Report on

Barrier Insertion Loss Measurements

Undertaken at Leeds Metropolitan University, Portland Crescent, Leeds, West Yorkshire, LS1 3HB

Wednesday18thDecember 2013

Student Number: 77144104

IOA Diploma 2013-14 Summary

Student Number: 77144104

This report describes the processes involved when undertaking insertion loss measurements using a white noise source for three sizes of acoustic barriers (100mm, 200mm and 300mm height). Various configurations were tested to understand if the measured insertion loss values were comparable with the insertion loss prediction models of (Maekawa, 1968) and (Kurze and Anderson, 1971). It was found that correlation in octave bands was poor whilst there was good agreement in broadband insertion loss. The measurements confirmed the theories of (Maekawa, 1968) and (Kurze and Anderson, 1971) in that when one considers designing an acoustical barrier the height and diffraction paths are of vital importance as is the distance from the source and/or receiver to the barrier. Measurements also confirmed that acoustical barriers are not effective at low frequencies (<250Hz) due to the wavelength of the sound being greater than the dimensions of the barrier and that at higher frequencies (>4000Hz) the sound diffracts over the barrier.

IOA Diploma 2013-14

Student Number: 77144104

Contents List of Figures ............................................................................................................................ii List of Tables ........................................................................................................................... iii Nomenclature ............................................................................................................................ iv 1. Introduction ........................................................................................................................ 1 2. Instrumentation& equipment .............................................................................................. 2 2.1. Sound level meter ........................................................................................................ 2 2.2. Calibrator ..................................................................................................................... 2 2.3. Microphone ................................................................................................................. 2 2.4. Equipment ................................................................................................................... 2 2.5. Materials ...................................................................................................................... 3 3. Measurements ..................................................................................................................... 4 3.1. Measurement layout .................................................................................................... 4 3.2. Measurement procedure .............................................................................................. 5 4. Measurement Results.......................................................................................................... 6 4.1. Insertion Loss results ................................................................................................... 7 5. Analysis .............................................................................................................................. 8 5.1. Comparison models ..................................................................................................... 8 5.1.1. 5.1.2. Maekawa model ....................................................................................................... 8 Kurze and Anderson model ................................................................................... 11

5.2. Prediction model comparison .................................................................................... 11 5.3. Scaled up version of Maekawa and Kurze and Anderson models ............................ 12 6. Conclusions ...................................................................................................................... 13 6.1. Observations from measurements ............................................................................. 13 6.2. Prediction model conclusions.................................................................................... 15 6.3. Measurement inaccuracies ........................................................................................ 16 6.4. Possible designs......................................................................................................... 17 6.5. Further Work ............................................................................................................. 18 References ................................................................................................................................ 19 Appendices .............................................................................................................................A-1 Appendix A:Measurement results..........................................................................................A-1 Sound Pressure Levels measurements, dB.............................................................................A-1 Insertion Loss measurements,dB ...........................................................................................A-2 Appendix B:Matlab Scripts for Maekawa and Kurze and Anderson Methods .................. B-1 Appendix C:Calculation examples ........................................................................................ C-1 Maekawa Model..................................................................................................................... C-1 i

IOA Diploma 2013-14

Student Number: 77144104

Kurze and Anderson Model ................................................................................................... C-4

ii

IOA Diploma 2013-14

Student Number: 77144104

List of Figures

Figure 3-1 Side View of measurement .................................................................................... 4 Figure 3-2 Photograph of measurement layout ........................................................................ 4 Figure 4-1 Insertion loss results, dB ....................................................................................... 7 Figure 5-1-Path difference ......................................................................................................... 9 Figure 5-2-I.L predictions (Maekawa, 1968) and (Kurze and Anderson, 1971) ..................... 11 Figure 5-3- Scaled up prediction.............................................................................................. 12 Figure 6-1- Noise paths of a barrier ......................................................................................... 15 Figure 6-2 Broadband overall Insertion Loss, dB ................................................................. 16 Figure 6-3 Possible barrier design to increase I.L ................................................................ 17 Figure C-1-Dimensions of measured barrier ......................................................................... C-1

