Characterizing noise barriers: SOPRANOISE final report and outcomes Jean-Pierre Clairbois 1 A-Tech / Acoustic Technologies Boulevard Jamar 19 box A 0.01, 1060 Brussels, Belgium Massimo Garai, Paolo Guidorzi 2 University of Bologna, DIN Department of Industrial Engineering Viale Risorgimento 2, 40136 Bologna, Italy Wolfram Bartolomaeus, Michael Chudalla, Fabio Strigari 3 Federal Highway Research Institute, Brüderstraße 53, 51427 Bergisch Gladbach, Germany Marco Conter, Andreas Fuchs 4 AIT Austrian Institute of Technology Giefinggasse 4, 1210 Vienna, Austrian Christophe Nicodeme 5 ERF - European Union Road Federation Rue Belliard20 Box 7, 1040 Brussels, Belgium ABSTRACT SOPRANOISE targets a three-step approach for assessing the in-situ intrinsic acoustic performance of road/railway noise barriers in regard to airborne sound insulation and sound reflection under direct sound field conditions. This approach goes from the in-situ inspections method (targeting at the acoustic effect in the far-field) to the standardized measuring methods EN 1793-5 and 6, with “quick and safe” measuring methods in between. Papers about SOPRANOISE have already been presented in 2020 and 2021 as progress reports. The research ended in March 2022 and the final outcomes are now presented. The main part of the pending work has been done within Work Package 4 (WP4). Herein, a new “quick and safe” method was designed and validated both in the lab and in situ. The new measuring equipment is portable, lightweight and uses a linear antenna; the measurement itself is fully automated and the results show a very good correlation with full EN 1793-5/-6 results. WP4 results are particularly successful and, together with the in-situ inspection, those methods are ready to be submitted to the relevant CEN/TC226/WG6 working group to draft a new framework on noise barriers assessment. All main reports are public available on the SOPRANOISE and CEDR webpages. 1 jpc@atech-acoustictechnologies.com 2 massimo.garai@unibo.it , paolo.guidorzi@unibo.it 3 bartolomaeus@bast.de , chudalla@bast.de , strigari@bast.de 4 marco.conter@ait.ac.at , andreas.fuchs@ait.ac.at 5 c.nicodeme@erf.be

inter noise 21-24 AUGUST SCOTTISH EVENT CAMPUS GLASGOW

1. INTRODUCTION

Initiated by a call from the Conference of European Directors of Roads (CEDR) about “Noise and Nuisance”, the SOPRANOISE (Securing and Optimizing the Performance of Road traffic noise barriers with New methOds and In- Situ Evaluation) project started in December 2019 and ended in March 2022: its aim was to improve the assessment of the acoustic characteristics of noise barriers installed along roads and/or railways. Its first results have been presented as progress reports in 2020 [1] and in 2021 [2]. The pending tasks were Task 4.3 “Validation of quick methods by comparison with full methods in-situ” and Task 4.4 “Report on the new quick methods”; as well as Task 5.3 “Assessment of the intrinsic performance” and Task 5.5 “Guidelines and scientific report”. This paper is concentrated on the successful new quick method being now named the “SOPRA” method, and on its insertion within the SOPRANOISE 3-step approach that will soon be proposed to European standardization. This approach greatly improves and facilitates the assessment of the acoustic performances of noise barriers, from the simplest but inexpensive “in-situ inspections”, up to the most complex but accurate measurements according to “full EN 1793-5 [3] and 1793-6 [4]”. The new SOPRA method fills the gap in between. With this approach, it is now possible to establish relevant information throughout the whole lifecycle of noise barrier projects to improve their use. All the deliverables of the SOPRANOISE research are available on the website platform https://www.enbf.org/sopranoise/outcome/ .

2. QUICK AND SAFE METHODS ALONGSIDE ROADS

Work package 4 is managed by the University of Bologna (UNIBO). Task 4.1 and Task 4.2 have already been presented in [2]: therein, the new method was developed and tested in the laboratory. From now on, this method is named the “SOPRA” method. The next step was, still continuing to improve the method, to validate it alongside real highway sites: this has been done in Task 4.3.

