Sound propagation for a low height impulsive source over an absorbing ground Frits van der Eerden 1 , Frank van den Berg, Ad van Heijningen TNO − Acoustics & Sonar department Oude Waalsdorperweg 63, 2597 AK, The Hague, The Netherlands

ABSTRACT Noise contours around military training areas are calculated by using: a) the measured source strength for the impulsive sources and b) the calculated sound propagation using a representative set of meteorological situations. For high-energy impulsive sounds the source strength is measured at distances beyond 100 meters, where peak levels are below 154 dB. A linear model is used to correct for the sound propagation from the source to the measurement position. In this way a linear source strength is determined that can be used with linear sound propagation models. Previous results showed that the source strengths at higher frequencies (above 250 to 500 Hz) were overestimated when the impulsive source is on the ground. Apparently, the calculated sound propagation over an absorbing ground, used to determine the source strength, is accounting for too much ground absorption effect. Note that the actual meteorology and ground absorption are measured during the measurements. In this paper an impulsive reference source is used at two heights and the sound propagation is measured and calculated at increasing distances. It is shown that the measured ground absorption effect saturates for higher frequencies. Next, an unknown impulsive source was measured and the effect of saturation was demonstrated.

1. INTRODUCTION

In general the noise contours for an area of interest can be determined using known acoustic sources and a calculation model for the sound propagation. For The Netherlands Ministry of Defence (MoD-NL) the impulsive sources used at military training areas are measured according to one of three techniques, depending on the peak level of the shockwave. For small arms the measurement distance for the microphones around the weapon is about 10 meters [1]. For larger weapons a distance of more than 100 meters is used [2]. The source strength is corrected for the sound propagation (or sound transfer) using a model that accounts for the actual meteorology and ground effect, such as the parabolic equation model [3]. By using a larger distance the measured peak levels are below 154 dB so that a linear sound propagation model can be used and also the source strength can be further used with linear models. Many measurements were carried out for large weapons, such as muzzle blasts from artillery fire and grenade detonations. For a muzzle blast the source height can be for example 2 meters, but for grenade detonations also zero height is used. It was seen that the source strengths at higher frequencies (above 250 to 500 Hz) were overestimated using this procedure when the source is on an absorbing ground.

1 Frits.vandereerden@tno.nl

Jai. inter noise 21-24 AUGUST SCOTTISH EVENT CAMPUS ? O ? . GLASGOW

In Figure 1 (left hand side) a source spectrum for a relatively high-energy impulsive source is drawn, with a logarithmic scale for the frequency, usually in one-third octave bands. The figure also shows a source spectrum with an overestimation at higher frequencies.

SEL (dB) source spectrum too high source spectrum correct Frequency (Hz)

Figure 1: Left: Sketch of a source spectrum with correct levels and with too high levels due

to incorrect calculation of the sound transfer. Right: Measured and calculated sound transfer between source and microphone position.

Frequency (Hz) OL 9 0S (ap) "1p OL-

The objective of this paper is to show that this overestimation is due to the calculated sound transfer. The source spectrum is determined by using measurements and corrections for the sound transfer (meteo & ground), distance and air absorption. The right hand side figure shows an impression of the calculated sound transfer dL tf for a source on the ground. At low frequencies there is almost no ground absorption and the sound levels are doubled (coherently, i.e. +6 dB). The absorbing ground effect, which can be calculated with a complex impedance model such as Wilson [4] or Delany & Bazley, reduces the sound levels at higher frequencies. This is the result of destructive interference between two point sources. It will be shown that the sound transfer can also be measured with a known reference source. This resulted in a corrected approach for the calculated sound transfer using a saturation level of -5 dB for the heather ground on which measurements normally take place.

Figure 2: Grenade at 0m and 4m height (left). Reference source at 0m and 4m height (right).

