, but they do give results with systematic differences. One task of the European project “EMRP JRP SIB 56 SoundPwr” has been to develop and investigate traceable and consistent methods for sound power measurements in essentially diffuse or free fields. By using a calibrated reference sound source as a traceability source, qualification procedures for setups for sound power determinations in different environments have been investigated. Uncertainty sources such as sound field sampling, directivity, frequency content and impedance of the device under test have been analysed. The results are presented in the contribution. Keywords: Sound power, traceability. I-INCE Classification of Subjects Number(s): 72.4. 1. INTRODUCTION The quantity sound power is today not traceable to any primary sound power standard. Instead sound power determinations normally are based on sound field measurements such as sound pressure or sound intensity. Standardised calibration procedures for sound pressure and sound intensity measurement devices are published. However these indirect traceability chains are relying on sound field assumptions which are hard to verify. A primary sound power standard is described in (8) and the dissemination of the quantity sound power by use of a transfer standard is described in (9). The sound power transfer standard can be the commercially available aerodynamic reference sound source (RSS), but can also be an electroacoustic sound source which may be tonal (10). By using a calibrated sound power transfer standard to determine a correction term for a certain measurement surface in a certain measurement environment it is possible to establish a more straight-forward traceability chain for sound power measurements.. 2. PRESENT CONCEPTS 2.1 Free Field Methods The standardised sound power determination methods ISO 3744, ISO 3745 and ISO 3746 are all based on the following equation. LW. L p 10 log10 S ¦ K n .. (1). where L p is denoting the surface-averaged sound pressure level measured over an enveloping surface, S is the total area of the measurement surface and K 0 is correcting for the atmosphere,. K 1 for the background noise and K 2 for the room. Not all corrections apply to all methods mentioned. 1 2. hakan.andersson@sp.se Volker.Wittstock@ptb.de. 6783.

(2) INTER-NOISE 2016. Different methods are suggested to determine the environmental correction K 2 , in some cases without consideration of the measurement surface. 2.2 Diffuse Field Methods The standards for sound power determinations in diffuse field are describing two fundamentally different methods, the comparison method and the direct method. In the comparison method the sound power level is determined by substitution with a reference sound source calibrated according to ISO 6926. In the direct method the sound power level is determined by the following formula.. LW. § Sc § A· L p 10 log 10 A 4,34¨ ¸ 10 log 10 ¨¨1 ©S¹ © 8Vf. · ¸¸ C 6 . ¹. (2). were L p is the mean sound pressure level in the room, A is the equivalent absorption area of the room, S is the surface area of the room, V is the volume of the room, c is the speed of sound, f is the measurement frequency and C is the sum of some minor climate-dependent corrections. The equivalent absorption area is normally determined by measurement of the reverberation time in t he room. A procedure for qualifying reverberation test rooms is described in ISO 3741. The principle of it is to check the standard deviation of the mean sound pressure level when the sound source is moved between different positions in the reverberant room.. 3. TESTING THE APPLICABILITY OF THE TRACEABILITY CONCEPT The basic idea of traceable sound power measurements is that the free -field sound power level of the transfer standard LWfree,TS is determined by substituting the transfer standard by the primary source in a hemianechoic or reverberant field. The sound field quantities sound pressure or sound intensity may be used in hemianechoic fields and the averaging is performed on an enveloping surface. In diffuse fields, the sound pressure is to be averaged over the room volume. Field quantity measurements have to be performed for both sources yielding the desired free-field sound power level of the transfer standard LW,free,TS LW,free,PS Lp/I,TS Lp/I,PS . (3) The sound power measurement of a device under test follows the same princ iple, only under less ideal conditions, means no hemianechoic or reverberation rooms. The sound power level for a device under test is then LW,free,DUT LW,free,TS Lp/I,DUT Lp/I,TS . (4) A general verification of this approach is not easily obtained. Nevertheless, the appl icability of this approach requires that the level difference of the averaged field quantities 'Lp/I Lp/I,DUT Lp/I,TS . (5) is constant for a specific source in different environments. This is tested in this contribution for different sources in different environments.. 4. TESTED ENVIRONMENTS 4.1 Free sound fields The standardised free field sound power determination methods have certain requirements on the test environment. This is achieved by setting allowable limits to the environmental correction K2 , e.g. in ISO 3744 the maximum allowed K 2 for A-weighted levels, K2A, is 4 dB. The procedure for determining the K 2 may be by comparison to a calibrated reference sound source, measurement of the reverberation time in the room or, in ISO 3746, just approximate the sound absorption by visual inspection. Another concept for room qualification for a certain microphone setup, e.g. a 9 microphone box, is to determine the surface-averaged sound pressure level (SPL) emitted from a reference sound source, L p , RSS ,room . Then repeat the measurement in a hemi-anechoic room with the same source and the same. 6784.