2.1. The SOPRA method The SOPRA measurement procedure is similar to that of the standards EN 1793-5 [3] and 1793-6 [4]. It integrates the necessary simplifications to apply those “full methods” in a much simpler, quick and much less expensive way with the least possible changes but keeping the fundamental principles of the EN standard methods. Main changes are the use of a vertical 6 microphones array (keeping a 0,4 m distance between two successive microphones) instead of the 9 microphones square array (0,4 by 0,4 m), and a completely redesigned measuring equipment that can widely simplify the in-situ measurements (see Figure 1).

ny ba \ — UNIBO UNIBO _

Figure 1: UNIBO (left) and AIT (right) equipment, loudspeaker units and microphone line arrays.

2.2. UNIBO measuring equipment The UNIBO equipment has been already presented in [2], it has the following characteristics: it is lightweight, portable, autonomous (battery operated), operates with wireless connections (see Figure 2), designed for both sound absorption and airborne sound insulation measurements, with a new control and processing unit which is easy and quick to use with interactive controls (see Figure 3), including environmental conditions recording (air, temperature, wind, etc.).

Figure 2: UNIBO wireless system at the loudspeaker unit and at the control and processing unit.

Figure 3: UNIBO new control & processing unit.

2.3. AIT measuring equipment The AIT equipment respects the same needs/characteristics as the UNIBO one, being very similar but not identic: AIT control unit uses a Raspberry Pi 4, while the UNIBO equipment uses a Teensy 4.1 system / Arm Cortex-M7 processor; both setups use wireless connections. For the loudspeakers and microphone arrays, Figure 1 already showed the two systems.

2.4. In-situ tests UNIBO carried out in-situ measurement sessions along the A22 motorway (Autostrada del Brennero / Brennerautobahn) in Italy. Two different kinds of noise barriers have been tested there: one metal noise barrier that is about 2 years old but in perfect conditions still looking as good as new, and another one in timber that is about seven years old, but on which ageing shows worse effects.

Metal noise barrier The metal noise barrier tested is sound absorptive and 5 m high, with a post every 2,67 m (see Figure 4).

Figure 4: the metal noise barrier tested by UNIBO along the A22 motorway. In about 6 hours, 22 SOPRA sound absorption tests (20 fields plus 2 repeated tests) and 11 SOPRA airborne sound insulation tests (10 fields plus 1 repeated test) have been successfully carried out. For the sound absorption tests, analysis has been done for all the 6 microphones (from M1 at 1,2 m, up to M6 at 3,2 m above the ground) and processed in three different ways: i) with all the six microphones, ii) with only five microphones M1 to M5, and iii) with only four microphones from M2 to M5, while the loudspeaker height was kept at 2 m above the ground for all the tests. The analysis of each of those three microphone combinations and their comparison with “full EN” tests clearly shows that, for a noise barrier with a height in the range of 4 to 5 m, it is better to exclude the bottom microphone (M1), which is too much influenced by the reflection on the ground, and the top microphone (M6), which is too much influenced by the diffraction at the barrier top edge. Figure 5 shows the “quick reflection index” values RI Q for 1/3 rd octave bands for each of the twenty-two measurements along the noise barrier, as well as for the “full EN” results of tests carried out on Field 1. In addition, the “tolerance curves” obtained by adding/subtracting to/from the “full EN” results the expanded measurement uncertainty at 95% confidence level, U 95 , are also presented (dotted black curves).

lo, dB 1,00 0,80 0,60 0.40 0,20 0,00

Figure 5: metal noise barrier, 1/3 rd octave “quick reflection index” RI Q per field.

20 -0,40 200 250 315 400 500 630 800 1000 1250 1600 2000 2500 3150 4000 5000 Frequency, Hz Field 1 Field 1 rep. Field 2 Field 3 Field 4 Field S FieldS rep. Field 6 Field 7 Field 8 Field 9 Field 10 Field 11 Field 12 Field 13 Field 14 Field 15 Field 16 Field 17 Field 18 —Field 19 —Field 20 Field 1 fullEN ~~ ~Field 1 EN -U9S ~~ ~Field 1 EN + U95