This paper describes the measurements with a reference source (250 gram PETN) at 4m and 0m height and the measurements with a grenade, also at both heights, see Figure 2. The source strength for the reference source is determined by using measurements also at closer distances (and including a nonlinear propagation model). Finally, the source strength of the grenade is determined using an empirically corrected sound transfer when the source is detonating on an absorbing ground. 2. MEASUREMENTS WITH A REFERENCE SOURCE

Measurements of 250 grams PETN based explosions (TNT equivalent 415 grams) were carried out at three different heights (0, 2m and 4m) and at four distances (15; 30; 60 and 120m), see Figure 3 . Also, ground impedance measurements were carried out using the two microphone technique and a one parameter model [4]. The wind and temperature profiles were determined using a meteo mast. The measurement were performed on a so called “smoke field” where test are performed with combustible materials. This terrain is regularly plowed to prevent the ground from catching fire. It was expected that the soil would be acoustically very soft, but rather high values for the flow resistivity were found, in the order of 4000 kPa.s.m -2 . When the ground was compacted by foot, and a bit rougher as a consequence, lower values were measured, in the order of 500 kPa.s.m -2 .

Figure 3: Set-up of measurements with a reference source at distances of 15, 30, 60 and 120m.

At 4m and 0m height. Microphones at 5m height.

The effect of the wind and temperature profiles is illustrated in Figure 4. The left hand side shows the measured wind and temperature at several heights during a measurement series of 5 detonations. The average profiles (Businger-Dyer) are shown in blue. The resulting profile for the speed of sound causes a refraction of the sound rays as shown in the right hand side of the figure. In this case there is a downwind effect. The source is depicted on the left in this figure at 2m height. The microphone is on the right at 5m height.

Meteoprofielen

Figure 4: Wind and velocity profiles during the measurements (left) and corresponding refracting

sound rays, for a source height of 2m and a distance of 120m (right). Microphone indicated at 5m height.

The measurements for the detonations are given in the next section. They are fitted to a Friedlander shockwave at 15m distance and propagated to 120m including nonlinear effects.

3. FRIEDLANDER WAVEFORM AND NONLINEAR PROPAGATION

The results from the measurements at 4 distances are depicted in Figure 5, with the top figure for 15m and the bottom one for 120m distance. A series of five measurements is shown with the thin lines. The 4 figures at the left hand side show the 4m high detonation results, at the right hand side the height of the detonation is 0m. At 15m distance, at 4m detonation height, the directed and ground reflected shockwave are separated by about 7ms. A Friedlander shockwave was fitted to the direct wave only and the ground reflection was added (using a reflecting ground for the model). This is shown with the red line. Next, the Friedlander shockwave is propagated nonlinearly [5] to the distances of 30, 60 and 120m distance to be able to compare this with the measurements. For the 4m high source a good comparison can be seen, although the reflected amplitude is somewhat overestimated. For the 0m high source the same Friedlander fit as for the 4m height has been used. The comparison with measurements shows a too high peak amplitude, but the rest of the waveform is captured rather well. It can be seen that the positive phase duration of the shockwave increases from about 4 to 5ms due to nonlinear effects. The broken lines in the figure show the propagation of the Friedlander shockwaves without nonlinear effects.

Pressure (Pa) Pressure (Pa) Pressure (Pa) Pressure (Pa) 2000 1000 1000 500 time signals per microphone Height am, mic. 15m = = Feediander, 7 =7.2 ms er (noi), 1 ms) ° 5 10 15 20 Time (ms) Height am, me. 30m = = Friodiander, 67=3.8 ms — Freatander (nin). 3 me) 0 5 10 6 20 Time (ms) Height am, mic. 6Om [=~ Friestander, oT = 1.9ms — Frestander (nonlin) oT =: ° 5 10 18 20 Time (ms) a Height 4m, mie. 120m = =Frediander, T= 1s tdlander (pot ams) ° 5 10 16 20 Time (ms) 25

Figure 5: Measured pressure-time signals for the reference detonations at 4 distances (15, 30, 60, 120m)

and detonation heights of 4m (left) and 0m (right). The Friedlander fit is done at 15m distance and propagated to further distances, with and without nonlinear effects.

The nonlinear propagation is also shown in Figure 6. The blue line is at 15m distance (using a 4m high source and a 5m high microphone), the magenta lines are the initially calculated shockwaves where the higher amplitudes travel faster than lower amplitudes. By using the equal area rule the shockwaves, shown in black, result [5].

