Laboratory and full-scale experimental evaluation of the acoustic behaviour of sonic crystal noise barriers

Transcription

1 PROCEEDINGS of the 22 nd International Congress on Acoustics Phononic Crystals and Acoustic Metamaterials: Paper ICA Laboratory and full-scale experimental evaluation of the acoustic behaviour of sonic crystal noise barriers Paulo Amado-Mendes (a), Luís Godinho (a), Pedro Gil Santos (a), Alfredo G. Dias (a), Mário Martins (b) (a) ISISE, Department of Civil Engineering, University of Coimbra, Portugal, (b) I.P.C., Instituto Superior de Engenharia de Coimbra, Portugal, Abstract Sonic crystals represent a particular case of acoustic metamaterials, corresponding to periodically arranged structures, with individual scatterers regularly spaced. These structures present a particular acoustic behaviour, revealing high levels of sound attenuation in a range of frequencies known as the band-gap. By properly selecting some definition parameters of the sonic crystals, for example, among others, the type of lattice, the lattice spacing or the diameter of the scatterers, the observed band-gap can be adjusted to match a given frequency of interest. This idea was explored to study the use of timber logs, arranged in periodic ways, to build sustainable noise traffic barriers. In the present work, the different steps of the experimental evaluation, performed to validate this concept, are described and the results of some selected experimental cases are presented, corresponding to reduced-scale laboratorial setups and to a full-scale prototype analysed in outdoor conditions. Keywords: sonic crystal noise barrier, timber logs, laboratorial evaluation, full-scale prototype

2 Laboratory and full-scale experimental evaluation of the acoustic behaviour of sonic crystal noise barriers 1 Introduction In the last two decades, there has been an increasing interest in the research of noise reduction provided by sonic crystals, i.e. composite structures exposed to sound waves with its elements designed in periodic or quasi-periodic arrangements. In fact, the presence of these arrays of sonic scatterers in the fluid propagation medium leads to the appearance of band gaps, a range(s) of frequencies where the acoustic waves are not allowed to propagate, being totally reflected by the sonic crystal solid elements [1]. Martínez-Sala and co-authors [2] demonstrated experimentally the evidence of sound filtering capabilities in a real (but unplanned) sonic crystal, created by Eusebio Sempere s sculpture, with a two-dimensional periodic arrangement of steel tubes installed in the gardens of the Fundación Juan March, in Madrid. The authors noted that the periodic arrangement of the metallic rods leads to multiple constructive and destructive interferences (in a similar way to Bragg scattering/diffraction) between acoustic waves, depending on the frequency, thus attenuating the sound spectra of the transmitted waves at certain frequencies (the forbidden bands or band gaps), while not affecting others. In various works, researchers have pointed out the effect on noise reduction of this type of periodic structures, with different geometrical configurations and materials [3, 4, 5]. The particular interest on the use of sonic crystals as a noise mitigation measure has also been demonstrated in different studies [6, 7]. Recent published developments correspond to: the adoption of sound absorbent cylinders [8], the use of a periodic array of trees or recycled materials [9, 10], and the combination of sonic crystals aligned parallel to the ground with a traditional rigid noise barrier [11]. Many of the published works regarding the acoustic behaviour of sonic crystals correspond to either theoretical and analytical studies or to experimental and laboratorial approaches. The use of different numerical techniques has also been proposed for modelling acoustic waves propagation in the presence of these periodic structures, in two-dimensional (2D) and threedimensional (3D) analyses. Given their competitiveness in terms of discretization and required computational cost, efficient meshless methods, that do not need to discretize the domain or the scatterers boundaries, for example the Method of Fundamental Solutions (MFS), have been recently implemented and suggested [12, 13]. Although most of the prominent phenomena related to periodic structures are still observable in 2D analyses, in some occasions 3D approaches might be considered to capture more detailed acoustic features, despite the higher computational demands. In the present paper, the main experimental developments of a research project project SCLog developed at the University of Coimbra are described. Its aim was to define an innovative technological solution for traffic noise mitigation, based on timber logs forming sonic crystals. In this work, the developed concepts and methodologies are described, regarding the 2