ii

IOA Diploma 2013-14

Student Number: 77144104

List of Tables
Table A-1 SPL (No Barrier) ...............................................................................................A-1 Table A-2 SPL (100mm barrier, 1m from Mic/Loudspeaker) ............................................A-1 Table A-3 SPL (200mm barrier, 1m from Mic/Loudspeaker) ............................................A-1 Table A-4 SPL (300mm barrier, 1m from Mic/Loudspeaker) ............................................A-1 Table A-5 SPL (100mm Barrier, 100mm from Loudspeaker Only) ..................................A-1 Table A-6 SPL (100mm Barrier, 200mm from Loudspeaker Only) ..................................A-2 Table A-7 SPL (100mm Barrier, 300mm from Loudspeaker Only) ..................................A-2 Table A-8 SPL (100mm Barrier, 400mm from Loudspeaker Only) ..................................A-2 Table A-9 SPL (100mm Barrier, 500mm from Loudspeaker Only) ..................................A-2 Table A-10 Insertion Loss (100mm barrier, 1m from Mic/Loudspeaker) ..........................A-2 Table A-11 Insertion Loss (200mm barrier, 1m from Mic/Loudspeaker) ..........................A-2 Table A-12 Insertion Loss (300mm barrier, 1m from Mic/Loudspeaker) ..........................A-3 Table A-13 Insertion Loss (100mm Barrier, 100mm from Loudspeaker Only).................A-3 Table A-14 Insertion Loss (100mm Barrier, 200mm from Loudspeaker Only).................A-3 Table A-15 Insertion Loss (100mm Barrier, 300mm from Loudspeaker Only).................A-3 Table A-16 Insertion Loss (100mm Barrier, 400mm from Loudspeaker Only).................A-3 Table A-17 Insertion Loss (100mm Barrier, 500mm from Loudspeaker Only).................A-3 Table C-1- Insertion Loss (Maekawa Model) ........................................................................ C-3 Table C-2-Insertion Loss (Kurze and Anderson Model) ....................................................... C-4

iii

IOA Diploma 2013-14

Student Number: 77144104

Nomenclature
I.L Mic SLM Insertion Loss, dB Microphone Sound Level Meter

iv

IOA Diploma 2013-14

Student Number: 77144104

1.

Introduction

Sound pressure level measurements of a plywood acoustic noise barrier were undertaken within the anechoic chamber at Leeds Metropolitan University, Portland Crescent, Leeds, West Yorkshire, LS1 3HB. The supplied noise signal was that of white noise. This was amplified via a power amplifier and relayed through a loudspeaker. Various configurations were tested. First of all sound pressure levels were measured without a barrier in place .This measurement was then repeated but for barriers of heights 100mm, 200mm and 300mm of which all had source to barrier and receiver to barrier distances of 1m and heights of 0.3m from floor to receiver and source. To understand the effects of having the noise source closer/further away from the barrier a series of measurements were then undertaken using the 100mm barrier. These involved moving the loudspeaker to as close as 100mm and as far as 500mm away from the barrier (in 100mm stages). In hindsight the 100mm barrier should not have been chosen given that there was a direct line of sight between the source and receiver. However the general observations of acoustical barrier theory were observed i.e. Height, source to receiver and diffraction effects. Given this the 300mm barrier was chosen for subsequent analysis. This was chosen due to there being no line of sight between the source and receiver. The I.L measurements of the 300mm barrier (derived from the sound pressure level measurements) were compared with the insertion loss prediction models of (Maekawa, 1968) and (Kurze and Anderson, 1971). It was found that in octave bands correlation was poor whilst broadband insertion loss gave good agreement.

IOA Diploma 2013-14

Student Number: 77144104

2.
2.1.

Instrumentation & equipment

Sound level meter
Norsonic 118 1 31310 Not specified

Manufacturer Model Class type Serial Number Calibration due 2.2.

Calibrator
Norsonic 4231 2389088 Not specified

Manufacturer Model Serial Number Calibration due 2.3.