Pressure (Pa) Pressure (Pa) Pressure (Pa) Pressure (Pa) time signals per microphone 6000 4000 —Heightom, mic. 15m = =Feiediander, a7 =0 ms 2000 — Frediander(pntin). dT = 0 ms, ° 2000 0 5 10 18 20 Time (rms) ein) Height om, mic. 30m = —Feediander, 67 = 0 ms 1000 — Friedander(ponin). dT = 0 ms) ° 1000, " i x ° 5 10 15 20 Time (ms) 1500 +000 Height om, mic. 60m (= ~Fredtander, oT =0 ms +500 Friedlander (nonin) a = 0 ms ot 500 0 A 10 16 20 Time (ms) 1000 -—yy \ Height Om, mic. 120m 500 |= ~Frediander, oT = 0 ms 500 a 10 16 20 Time (ms) 25

Figure 6: Starting Friedlander waveform at 15m distance (blue) and nonlinear propagated

waveforms for further distances (magenta), including the ground reflected ones. The black waveforms use the equal area rule to obtain the final waveform.

The measurements at 120m distance are used to determine the source strength. However, it needs to be corrected for the ground and the meteo effects. It can be seen (also by comparing spectral results) that the Friedlander results with a 4m high source at 120m distance show a good correspondence with the measurements. Therefore, by using the direct wave only, the source strength can be determined by only correcting for the distance (and air absorption). The results are shown in Figure 7.

Pressure (Pa) Waveform at distance R +5000 (6.0339 m. nial waveform) |= ~ Propagated waveform 17:5 m. Using equatrea |—R=30 m, Using equatarea ‘3m. Using equatarea R60 m. Using equat-aea 7m. Using equatarea 0 m. Using equal-area 20m. Using equalarea 4000 3000 2000 1000 5 4 3 2 4 Relative distance (m)

Figure 7: Source spectrum for the reference detonation using the fitted Friedlander

waveform for the direct path only, i.e. without ground reflection.

This reference detonation can then be used to measure the sound transfer from the source position to the different microphone positions using the following equation:

dL tf = L E, meas – L E, source

with

dL tf = transfer function; L E, meas = measured spectrum of the Sound Exposure Level at the microphone; L E, source = source spectrum as depicted in Figure 7 .

The measured transfer functions at 120m distance, for the 4m and 0m high sources, are shown in Figure 8 . The depicted sound transfer only accounts for ground and meteo effects, i.e. effects of

dB 170 160 150 140 130 120 110 Source spectrum reference explosion Direct only (nonlin) 16 32 63 125 250 500 Frequency (Hz) 1k 2k 4k

distance and air absorption are excluded. Also shown are the calculated sound transfers for three different values of the flow resistivity  , with decreasing absorption. A +6 dB value indicates sound pressure doubling due to the ground reflection. At 4m height the ‘ground dip’, the measured destructive interference between direct and ground reflected wave, can be seen around 250 Hz. A fair comparison with the calculated sound transfer for  =1000k can be seen (excluding wind here). At 0m height the measured sound transfer decreases from +6 dB, but saturates for this sandy ground around 0 dB. While the calculated sound transfer decreases below -10 dB for higher frequencies.  In the next section the results for grenade and reference detonations are given at 4m and 0m height above a heather ground type. The measured impedance was  =160 kPa.s.m -2 .

Sound transfer (Friedlander source) Distance 120m 0 16 32 63 125 250 500 Frequentie (Hz) 1k kak

Figure 8: Measured sound transfer (magenta) at 120m distance by using the known source spectrum

depicted in Figure 7 for 4m high (left) and 0m high source (right). Blue lines represent calculated sound transfer above absorbing ground with three different flow resistivities.

Sound transfer (Friedlander source) Distance 120m 16 32 63 125 250 500 tk 2k 4k Frequentie (Hz)

4. GRENADE AND REFERENCE DETONATIONS

Measurements were conducted at four positions around the source (see Figure 9). The grenade and reference source were placed at 4 m height and on the ground as shown in Figure 2.

Figure 9: Source location for the grenade and reference detonation measurements (left)

and layout of the microphone positions at 125m distance and 5m height (right).

A typical pressure-time signal for the grenade can be seen in Figure 10 for a 4m high source. The “noise” arriving before the shockwave is due to supersonically flying shrapnel from the grenade,

mic 4 mic 1 mic 3 prio meteo

causing small N-waves (visible when zooming in). By using a rectangular time window, shown in black, this “noise” is excluded from the analysis to determine only the spectrum of the detonation.