3 laboratorial evaluation of the acoustic behaviour of sonic crystal structures, in terms of noise reduction and sound dispersion, and the construction and testing of a full-scale prototype. 2 Description of timber logs sonic crystal noise barriers When considering the technological use of sonic crystals, one interesting application is that of noise mitigation devices, namely the construction of acoustic barriers for traffic noise attenuation (see a conceptual preview in Figure 1a). That important acoustic source in the urban environment presents dominant frequencies that can possibly be used to adjust the band gaps that are attenuated by the sonic crystal structure. From previous research and ongoing works it is possible to conclude that the sound reduction provided by periodic structures is mainly related to two mechanisms, namely the configuration of the structure (geometrical arrangement and distance between scatterers Figure 1b) and the acoustic properties (geometry and sound absorption) of the individual elements. This type of noise barriers present practical advantages, since they are light and can be easily built, they can use less material and be cost competitive, they exhibit some level of transparency and low visual impact, and they can be adequately tuned by making the correct arrangement of scatterers, in order to control sound attenuation properties for different situations. For instance, when considering road traffic noise, the most important frequency range goes from 500 to 1500Hz, being, in most of the cases, the noise peak around 1000Hz. a) b) Figure 1: a) Computer simulation of in situ application of the sonic crystal-type barrier; b) Photo of full-scale prototype of timber log sonic crystal barrier In the scope of SCLog research project, at the University of Coimbra, a possible practical solution for reducing traffic noise was studied, which hasn t yet been explored. The use of round timber elements, with small diameters, to build the sonic crystal acoustic barrier was proposed and analysed. These elements are thoroughly available from forest thinning operations and require little processing to be incorporated in the acoustic barrier, giving rise to an interesting, sustainable and viable solution that needs to be characterized in detail, namely in what regards its acoustic behaviour. The use of small diameter timber logs taken from forest thinning operations, in structural applications, such as in roofs, building frames, bridges, among others, has already been studied world-wide [14] and, specifically in Portugal, an extensive research work has been performed in 3

4 the characterization of small roundwood timber obtained from maritime pine (pinus pinaster) forest thinning operations [15]. This species is the main softwood growth in Portugal, with an abundant number of trees being taken from thinning operations in the country that fulfil the main requirements for the barrier elements (e.g. the size and mechanical properties). However, the incorporation of this renewable material in periodic arrays to form acoustic barriers hasn t been tried yet, to the authors knowledge. In terms of sustainability, the log trees that are being proposed for this solution only need low processing, such as lathe and a preservative treatment. The construction of the acoustic barrier illustrated in Figure 1 can be performed with similar equipment that is used for the installation of utility poles for communication lines, by introducing the logs into the soil at a given depth. The proposed solution can thus be seen as a technologically simple and sustainable solution, with high potential in the mitigation of traffic noise. Next, one may find the description of the experimental work developed towards both the acoustic characterization of this innovative solution and the validation of different predictive numerical methodologies/tools implemented within the SCLog project. 3 Laboratorial testing and analyses One important component was the evaluation, in laboratorial conditions, of the acoustic behaviour of the small-scaled sonic crystal barriers, in terms of insertion loss (noise attenuation) and acoustic dispersion. A number of reduced-scale physical models were idealized and built with small-sized elements placed in simple periodic arrangements, in order to allow for the evaluation of their acoustic behaviour in controlled conditions. On the other hand, in the present research, several numerical models were implemented, based either on the Method of Fundamental Solutions (MFS) or on the Boundary Element Method (BEM), allowing the analysis of these structures in 2D, 2.5D and 3D analyses. For the 2D simulations, that reproduce the experimental conditions, the numerical model is based on the MFS, leading to results with similar level of accuracy of those obtained by the classical FEM or BEM approaches, but requiring less computational resources. The implemented numerical models were previously described in the works by Martins et al. [12] and Santos et al. [13]. 3.1 Noise reduction / Insertion Loss assessment The first prototype corresponds to a periodic structure built in the scale 1:5. This structure was built using PVC hollow tubes (PVC_U 40 PN4), arranged with regular distribution. The process for its manufacturing was defined so that great flexibility was obtained for positioning the scatterers and allowing different scatterer configurations to be tested (e.g. different distances between tubes). A sample image of the prototype is presented in Figure 2. A triangular lattice configuration was chosen, as preliminary results suggested that better results, in terms of sound attenuation, could be achieved with such arrangement. In order to get greater flexibility in generating a larger number of different geometric arrangements, the model may have up to five rows of cylinders with up to 20 (or 21) cylinders each, enabling those rows to be totally filled or to have voids along them. In this setup, a driver source in combination with an acoustical horn (Beyma TD460N Horn), that would provide controlled sound directivity characteristics, was the adopted sound source (see Figure 3). 4