Microphone
Norsonic Unknown Unknown Not specified

Manufacturer Model Serial Number Calibration due

2.4.

Equipment
Wharfedale Speaker (White Noise Source Speaker) 30D 6 Unknown Not specified NTI Minarator (Signal Generator) MR Pro Not specified Not specified 2

Manufacturer Model Serial Number Calibration due Manufacturer Model Serial Number Calibration due

IOA Diploma 2013-14

Student Number: 77144104

Manufacturer Model Serial Number Calibration due

Bruel & Kjaer Power Amplifier Unknown Unknown Not specified

2.5.

Materials
10mm thick, 100m plywood board 10mm thick, 200m plywood board 10mm thick, 300m plywood board

Material 1 Material 2 Material 3

IOA Diploma 2013-14

Student Number: 77144104

3.
3.1.

Measurements
Measurement layout

Sound Source

Microphone

Barrier

Figure 3-1 Side View of measurement

Figure 3-2 Photograph of measurement layout

IOA Diploma 2013-14 3.2.

Student Number: 77144104

Measurement procedure

All measurements sound pressure level measurements were undertaken within the anechoic chamber at Leeds Metropolitan University, Portland Crescent, Leeds, West Yorkshire, LS1 3HB. The supplied noise signal was that of white noise. This was amplified via a power amplifier and relayed through a loudspeaker. All equipment was inspected prior to and after measurements. Measurements were made at various distances from the barrier. The first measurements involved measuring the sound pressure level with no barrier present. This was to allow for Insertion Losses (I.L) to be computed. Each barrier size was then tested at a distance of 1m from both the loudspeaker (source) and the Sound Level Meter (SLM), (receiver). To understand the effects of having the noise source closer/further away from the barrier a series of measurements were then undertaken using the 100mm barrier. These involved moving the loudspeaker to as close as 100mm and as far as 500mm away from the barrier (in 100mm stages). Throughout the measurement process both the SLM and loudspeaker were mounted on a tripod and were not moved. The measurement duration was 10 seconds. Prior to the commencing measurements, the calibration of the SLM described in Section 2.1 was undertaken using the calibrator (referred to in Section 2.2). The calibrator level was measured correctly at 114 dB(A). The calibration of the SLM was re-checked successfully on completion of the measurements.

IOA Diploma 2013-14

Student Number: 77144104

4.

Measurement Results

From the measured sound pressure levels (with and without the barrier present) I.L values were computed. These are illustrated within Figure 4-1. For all measurement results the reader is referred to Appendix A:

IOA Diploma 2013-14

Student Number: 77144104

4.1.

Insertion Loss results

Figure 4-1 Insertion loss results, dB 7

IOA Diploma 2013-14

Student Number: 77144104

5.
5.1.

Analysis
Comparison models

To understand if empirical insertion loss prediction models such as (Maekawa, 1968) and (Kurze and Anderson, 1971) are comparable with the computed insertion loss values of Figure 4-1 a series of simple calculations were undertaken. For ease of calculation all models were computed within a simple Matlab script1 and only one configuration (300m barrier at 1m from loudspeaker and mic) was chosen for analysis. The computed frequency range was 63Hz to 8000Hz.

5.1.1. Maekawa model (Maekawa, 1968) proposed two formulas for predicting insertion loss, dependent on receiver position. The first formula (Equation (5.1)) was dependent on the receiver being located within the illuminated zone (hence there being a direct line of sight between source and receiver) with the second (Equation (5.2)) being dependant on the receiver being within the shadow zone (no direct line of sight). An illustration of these zones is given within Figure 5-1.

I.L=10log10 ( 3-20N ) dB I.L=10log10 ( 3+20N ) dB

(5.1)

(5.2)

Maekawa presented his results through the way of a curve representing I.L (in dB) as a function of the Fresnels number N. The Fresnels number being a ratio of the path difference (the difference in distance between the diffracted path and the direct path of sound to the half wavelength).