Figure 10: Pressure-time measurement for a grenade detonation. The debris before the

shockwave consists of supersonically flying parts. The signal within the rectangular window is used for the analysis.

4.1. Reference detonation measurements The measured spectra for the four microphone positions for the reference detonation are depicted in Figure 11 for the 4m source height (left hand side). These spectra are corrected for the distance with 10 log 10 (4 π R 2 ) using R = 125m. The source spectrum of the reference detonation from Figure 7 is shown in the same figure by a black line. In the right hand side figures, the corresponding sound transfer function are shown, using the difference between the source and the measured spectra. The upper two figures are for a downwind situation and show a ground dip between 125 and 250 Hz. In the lower figures the wind direction towards the microphones is 67 and 127 degrees; for 127 degrees this is a headwind situation. As a result the interference due to the direct and ground reflected shockwaves is reduced. In general downwind situations are used to determine the sound source.

Measured spectrum (reference) Sound transfer spectrum (reference)

Figure 11: Left - Reference detonation measurements at 4m height at 4 microphone locations around the

source, corrected for distance only (125m) and compared to the known source spectrum (black). Right – Corresponding measured sound transfers using the known source spectrum.

In Figure 12 similar results are shown for the reference detonation at 0m height. These measurements were taken one day later and in this case the lower two figures are for a downwind situation. Two observations are made. First, the reference source on the ground emits less low-frequency energy compared to 4m the height one. These frequencies are assumed to be damped via the mass- spring behaviour of the ground. This is not the case for the grenade (as shown in Figure 13), as the metal cover reduces the interaction with the ground when detonating.

to Measured source spectra 1 Measured source spectra a Measured sound transfor 4 Measured sound transfor 160 160 B19 8 109] 140 140 NW wind: 53° (-downnend) ‘Ww +7° (dour) NW wind 53° (-downwind) 130 130 “0 16 32 63 125 250 00 tk & ak 16 32 63 125 250 500 hk & a 16 32 63 125 250 600 tk 2% 16 52 63 125 250 600 th 2% dk Frequentie He) Frequent (Hz) Froquonto (He) Frequontio Hz) roy +» Measured source spectra a Measured source spectra to Measured sound transter to Measured sound transfer 160 160 5 2180 8 100) go go 0 140 5 5 | NW wind: +872 NW wind: +127° (adie) NW wind: 167° NW wind: #127? (Deadwind) 10 130 0 0 16 32 63 125 250 00 tk & ak 16 32 63 125 20 00 hk & a 16 52 63 125 250 500 hk 2% 16 32 63 125 250 600 th & 4k Frequentie (Hz) Frequentie (Hz) Frequentio (Hz) Froquontio (Hz)

Second, the measured sound transfer shows a ‘saturation’ at about -5 dB at high frequencies, using the downwind microphones only. It shows that a coherent interference with the ground (which is assumed when calculated, using a ground with a complex impedance and neglecting turbulence effects) is not occurring in practice at higher frequencies.

Measured spectrum (reference) Sound transfer spectrum (reference)

__ Measured source spectra at6 ‘Measured source spectra 46 Measured sound transfer 7 Measured sound transfer 160) s 6 B 150 go Bo 140 $ 6 [NE wind: -142° (headwind) NE wind: -82° NE wind: -142° (headwind} NE wind: -82° 130 210 “10 = — - 16 32 63 125 250 600 tk & 4k 16 32 63 125 250 600 tk %& 4k 16 32 63 125 250 500 tk 2% 4k 16 92 63 125 250 500 tk m% 4k Frequent (Hz) Frequent (Hz) Frequentie (Hz) Frequent (Hz) ____ Measured source spectra 70, Measured source spectra @ Measured sound transfer e Measured sound transfer 160 5s 5 8 180 Bo Bo 140 6 [NE wind: -22° (downwind) [NE wind: +38° (dowrwind) NE wind: -22° (downing) [NE wind: #38° (downwind) 130 19 L- oe 19 L— —~ ~ —$- 16 32 63 125 250 500 tk 2% 4k 16 32 63 125 250 500 tk & 4k 16 92 63 125 250 500 tk 2% 4k 16 32 63 125 250 500 tk & Frequent (Hz) Frequentio (Hz) Frequentie (Hz) Frequentie (Hz)

Figure 12: Left - Reference detonation measurements at 0m height at 4 microphone locations around the

source, corrected for distance only (125m) and compared to the known source spectrum (black). Right – Corresponding measured sound transfers using the known source spectrum.