Finally, the single number ratings DL RIQ have been derived for each field from those 1/3 rd octave band results: the results show that, along the noise barrier, DL RIQ varies from 6,9 to 12,0 dB, what is a quite important information that is only available when the measurements are carried out at numerous fields, while it is much more difficult to conclude if the only tested elements were either the one with 6,9 dB or the one with 12,0 dB value. In the same, for the airborne sound insulation tests, analysis has been done for all the 6 microphones (from M1 at 1,2 m, up to M6 at 3,2 m above the ground) and processed in three different ways: i) with all six microphones, ii) with only five microphones M1 to M5, and iii) with only four microphones from M2 to M5, while the loudspeaker height was kept at 2 m above the ground for all the tests. The analysis of each of those three microphone combinations and their comparison with “full EN” tests done three months before clearly shows again (as for sound absorption) that, for a noise barrier with a height in the range of 4 to 5 m, it is better to exclude the bottom microphone (M1), which is too much influenced by the ground, and the top microphone (M6), which is too much influenced by the top edge, and this even more for airborne sound insulation. Figure 6 shows the “quick airborne sound insulation index” values SI Q by 1/3 rd octave bands for each of the ten fields (plus the repeated Field 3), as well as for the “full EN” results of tests carried out on Field 3. In addition, the “tolerance curves” obtained by adding/subtracting to/from the “full EN” results the expanded measurement uncertainty at 95% confidence level, U 95 , are also presented (dotted black curves).

55,0 cry Slay

Figure 6: metal noise barrier, 1/3 rd octave “quick airborne sound insulation index” SI Q per field. As for sound absorption, the single number ratings DL SIQ have been derived for each field from those 1/3 rd octave band results: the results show that, all along the tested zone, DL SIQ varies from 27,9 to 32,4 dB: this is again an important information on how the airborne sound insulation can vary along the noise barrier length. The SOPRA method allows to carry out more measurements that can help to better understand the performance of the numerous elements within long noise barriers.

Timber noise barrier The timber noise barrier tested is sound absorptive, it is about seven years old, while apparent defects already appeared for airborne sound insulation; the barrier height is 3,5 m and it has a post every 2,67 m (see Figure 7).

Lil a 15,0 10,0 5,0 0.0 200 250 315 400 $00 630 800 1000 1250 1600 2000 2500 3150 4000 5000 Frequency, Hz —Field 1 —Field 2 —Field 3 ——Field3 rep. =——Field 4 —Fiela s —riela 6 —Fiea7 —Fields —tiela 9 —Field 10 —Field 3 fullEN = —Field 3 EN-U9S — —Field 3 EN +U9S

Figure 7: the timber noise barrier tested by UNIBO along the A22 motorway. In about 5 hours, 8 SOPRA sound absorption tests (6 fields plus 2 repeated tests) and 7 SOPRA airborne sound insulation tests (6 fields plus 1 repeated test) have been successfully carried out. For the sound absorption tests, considering the results obtained on the metal noise barrier (see above), the data have been processed only for the four microphones from M2 to M5, while the loudspeaker height was kept at 2 m above the ground. Figure 8 shows the “quick reflection index” values RI Q for 1/3 rd octave bands for each of the six fields along the noise barrier, as well as for the “full EN” results of tests carried out on Field 3. Again, the “tolerance curves” obtained by adding/subtracting to/from the “full EN” results the expanded measurement uncertainty at 95% confidence level, U 95 , are presented (dotted black curves).

er 120 1100 080 060 Rig, dB

Figure 8: timber noise barrier, 1/3 rd octave “quick reflection index” RI Q per field. The single number ratings DL RIQ have been derived for each field from those 1/3 rd octave band results: the results show that DL RIQ varies from 2,7 to 4,2 dB, what again demonstrates the usefulness of the SOPRA method to carry out more numerous tests to get a more relevant figure of the performances along long noise barriers. For the airborne sound insulation tests, the data have also been processed only for the four microphones from M2 to M5, while the loudspeaker height was kept at 2 m above the ground.

0,40 0,20 0,00 200 250 315 400 500 630 800 1000 1250 1600 2000 2500 3150 4000 $000 Frequency, Hz Field 1 Field 2 Field 3 Field 4 Field S Field 6 Field 3 rep. 1 Field 3 rep. 2 Field 3 fulleN ~~ -Field 3 EN-U9S ~ ~~ Field 3 EN + U95

Figure 9 shows the “quick airborne sound insulation index” values SI Q by 1/3 rd octave bands for each of the tested fields, as well as for the “full EN” results of tests carried out on Field 1. The “tolerance curves” obtained by adding/subtracting to/from the “full EN” results the expanded measurement uncertainty at 95% confidence level, U 95 , are also presented (dotted black curves).