(3) INTER-NOISE 2016. microphone configuration, yielding a new surface-averaged SPL, L p , RSS ,hemi . Supposing the sound power level of the reference sound source, LW,RSS, is unchanged the following formula applies:. L p, RSS ,room 10 log10 S K1,room K 2,room. LW , RSS. L p, RSS ,hemi 10 log 10 S K1,hemi K 2,hemi. (6). As the measurement surfaces, S, are the same, they cancel out and as the hemi-anechoic room has low background noise and room reflections its K terms may be considered neglectable. Left of the formula above is:. K 2,room. Lp,RSS ,room Lp,RSS ,hemi K1,room. (7). Following the procedure described above SP and PTB determined the K 2,room in the environments used. The result is presented in figure 1 and 2. The measurements at SP were done by using a 9 microphone, 2,3 x 2,3 x 1,3 m³ box-shaped measurement surface according to ISO 3744. Measurements at PTB are yielded by a 9 microphone box-shaped array with the dimensions 2,5 x 2,5 x 1,5 m³. This setup was used in the hemianechoic room, in a partly open measurement site in a large hall and in a 50 m³ room. The latter was used once with a large amount of damping and once without any damping at all. The resulting room corrections cover a wide range. Very pronounced is the 40 Hz anomaly in the 50 m³ room which is the resonance frequency of a floating floor. At this frequency, the floor radiates a large amount of sound which results in the huge room correction.. Figure 1. K 2,room determined in different environments at SP.. 6785.

(4) INTER-NOISE 2016. 20 50 m³ empty 50 m³ damped. 15. cellar hem.an.room. K2 (dB). 10. 5 0 -5. 10. 100. 1000. 10000. 100000. Frequency (Hz). Figure 2. K 2,room determined in different environments at PTB.. 4.2 Diffuse sound fields An alternative to the standardised procedure for qualification of reverberant test setups is to use a transfer standard, e.g. a reference sound source, which is calibrated with traceability to a primary sound power standard. The room-averaged sound pressure level in the reverberation room is measured using the same setup and the same source positions as in normal sound power measurements. Under assumption of a reverberant field, the sound power is calculated from the mean sound pressure level by ' use of equation (2). This imaginary sound power level, LW , is compared to the traceable calibrated sound power level of the same transfer standard, LW ,TS , and the difference is determined.. 'LW. LW' LW ,TS. (8). The sound power level difference, 'LW , may serve as a quality indicator of the reverberant room and its setup. SP and PTB have determined 'LW for some of their test rooms. SP has done the tests in a 200 m3 reverberation room, in a 130 m 3 reverberant room for building acoustics and in a 50 m3 reverberant room. In all rooms the measurements have been done with the room em pty and with an amount of absorbers inserted. The measured reverberation times in the rooms at SP are given in figure 3 and the sound power level differences, 'LW , in figure 4. At PTB, measurements were performed in altogether three different reverberant environments. These are a 200 m³ rectangular reverberation room and a 50 m³ room which was once highly damped and once used without any damping material. The reverberation times were measured in these rooms and they turn out to cover a large range (figure 5).. 6786.

(5) INTER-NOISE 2016. Figure 3. Reverberation times in the tested reverberant rooms at SP.. Figure 4. The sound power level difference, 'LW , determined at SP. 100 200 m³ rev. room 50 m³ room empty 50 m³ room damped. T (s). 10. 1. 0.1 10. 100. 1000. 10000. 100000. Frequency (Hz). Figure 5. Reverberation times of the three reverberant environments used at PTB.. 6787.