For the complete Matlab script the reader is referred to Appendix B:

IOA Diploma 2013-14

Student Number: 77144104

Barrier

Illuminated zone Shadow zone b Receiver

Source

Hb c Hr

Hs ds dr Figure 5-1-Path difference

The Fresnels number Nis computed from Equation (5.3)

2 N =

(5.3)

Where is the path difference which is given by Equation (5.4)

= ( a + b) c

(5.4)

IOA Diploma 2013-14

Student Number: 77144104

From Figure 5-1 the path difference may be calculated by using Pythagoras theorem for three right angled triangles with a, b and c as the hypotenuse (Smith et al., 2011, p.47).

2 2 a = ( H b H s ) + ( d s )

0.5

(5.5)

2 2 b = ( H b H r ) + ( d r )

0.5

(5.6)

2 2 c = ( H r H s ) + ( d s + d r )

0.5

(5.7)

The Wavelength in metres is easily calculated from:

c f

(5.8)

Where c is the speed of sound in metres per second and f is the frequency of interest, Hz

Maekawa assumed the barrier was semi-infinite, thin and rigid and that all predictions are free field (Everaert et al., 2010). Due to this Equations (5.1) and (5.2) only include reflections on the barrier and diffraction on the top of the barrier. He further assumed that the noise barrier was of finite length and of a sufficient thickness and hence ground or wall transparency is not accounted for.

10

IOA Diploma 2013-14 5.1.2. Kurze and Anderson model

Student Number: 77144104

As with (Maekawa, 1968) the prediction model of (Kurze and Anderson, 1971) uses the Fresnel number to calculate I.L, however it is slightly different in that it introduces a tanh function and a 5dB constant .

I.L = 5 + 20 log10

2 N tanh 2 N

(5.9)

Where N is the Fresnels number computed from Equation (5.3)

5.2.

Prediction model comparison

Figure 5-2 is a plot of the computed values of the (Maekawa, 1968) and (Kurze and Anderson, 1971) prediction models. These values are plotted against the measured values from the anechoic chamber. For an in-depth explanation/example of how these values were computed the reader is referred to Appendices C1 and C2.

Figure 5-2-I.L predictions (Maekawa, 1968) and (Kurze and Anderson, 1971)

11

IOA Diploma 2013-14 5.3.

Student Number: 77144104

Scaled up version of Maekawa and Kurze and Anderson models

To recreate a real barrier situation the values used for the computations given within Appendices C1 and C2 were multiplied by a factor of 10 i.e. the results computed for 10000Hz describe the real life situation at 1000Hz. The values of this computation are illustrated within Figure 5-3.

Figure 5-3- Scaled up prediction

12

IOA Diploma 2013-14

Student Number: 77144104

6.
6.1.

Conclusions
Observations from measurements

The observations that maybe drawn from the measurements are as follows: Insertion Loss increases with height of barrier. Insertion Loss increases when the barrier is closer to the sound source/receiver. At frequencies below 250Hz the sound is able to diffract around the barrier. This is due the dimensions of the barrier being small compared to its wavelength i.e. at 250Hz the wavelength is approximately 1.36m which is larger than any dimension of the barrier. At higher frequencies (>4000Hz) the sound appears to diffract over the barrier. This causes the I.L performance to decrease. Previous experiments have indicated that the anechoic chamber is not completely freefield i.e. 6dB per doubling of distance does not occur. This is likely to have an effect on the measured I.L values. It was noted that there was a reflective surface under the barrier i.e. anechoic chamber is semi-anechoic. This may have increased the measured levels by up to 3dB. It was unclear what the sound reduction index of the barrier material was. From (Watson and Downey, 2008, p.85) the level of sound transmitting through the barrier should be at least 10dB lower than the level of sound passing around the barrier i.e. the mass per unit area of the barrier material should be large enough to block the required amount of broadband noise. Again this is likely to have an effect on the measured I.L values. The minimum mass maybe calculated from a formula given by (Watson and Downey, 2008, p.85):
( broadband attenution+10 14 ) ) minmass = 3 10 (

(6.1)

13

IOA Diploma 2013-14

Student Number: 77144104

For example if one wished for 20dB of broadband attenuation the minimum mass per unit area required would be as follows:

( -20+10 14 ) ) minmass = 3 10 (

( (
(

( -10 14 ) ) = 3 10 (

= 3 10

( 0.71)

= 15.4 kg m 2

Whilst undertaking the measurements it was noted that there was potential paths at both the sides and underneath the barrier.