Figure 9: timber noise barrier, 1/3 rd octave “quick airborne sound insulation index” SI Q per field. The single number ratings DL SIQ have been derived for each field from those 1/3 rd octave band results: the results show that, all along the tested zone, DL SIQ varies from 11,6 to 14,8 dB, what can only be found with several measurements locations instead of a limited amount of.

30,0 25,0 7 200 Sle, dB 15,0 100 > - 50 4 0.0 200 250 315 400 00 630 800 1000 1250 1600 2000 2500 3150 4000 S000 Frequency, Hz Field 1 Post 1-2 Field 2 Field 3 —Field 5 —Field 6 —Field 1 full EN Field 1 EN -U95

2.5. WP4 conclusion Thanks to the work achieved by UNIBO and the help of AIT, the new SOPRA method has been successfully designed and validated: the next step will be now to submit it to CEN TC226 (Road equipment Technical Committee) WG6 (Noise reducing devices Working Group) for standardization. The reports D4.1 [8] and D4.2 [9] contains all the details of the research that led to the SOPRA method. 3. THE SOPRANOISE 3-STEP APPROACH

3.1. Principles Now that the SOPRA method successfully fills the gap between the easiest way to assess the acoustic performance of noise barriers, i.e. the in-situ inspections, up to the most accurate but expensive one, i.e. the “full” EN 1793-5 and EN 1793-6 methods, the SOPRANOISE 3-step approach is ready to be used, while its standardization process will soon start. Figure 10 summarizes the main principles of the SOPRANOISE 3-step approach when assessing the acoustic performance of noise barriers.

Figure 10: The SOPRANOISE 3-step approach to assess the noise barriers acoustic performances.

Be tem t | SOPRANOISE 3-step approach “41 72 v3) ‘Measurements according to EN 1793.5,-6

The main principle is to use the right assessment tool for the right conclusion: using first the simplest method and, if fair conclusion cannot be achieved, then pass to the next step, carrying out further tests with higher accuracy to conclude, the last step being definitely the EN 1793-5 and EN 1793-6 methods. However, the validity of the conclusions may vary depending on what we want to do with the test results. In facts, we can have two main kinds of assessments: monitoring of the evolution of performances of existing barriers, and approval of newly installed noise barriers: Figure 11 shows the principles in case of monitoring, while Figure 12 shows the principles in case of approval .

Affected characteristic? | SOPRANOISE 3-step approach |

Figure 11: The SOPRANOISE 3-step approach applied to monitor existing noise barriers.

Reason for investigation? ISE 3-steps

Figure 12: The SOPRANOISE 3-step approach applied for existing noise barriers approval . To approve noise barrier performances, i.e. to verify if the effective performances successfully reach the specific quantified requirements, eventually the “full” EN 1793-5 and EN 1793-6 methods have to be applied. However, in-situ inspections and SOPRA quick methods can be very useful to avoid testing elements that are obviously defective and reduce the effort for the complex measurements according to EN 1793-5 and -6.

3.2. SOPRANOISE holistic approach of noise barriers projects Taking advantage of the outcomes of the SOPRANOISE research, being now able to assess the acoustic performance of noise barriers wherever and whenever, one can now consider noise projects in a new, different, more sustainable and holistic approach. In any new noise barrier project, we have to completely understand its whole process, ideally from “cradle to grave”, without forgetting any stage of its whole life cycle (LC): Figure 13 shows the SOPRANOISE holistic approach.

ONvudos * t = aa Ss Quick method

Figure 13: The SOPRANOISE holistic approach of noise barriers projects. Before noise barrier installations, one can benefit from all the existing tests already done but, from now on, applying the SOPRANOISE 3-step approach on every existing or new projects will ease the noise barrier assessment by continuously feeding the databases and improving the knowledge. At noise barriers installations, SOPRANOISE will help to choose the relevant samples and to limit EN 1793-5 and EN 1793-6 tests where appropriate. During the whole noise barriers use/lifetime, the SOPRANOISE approach is the convenient tool to monitor and/or to carry out relevant tests: maintenance plans could also benefit from a periodic assessment of the evolving acoustic performance. Finally, the SOPRANOISE approach can be used to objectively decide about the end of life of noise barriers. Report D5.2 [11] includes this approach as well as the SOPRANOISE final reports; it also includes guidelines based on the main outcomes of the whole research. 4. CONCLUSIONS