(6) INTER-NOISE 2016. In figure 4 it can be seen that the 'LW in the 50 m3 room is clearly deviating from the larger rooms, both for the empty and the absorbent test case. One may also note that there is almost no correlation between the reverberation time and the 'LW , which may be understood that the reverberation time only is not a good indicator if the room is suitable for sound power determinations by use of reverberation methods.. 5. SOURCES USED FOR THE TESTS At PTB, four different sound sources are used for the test. This is a vacuum cleaner, a compressor, a tapping machine running on a steel plate and an electroacoustic source developed for room aoustic measurements (10), see figure 6. As a transfer standard, an aerodynamic reference sound source is used. The sound pressure levels emitted by these sources were measured using the scanning apparatus mentioned in (11) on an enveloping surface in PTBs hemianechoic room. From these sound pressure levels, sound power levels were calculated according to eq. 1. For this measurement, one -third octave bands and FFT bands were used. The sources exhibit a broad spectrum was a considerable amount of individual tones (figure 7). Furthermore, they have a directivity which clearly distinguishes from the aerodynamic sound source used as a transfer standard.. Figure 6. Sources used at PTB 100 90. LW dB re 1 pW. 80. 70 60 50 40 30. vacuum cleaner compressor RSS. 20. tapping machine volume source. 10 10. 100. 1000 Frequency (Hz). 10000. 100000. Figure 7. Sound power levels in FFT bands as measured in PTB’s hemianechoic room At SP, four different sound sources are used for the test. This is a garden cleaning blower, an electric screwdriver, an angle grinder and an emery grinding machine, see figure (8). The first two of them are battery operated. Some of them had a tonal spectrum. As a transfer standard, an aerodynamic reference sound source is used. All sources are placed on floor positions.. 6788.

(7) INTER-NOISE 2016. Figure 8. Sound sources used at SP. 6. TEST RESULTS For the four sources under investigation at PTB, the sound pressure level differences according to eq. 5 are quite constant for the same source over different approximated hemi-free and reverberant environments (figure 9). Large differences occur for the tapping machine due to a temporal instability. For the volume source, the room with the floating floor shows a significant peak at 40 Hz. At this frequency this floor is excited by the transfer source (figure 2) which results in an exceptional large sound pressure level for the transfer source. This effect is less obvious for the other sources since all these sources also excite the floating floor. 40. 40. appr. free field appr. diff. field. 10 0 vacuum cleaner. -20 10. 100. 40. 1000 10000 Frequency (Hz). -10. 100. compressor. 10. tapp.machine 10. -10 100. 100000. 30 20 10 0. volume source. -10 1000 10000 Frequency (Hz). 1000 10000 Frequency (Hz) appr. free field appr. diff. field. 40. 0. -30. 0. 50. 10. -20. 10. 100000. appr. diff. field. 20. 20. -20. appr. free field. 30. Lp,TS - Lp,DUT (dB). Lp,TS - Lp,DUT (dB). 20. -10. appr. free field appr. diff. field. 30. Lp,TS - Lp,DUT (dB). Lp,TS - Lp,DUT (dB). 30. 100000. 10. 100. 1000 10000 Frequency (Hz). 100000. Figure 9. Sound pressure level differences acc. to eq. 5 determined by PTB It is obvious, that the sound pressure level differences do not depend significantly on the nature of the surrounding sound field. Results for approximated free and diffuse fields show a normal scatter but no systematic deviation. The major exception to this is the room with the floating floor. Since the results look rather promising, the standard deviation of the soun d pressure level differences from all seven environments has been calculated for the different sources (figure 10).The sound pressure level differences are added to the sound power level of the transfer standard to yield the sound power level (eq. 4). Since the sound power level of the transfer standard is constant in the context of this contribution, the standard deviation of the sound pressure levels equals the standard deviation of the sound power levels determined by the substitution method. Despite the fact that the sources exhibit a tonal behaviour and a pronounced directivity, only the tapping machine shows larger standard deviations. The other values are around one dB for frequencies above 100 Hz which is rather small for the large range of rooms an d sources covered. But also for the lower frequencies, results are promising. Keeping in mind the problem with the floating floor one can. 6789.