14

IOA Diploma 2013-14

Student Number: 77144104

Further to the above there are several paths that the sound can take which will decrease the amount of I.L afforded by the barrier. Some of these paths are illustrated within Figure 6-1 which was extracted from (pcfarina.eng.unipr.it, n.d)

Diffraction at upper of the screen: i.e. the barrier is very long and the effective heights not too high

Diffraction at sides of the screen: when the distance between the source and the screen edge is less than 5 times the effective height

Passing through the screen

Reflection over other surfaces present in proximity: when the location allows it

Figure 6-1- Noise paths of a barrier

6.2.

Prediction model conclusions

With respect to the I.L models outlined within section 5 there was not a good octave band correlation achieved between the measured and predictive models (Maekawa, 1968) and (Kurze and Anderson, 1971). This maybe in part due to the inaccuracies outlined above. However as we can see from Figure 6-2 the agreement between the measured broadband I.L and the models of (Maekawa, 1968) and (Kurze and Anderson, 1971) is excellent. The measured broadband I.L was 23.4dB where as the model of (Kurze and Anderson, 1971) was computed at 24.7dB with the (Maekawa, 1968) model being 25.1dB. For the worst case (Maekawa model) this is an increase of 7.3% from the measured I.L.

15

IOA Diploma 2013-14

Student Number: 77144104

Figure 6-2 Broadband overall Insertion Loss, dB The scaled up model predicted a broadband I.L of 35.2dB. For a thin rigid barrier this is not realistic given that (Smith et al., 2011, p.47) places a limit on barrier I.L at 15-20dB.

6.3.

Measurement inaccuracies

As the comparison is between two measurements taken at the same location at relatively the same time the effect of the position and reading on the SLM is minimal. The SLM does however have an inherent inaccuracy of approximately 1.1dB, hence no measurement can be taken to be better than this. Furthermore the fact that only one measurement at one position was carried out constrains the conclusions that can be drawn. To improve measurement inaccuracies it would be beneficial to measure at varying positions. Directionality effects were also not considered. Further to this whilst undertaking the measurements there were three adults within the anechoic chamber. Therefore it is likely that the absorption properties of these adults will have had an effect on the measurements. Effects of temperature, humidity, air pressure were assumed to be negligible.

16

IOA Diploma 2013-14 6.4.

Student Number: 77144104

Possible designs

Important design parameters can be drawn from this experiment these being: The more the height of the barrier the greater the I.L. Although there is a limit to this.(Munjal, 2013, p.145) puts this limit at 5m. For greater I.L the barrier should be placed close to the sound source or receiver i.e. located within the shadow zone. The mass per unit area of barrier should be of a sufficient mass. All flanking/diffraction paths shall be minimised.

Typical designs extracted from (Munjal, 2013, p.147) that will increase I.L are: a. Thicker barrier b. Double barrier c. Barrier with flat cap d. Barrier with forked end

(a)

(b)

(c)

(d)

Figure 6-3 Possible barrier design to increase I.L

17

IOA Diploma 2013-14 6.5.

Student Number: 77144104

Further Work

To verify the conclusions and design guidance given above one may wish to undertake further testing. This could include testing different barrier materials and configurations i.e. adding absorption to barrier surfaces. Other prediction models maybe analysed to see if they are more accurate than those of (Maekawa, 1968) and (Kurze and Anderson, 1971). For this one may wish to consult (Li and Wong, 2005) who have reviewed several I.L prediction models and compared these to actual measurements.