Started in December 2019, SOPRANOISE successfully ended in March 2022 with the following main outcomes:  WP2 established a comprehensive review of the significance of the EN 1793 standards (report D2.1 [5]), and a relevant and well populated database of the acoustic performances of European noise barrier products (report D2.2 [6]).  WP3 created and validated in-situ inspection protocols to assess existing noise barriers in an easy way (report D3.1 [7]).  WP4 developed and validated the new “quick & safe” SOPRA method (reports D4.1 [8] and D4.2 [9]).  WP5 reported different relevant information on the use and the acoustic performance of noise barriers (report D5.1 [10] and D5.2 [11]). SOPRANOISE gives road and railways authorities, as well as to all the noise barrier market stakeholders, the right tools to better understand and assess the noise barriers acoustic performances. 5. ACKNOWLEDGEMENTS

We gratefully acknowledge the CEDR (Conference of European Directors of Roads) that supported this research as part of their own Research Programme (CEDR TRANSNATIONAL ROAD RESEARCH PROGRAMME - Call 2018 – Noise and Nuisance – Topic B: Optimization and Securing the Performance of Noise Barriers).

Noise Barrier Planning Noise Barrier Design Noise Barrier Procurement NB End of Life ‘NB Installation Decomissioni Noise Barrier Approval (product & installation) NB Effective Use / Lifetime Noise Barrier Monitoring Noise Barrier Lifetime Tests Noise Barrier Maintenance

6. REFERENCES

1. Clairbois, J-P., Garai, M., Bartolomaeus, W., Chudalla, M., Strigari, F., Conter, M., Fuchs, A. & Nicodeme, C., SOPRANOISE: EU Research on new techniques to characterize Noise Barriers acoustic performances. Proceedings of INTER-NOISE 2020 , paper 16_4_468. Seoul, South Korea, August 2020. 2. Clairbois, J-P., Garai, M., Bartolomaeus, W., Chudalla, M., Strigari, F., Conter, M., Fuchs, A. & Nicodeme, C., Characterizing noise barriers: SOPRANOISE half-term progress report. Proceedings of INTER-NOISE 2021 , paper 16_4_468. Washington, DC, U.S.A., August 2021. 3. CEN, “EN 1793-5: Road traffic noise reducing devices - Test method for determining the acoustic performance - Part 5: Intrinsic characteristics – In-situ values of sound reflection under direct sound field conditions,” CEN. Brussels, Belgium, 2016 /AC:2018. 4. CEN, “EN 1793-6: Road traffic noise reducing devices - Test method for determining the acoustic performance - Part 6: Intrinsic characteristics – In-situ values of airborne sound insulation under direct sound field conditions,” CEN. Brussels, Belgium, 2018 +A1:2021. 5. Reiter, P., Fuchs, A. et. al., SOPRANOISE Deliverable D2.1, “Review of the physical significance of EN 1793-1, EN 1793-2, EN 1793-5 and EN 1793-6”; 2020. 6. Conter, M. et. al., SOPRANOISE Deliverable D2.2, “Final report on the main results of WP2 (including M2.1, M2.2. and M2.3) – Acoustic assessment of the intrinsic performances of noise barriers”; 2021. 7. Chudalla, M., Strigari, F. and Bartolomaeus W., SOPRANOISE Deliverable D3.1, “Final report on the main results of WP3 (including M3.1, M3.2 and M3.3) – In-situ inspection tools”; 2021. 8. Garai, M., Guidorzi, P., SOPRANOISE Deliverable D4.1, “Report on the development of the new quick methods in laboratory”; 2021 9. Garai, M., Guidorzi, P., SOPRANOISE Deliverable D4.2, “Report on the validation of the new quick methods in-situ with recommendations for proper use”; 2022 10. Clairbois, J-P., Houtave, P., Lebasi, N., Nicodème, C. and Garai, M. SOPRANOISE Deliverable D5.1, “WP5 Intermediate progress report including M5.1, M5.2 and M5.3”; 2021 11. Clairbois, J-P., Conter, M., Chudalla, M. , Strigari, F., Bartolomaeus W., Garai, M., Lebasi, N. and Nicodème, C., SOPRANOISE Deliverable D5.2, “SOPRANOISE Final report -Guidelines for NB use and scientific report”, 2022