4.2. Grenade detonation measurements At left hand side in Figure 13 the measured spectra for the grenade at two source heights are shown in one-third octave bands. Only the downwind results are given and the spectra are corrected for distance. As a comparison the measured spectra for the reference source are given at the right hand side. The source spectrum of the reference source is given with a black line in both figures. For the grenade the sound levels up to 125 Hz do not differ much, but for the reference source the levels for the 0m high source are about 4 dB lower. At 4m height the ground interference between 125 and 250 Hz can be clearly seen in the measurements. Above 250 Hz the 4m high source shows higher levels compared to 0m height, due to the absorbing ground.

Grenade measurements (corrected for 125m distance) Reference measurements (corrected for 125m distance)

dB 170 Measured source spectra 81 mm grenade 165} 160 1554 150} 145 140 135 130 Z == meas. om +38, ‘Source lev. ref —— meas. 4m 53° — meas. 4m +7° = = meas. om-22° 16 32 63 125-250 500 Frequency (Hz) 1k 2k 4k

Figure 13: Left – Grenade measurements corrected for distance (125m) at 4m and 0m height , only for the

downwind microphone locations and compared to the known reference source spectrum (in black). Right – Idem, for the reference detonations. Values in degrees indicate angle with wind direction.

dB 170 Measured source spectra reference source 165 160 155 150} 145 140 135 130 Source lev. ref ‘meas. 4m -53° — meas. 4m 47° = = meas. om-22° == meas. om +38° 16 32 63 125 © 250 500 Frequency (Hz) 1k 2k 4k

5. SOURCE STRENGTH AND SOUND TRANSFER SPECTRA

The source spectra, corrected for the ground and meteo are derived using a calculated sound transfer, either with or without a saturation. Figure 14 shows the reference source spectrum measured at 4m and 0m height for two downwind microphones positions. The black line is the source spectrum of the reference source derived in section 3. The source levels at 0m height are shown with broken lines for the originally calculated sound transfer, without a saturation at -5 dB. It can be seen that the source levels above 250 Hz match the originally determined source levels. Below 63 Hz the reference source at 0m height emits about 5 dB less sound due to interaction with the ground. For the 4m high source no saturation is applied. The peaks at 1000 and 2000 Hz are attributed to the calculated interference pattern as shown in Figure 16 (broken lines in top figure for 4m height).

Figure 14: Source spectrum for the reference detonation at 4m and 0m height using a

175 170 145 140 Reference source spectra (4m & 0m height) — Ret. source 4m (Friedlander) —Meas. source 4m (original) —Meas. source 4m (original) = =Meas. source Om (original) = =Meas. source Om (original) —Weas. source Om (saturated) —eas. source Om (saturated) 16 32 63 125 250 500 1k 2k 4k Frequency (Hz)

calculated sound transfer. At 0m height a saturation for the sound transfer has been used (dashed lines: without saturation).

The same approach has been used for the grenade detonations, the results are shown in Figure 15. Note that the source levels for 0m and 4m height are comparable up to 63 Hz, so a reduced level of interaction with the ground is absent for the grenade.

Figure 15: Source spectrum for the grenade detonation at 4m and 0m height using a

calculated sound transfer. At 0m height a saturation for the sound transfer has been used (dashed lines: without saturation).

dB 175 170 165 160 185 150 145 140 81mm source spectra (4m & Om height) —eas. —Meas. = —Meas. = —Meas. —Neas. —— Ref. source 4m (Friedlander) source 4m (original) source 4m (original) source 0m (original) source Om (original) source Om (saturated) ‘Meas. source Om (saturated) 7 Not 16 32 63 125 250 500 Frequency (Hz) 1k ok 4k