18

IOA Diploma 2013-14

Student Number: 77144104

References
Everaert, V. , Progneaux, A. and Bouillard, Ph. (2010). Simulation of Sound Wave Propagation in Presence ofNoise Barrier by using the Partition of the Unity FiniteElements Method : a comparison with analytical methodsin 2D cases. [Online]. Available from: https://www.isma-isaac.be/past/conf/isma2010/proceedings/papers/isma2010_0182.pdf. [Accessed 5th January 2014] Kurze, U.J. and Anderson, G.S. (1971). Sound attenuation by barriers. Applied Acoustics. [Online]. 4(1), pp.35-53. Available from: http://www.sciencedirect.com/science/article/pii/0003682X71900247. Li, K.M. and Wong, H.Y. (2005). A review of commonly used analytical and empirical formulae for predicting sound diffracted by a thin screen. Applied Acoustics. [Online]. 66(1), pp.45-76. Available from: http://www.sciencedirect.com/science/article/pii/S0003682X04000842. Maekawa, Z. (1968). Noise reduction by screens. Applied Acoustics. [Online]. 1(3), pp.157173. Available from: http://www.sciencedirect.com/science/article/pii/0003682X68900200. Munjal, L.M. (2013). Acoustics of Rooms, Partitions, Enclosures and Barriers. In: Munjal, L.M. ed. Noise and Vibration Control. London: World Scientific Publishing, pp.132-167. pcfarina.eng.unipr.it. (n.d). Noise Barriers. [Online]. Available from: http://pcfarina.eng.unipr.it/Public/Acoustics-Course/.../08_ghiacci_222030.doc. [Accessed 12th January 2014] Smith, J.B. , Peters, J.R. and Hollins, M. (2011). Acoustics and Noise Control. 3rd ed. Harlow: Pearson Education Limited. Watson, R. and Downey, O. (2008). Little Red Book of Acoustics 2nd ed. Great Britain: Blue Tree Acoustics.

19

IOA Diploma 2013-14

Student Number: 77144104

Appendices Appendix A: Measurement results

Sound Pressure Levels measurements, dB Table A-1 SPL (No Barrier) 63 40.4 125 37.6 250 51.2 Octave Band, Hz 500 1000 2000 55 63.6 70.7 4000 81.2 8000 72.9

No Barrier,dB

Table A-2 SPL (100mm barrier, 1m from Mic/Loudspeaker) 63 37.5 125 37.4 250 52.5 Octave Band, Hz 500 1000 2000 54.9 61.6 66.2 4000 68.8 8000 61.2

100mm Barrier, dB

Table A-3 SPL (200mm barrier, 1m from Mic/Loudspeaker) 63 35.7 125 36.8 250 52.4 Octave Band, Hz 500 1000 2000 53.1 57.8 58.2 4000 64.1 8000 64.2

200mm Barrier, dB

Table A-4 SPL (300mm barrier, 1m from Mic/Loudspeaker) 63 35.7 125 36.8 250 52.4 Octave Band, Hz 500 1000 2000 53.1 57.8 58.2 4000 64.1 8000 64.2

300mm Barrier, dB

Table A-5 SPL (100mm Barrier, 100mm from Loudspeaker Only) 63 40.9 125 40.7 250 47 Octave Band, Hz 500 1000 2000 40.9 41.5 48.5 4000 53.2 8000 49.4

100mm Barrier, dB

A-1

IOA Diploma 2013-14

Student Number: 77144104

Table A-6 SPL (100mm Barrier, 200mm from Loudspeaker Only) 63 40.4 125 34.3 250 48 Octave Band, Hz 500 1000 2000 44.5 44.3 49.2 4000 57.1 8000 52

100mm Barrier, dB

Table A-7 SPL (100mm Barrier, 300mm from Loudspeaker Only) 63 34.9 125 34 250 49.3 Octave Band, Hz 500 1000 2000 46.1 47 50.9 4000 56.6 8000 53.3

100mm Barrier, dB

Table A-8 SPL (100mm Barrier, 400mm from Loudspeaker Only) 63 38.5 125 35.9 250 50.2 Octave Band, Hz 500 1000 2000 47.8 47.7 51.6 4000 54 8000 53.5

100mm Barrier, dB

Table A-9 SPL (100mm Barrier, 500mm from Loudspeaker Only) 63 33.6 125 35.4 250 50.9 Octave Band, Hz 500 1000 2000 49 50.3 50.1 4000 56.1 8000 53.5