(8) INTER-NOISE 2016. hope to determine sound power levels down to 20 Hz with a standard deviation of about 5 dB. This value would also be a good estimate for the uncertainty, since the realisation of the unit and the dissemination would not contribute large uncertainty components in this frequency range.. standard deviation (dB). 12. 10. vacuum cleaner compressor tapping machine volume source. 8 6 4 2 0 10. 100. 1000 10000 Frequency (Hz). 100000. Figure 10. Standard deviation of the sound pressure level difference from figure 9 The four sound sources at SP were investigated in the same way as described above , in three approximated free fields and four approximated diffuse fields, and the result is given in figure 11. As the majority of the sound sources had a low sound power level at low frequencies those measurement had problems with background noise and the results below 63 Hz had to be omitted. The varying result of the electric screwdriver at lower frequencies is also a result of background noise . The standard deviation of the sound pressure level differences from the seven environments has been calculated for the different sources, figure 12. For the electric screwdriver no result is given below 500 Hz due to the background noise problem mentioned above. Also the emery grindi ng machine was suffering of instability.. Figure 11. Sound pressure level differences acc. to eq. 5, measured by SP. 6790.

(9) INTER-NOISE 2016. Figure 12. Standard deviation of the sound pressure level difference from figure 11. 7. CONCLUSIONS The application of the traceability concept for sound power determination requires the assumption that the ratio between the free-field sound power and the sound pressure level (or sound intensity level) is identical for the devices under test and for the transfer standard. The sound pressure level may be averaged in a room volume for (approximated) diffuse fields or averaged over an enveloping surface for (approximated) free fields. Altogether eight different sound sources were tested in a very wide frequency range (20 Hz to 20 kHz) in 13 different environments. The environments covered a range from hemianechoic rooms to reverberation rooms. It turns out that the sound pressure level differences are close to constant for the same source in different environments for frequencies above 100 Hz. The standard deviation is about 1 dB in this frequency range. Towards lower frequencies, deviations increase cons iderably. This increase may be caused by a different change of the emission of the device under test and the transfer source. Future work should focus on this aspect.. ACKNOWLEDGEMENTS Measurements at SP have been performed by Niklas Rosholm. Measurements at PTB have been performed by Christian Bethke, Heinrich Bietz and Spyros Brezas. Data analysis at PTB was supported by Martin Schmelzer. This work was carried out within EMRP Joint Research Project SIB56 SoundPwr. The EMRP is jointly funded by the EMRP participating countries within EURAMET and the European Union.. REFERENCES 1. 2. 3. 4. 5. 6. 7. 8.. ISO 3741 Acoustics - Determination of sound power levels and sound energy levels of noise sources using sound pressure - Precision methods for reverberation test rooms; 2010. ISO 3744 Acoustics - Determination of sound power levels and sound energy levels of noise sources using sound pressure - Engineering methods for an essentially free field over a reflecting plane; 2010. ISO 3745 Acoustics - Determination of sound power levels and sound energy levels of noise sources using sound pressure - Precision methods for anechoic rooms and hemi-anechoic rooms; 2012. ISO 3746 Acoustics - Determination of sound power levels and sound energy levels of noise sources using sound pressure - Survey method using an enveloping measurement surface over a reflecting plane; 2010. ISO 9614-1 Acoustics - Determination of sound power levels of noise sources using sound intensity Part 1: Measurement at discrete points; 1993. ISO 9614-2 Acoustics - Determination of sound power levels of noise sources using sound intensity Part 2: Measurement by scanning; 1996. IEC 61043 Electroacoustics - Instruments for the measurement of sound intensity - Measurements with pairs of pressure sensing microphones; 1993 Cafer Kirbas, Håkan Andersson, Claudio Guglielmone, Volker Wittstock: Primary Sound Power. 6791.

(10) INTER-NOISE 2016. Sources for the Realisation of the Unit Watt in Airborne Sound, Internoise, Hamburg, August 2016. Spyros Brezas, Patrick Cellard, Håkan Andersson, Claudio Guglielmone, Cafer Kirbas: Dissemination of the unit Watt in airborne sound: aerodynamic reference sound sources as transfer standards, Internoise, Hamburg, August 2016. 10. Heinrich Bietz, Volker Wittstock, Spyros Brezas: Investigations on the suitability of an electroacoustic sound source as secondary sound power standard, Internoise, Hamburg, August 2016. 11. Patrick Cellard, Håkan Andersson, Spyros Brezas, Volker Wittstock: Automatic sound field sampling mechanisms to disseminate the unit watt in airborne sound, Internoise, Hamburg, August 2016. 9.. 6792.

(11)

No results found