The main result is that for the source on the ground, the source levels above 250 Hz are largely reduced when a saturation for the sound transfer is applied. Compared to the reference source spectrum (black line) the levels for the grenade on the ground are a bit higher though. For the 4m high grenade peaks at 1000 and 2000 Hz can be seen, similar to the reference detonation. These are the result of the calculated interference from the direct and ground reflected shockwaves, as shown in Figure 16. The upper figures show the measured (solid lines) and calculated sound transfers for a 4m high source above a heather ground including wind (broken lines). The calculated ones show more interference than measured. When correcting the measurements with the calculated sound transfer these dips become peaks in the source spectrum. The measured sound transfer indicated less pronounced peaks, so some saturation may be assumed for a higher source as well. However, when calculating noise contours usually octave bands are used (instead of one-third octave bands) making these peaks less pronounced. Figure 16 also shows the calculated sound transfers without the effects of wind. The effect of wind is small. The largest effects are for a 4m high source (compare green broken line to solid magenta line).

Figure 16: Comparison of measured (solid line) and calculated (broken line) sound

transfers. Top – source at 4m height. Bottom – source at 0m height. Using a measured flow resistivity for the ground of 160k Pa.s.m -2 and a downwind. Thin lines: three different flow resistivities without wind.

6. CONCLUSIONS

To determine the source strength of high-energy impulsive sources, such as artillery fire or grenade detonations, measurements are done at distances of 100m or more. The sound transfer between the source and the microphone is used to account for ground and meteo effects. For sources at the ground or at low-height the calculated sound transfer shows relatively low propagation levels for an absorbing ground. As a result relatively high levels for the source are obtained. Measurements of the sound transfer with a reference impulsive source close above a sandy or heather ground show a saturation or minimum for these low levels, of 0 and -5 dB respectively. This empirically obtained

Measured sound transfer Measured sound transfer 10 —Measured ~@=50k no wind ‘2=500k no wind NW wind: -53° (~fownwind) 10 10 16 32 63 125 250 500 1k 2k 4k Frequency (Hz) Measured sound transfer 16 32 63 125 250 500 1k 2k 4k Frequency (Hz) Measured sound transfer NE wind: -22° (downwi 16 32 63 125 250 500 1k 2k 4k Frequency (Hz) 16 32 63 125 250 500 1k 2k 4k Frequency (Hz)

saturation, by using a reference source, is applied in the MoD-NL analysis method for the determination of the acoustic source strength of a muzzle blast or detonation by measurement (NIELS – method III). The uncertainty for the sound transfer at these large distances can be reduced by using (much) shorter distances. It also reduces the amount of work in the field. In that case a nonlinear sound propagation model is needed to propagate the measured shockwave to the linear region, at which distance a linear sound source can be determined [6]. This method will be further investigated. 7. ACKNOWLEDGEMENTS

This work has been done for The Netherlands Ministry of Defence (BMW-ost contract). The measurements were carried out by KCWM from the MoD-NL and the cooperation is gratefully acknowledged. Also acknowledged are the interesting discussions with CERL of the US-Army. 8. REFERENCES

1. ISO 17201-1:2018, Acoustics – Noise from shooting ranges – Part 1: Determination of muzzle

blast by measurement, International Standard ISO , 2018, https://www.iso.org/standard/ 66940.html 2. E. Buchta, Standard method to measure the sound exposure emissions and immisions from large

weapons, No. 8/559 in Meβ, Steuerungs- und Regelungstechnik, Fortsschritt-Berichte (In English and German) VDI Verlag , Düsseldorf, 1996 3. K.E. Gilbert and M.J. White, “Application of the parabolic equation to sound propagation in a

refracting atmosphere”, J. Acoust. Soc. Am. , 85 (630-637), 1989 4. D.K. Wilson, Simple, relaxational models for the acoustical properties of porous media, Applied

Acoustics 50(3) , 1997 5. A.D. Pierce, Acoustics. An introduction to its physical principles and applications , AIP, New

York, 1991, Chapter 11. 6. F.J.M van der Eerden and F.H.A. van den Berg, The acoustic source strength of high-energy blast

waves: combining measurements and a non-linear model, Proceedings of the 20 th International Congress on Acoustics , ICA 2010, Sydney, 23-27 August 2010