100mm Barrier, dB

Insertion Loss measurements,dB Table A-10 Insertion Loss (100mm barrier, 1m from Mic/Loudspeaker) 63 2.9 125 0.2 250 -1.3 Octave Band, Hz 500 1000 2000 0.1 2 4.5 4000 12.4 8000 11.7

100mm Barrier, dB

Table A-11 Insertion Loss (200mm barrier, 1m from Mic/Loudspeaker) 63 4.7 125 0.8 250 -1.2 Octave Band, Hz 500 1000 2000 1.9 5.8 12.5 4000 17.1 8000 8.7

200mm Barrier, dB

A-2

IOA Diploma 2013-14

Student Number: 77144104

Table A-12 Insertion Loss (300mm barrier, 1m from Mic/Loudspeaker) 63 4 125 0.8 250 -0.2 Octave Band, Hz 500 1000 2000 4.1 11.3 20 4000 18.9 8000 13.4

300mm Barrier, dB

Table A-13 Insertion Loss (100mm Barrier, 100mm from Loudspeaker Only) 63 -0.5 125 -3.1 250 4.2 Octave Band, Hz 500 1000 2000 14.1 22.1 22.2 4000 28 8000 23.5

100mm Barrier, dB

Table A-14 Insertion Loss (100mm Barrier, 200mm from Loudspeaker Only) 63 0 125 3.3 250 3.2 Octave Band, Hz 500 1000 2000 10.5 19.3 21.5 4000 24.1 8000 20.9

100mm Barrier, dB

Table A-15 Insertion Loss (100mm Barrier, 300mm from Loudspeaker Only) 63 5.5 125 3.6 250 1.9 Octave Band, Hz 500 1000 2000 8.9 16.6 19.8 4000 24.6 8000 19.6

100mm Barrier, dB

Table A-16 Insertion Loss (100mm Barrier, 400mm from Loudspeaker Only) 63 1.9 125 1.7 250 1 Octave Band, Hz 500 1000 2000 7.2 15.9 19.1 4000 27.2 8000 19.4

100mm Barrier, dB

Table A-17 Insertion Loss (100mm Barrier, 500mm from Loudspeaker Only) 63 6.8 125 2.2 250 0.3 Octave Band, Hz 500 1000 2000 6 13.3 20.6 4000 25.1 8000 19.4

100mm Barrier, dB

A-3

IOA Diploma 2013-14

Student Number: 77144104

Appendix B:

Matlab Scripts for Maekawa and Kurze and Anderson Methods

Octave Band Frequency's , Hz

f=[63 125 250 500 1000 2000 4000 8000];

Prompts for required inputs

prompt = 'Enter height of source from ground in m '; Hs = input(prompt); prompt = 'Enter height of receiver from ground in m '; Hr = input(prompt); prompt = 'Enter height of barrier from ground in m '; Hb = input(prompt); prompt = 'Enter distance of source from the barrier in m '; ds = input(prompt); prompt = 'Enter distance of receiver from the barrier in m '; dr = input(prompt);

Speed of Sound
C=340; %Default 340 m/s%

Computations
%Wavelength, m% lamda=(C./f); %Pythrogoras% a=((Hb-Hs)^2+(ds)^2)^0.5; b=((Hb-Hs)^2+(dr)^2)^0.5; c=((Hr-Hs)^2+(ds+dr)^2)^0.5; %Path Difference, m% PD=(a+b)-c; %Fresnel Number% N=(2.*PD./lamda);

Computed Insertion Loss (Maekawa Method)

IL_Maekawa=10.*log10(3+(20*N));

B-1

IOA Diploma 2013-14 Computed Insertion Loss (Kurze and Anderson Method)
x=sqrt(2.*pi.*N); y=tanh(x); IL_Kurze_Anderson=5+20*log10(x./y);

Student Number: 77144104

300mm Barrier, 1m from Mic/Loudspeaker (Measured Values)

Test_3_IL =[4 0.8 -0.2 4.1 11.3 20 18.9 13.4];

Plot of measured Insertion Vs Maekawa and Kurze and Anderson Methods

figure(1) clf semilogx(f,Test_3_IL ,'mo','LineWidth',2.5'), grid hold on semilogx(f,IL_Maekawa,'r--','LineWidth',2.5') semilogx(f,IL_Kurze_Anderson,'b','LineWidth',2.5') hold off xlabel(['\fontsize{14}','Octave Band Hz']); ylabel(['\fontsize{14}','Measured Insertion Loss (dB)']); set(gca,'XTick',[63,125,250,500,1000,2000,4000,8000]) set(gca,'YTick',[-1,0,5,10,15,20,25]) xlim([63 8000]); ylim([-1 25]); legend('Measured Insertion Loss for 300mm Barrier, 1m from Mic/Loudspeaker',... 'Calculated Insertion Loss,(Maekawa Method)','Calculated Insertion Loss (Kurze and Anderson Method)'... ,'Location','NorthWest');

Displaying Data
disp(['Insertion Loss (Maekawa) = ', num2str(RoundOff(IL_Maekawa,0.1)), ' dB']); disp(' '); disp(['Insertion Loss (Kurze and Anderson) = ', num2str(RoundOff(IL_Kurze_Anderson,0.1)), ' dB']); disp(' ');

B-2

IOA Diploma 2013-14

Student Number: 77144104

Appendix C:
Maekawa Model

Calculation examples

Barrier

Illuminated zone Shadow zone b Receiver

Source

0.3m c 0.1m

0.1m 1m 1m

FigureC-1-Dimensions of measured barrier One can use Equations (5.5)(5.6) and (5.7) to calculate the path difference,
0.5

2 2 a = ( H b H s ) + ( d s )

2 2 = ( 0.3 0.1) + (1)

0.5

= 1.02m

2 2 b = ( H b H r ) + ( d r )

0.5

2 2 = ( 0.3 0.1) + (1)

0.5

= 1.02m

C-1

IOA Diploma 2013-14

Student Number: 77144104

2 2 c = ( H r H s ) + ( d s + d r ) 0.5

2 2 = ( 0.1 0.1) + (1+ 1)

0.5

= 4m

Now using Equation (5.4) we calculate the path difference

= ( a + b) c
= (1.02 + 1.02 ) 4 = 0.04m

If we assume the speed of sound in air to be 340 m/s then the wavelength, maybe calculated from Equation (5.8). For this example the frequency of interest, f is 1000 Hz

c f

340 = = 0.34m 1000

One can then compute the Fresnel Number, N from Equation (5.3)

2 2 0.04 N = = = 0.235 0.34

C-2

IOA Diploma 2013-14

Student Number: 77144104

As there is no line of sight the insertion loss at 1000 Hz is computed using Equation (5.2)

I.L=10log10 ( 3+20N ) = 10log10 ( 3+ ( 20 0.235 ) ) 8.8dB

The values at all other frequencies are given within TableC-1

TableC-1- Insertion Loss (Maekawa Model) Octave Band, Hz 500 1000 2000 7.3 8.8 10.9

Calculated Insertion Loss, dB

63 5.2

125 5.5

250 6.2

4000 13.4

8000 16

C-3

IOA Diploma 2013-14 Kurze and Anderson Model

Student Number: 77144104

If we use the value of N from above the insertion loss (at 1000 Hz) for the Kurze and Anderson model maybe calculated from Equation (5.9)

I.L = 5 + 20 log10

2 N tanh 2 N 2 0.235

= 5 + 20 log10

tanh 2 235 1.21 tanh (1.21)

= 5 + 20 log10

1.21 = 5 + 20 log10 0.838 = 5 + 3.2 = 8.2dB

The values at all other frequencies are given within Table C-2

Table C-2-Insertion Loss (Kurze and Anderson Model) Octave Band, Hz 500 1000 2000 6.8 8.2 10.2

Calculated Insertion Loss, dB

63 5.3

125 5.5

250 6.0

4000 12.8

8000 15.7

C-4