5 Figure 2: View of the first small-scale sonic-crystal prototype (1:5 scale) For this first analysis, the tests were performed in an open-air environment (Figure 3), in wellcontrolled conditions. Although this setting was quite limited, it allowed a first set of tests to be performed, and an initial validation to be obtained. In relation to the frequency analysis domain, as the sonic crystal noise barrier is originally intended to tackle road traffic noise, the most significant frequency range can be seen as spanning from 500 to 1500Hz, with noise peaks at around 1000Hz in most cases. Since the model scale is 1:5, the relevant frequency analysis domain to be used in the tests should span at least from 2500 to 7500Hz, although during measurements the surveyed range was 125 Hz-12.5kHz. Figure 3: Experimental setup and first sonic-crystal prototype being tested These tests allowed some interesting results and data to be measured, and indicated a very good accuracy of the numerical models. One sample measurement results (given by IL Insertion Loss (db) as the difference between the average Sound Pressure Levels (SPL) measured with and without the sonic crystal barrier) are presented in Figure 4, together with the corresponding numerical calculations. These tests allow to experimentally observing the existence of band gaps with very high attenuation in the presence of Sonic Crystal structures. 5

6 IL (db) Numer. Experim Freq. (Hz) Figure 4: Experimental and numerical Insertion Loss (IL) results in 1:5 scale sonic-crystal tests A second set of reduced-scale periodic structures was built using smaller PVC tubes (Figure 5). The tested specimens correspond to 1:10 reduced scale models of a full-scale barrier and several configurations were tried out for the sonic crystal barrier, with variants to the traditional rectangular and triangular lattices and different numbers of rows and spacing between the cylindrical scatterers (see the schematic views on Figure 6). The cylinders used to build the reduced-scale barriers comprised a set of PVC hollow tubes with diameters of 20mm, which corresponds to 200mm cylindrical elements in full-scale acoustic barriers. Figure 5: View of the 1:10 small-scale sonic-crystal prototype and a prototype being tested This second set of the sonic crystal structures were tested on a previously prepared quasianechoic closed room, in order to avoid possible interferences generated by reflections in the vicinity of the crystals. The frequency range of interest in these tests ranges from 5000 to 15000Hz, as it represents, at the 1:10 scale, the usual range for traffic noise for the real scale. The sound attenuation provided by the acoustic barrier, is defined by the registered Insertion Loss, IL, calculated using the SPL registered at a square grid of 25 receiver points (Figure 6). For the two selected cases, the Insertion Loss was determined for a frequency range between 100 to 20000Hz, considering thin bands with a width of 100Hz. In Figure 6, the red lines represent the experimental results and the blue ones the numerical results. In general, the computed results closely follow the experimental measurements, except for lower frequencies of up to 2500Hz. At these lower frequencies, it is possible that the observed differences are due to the modelling of the sound source (which predominantly makes use of the woofer component), and to a less efficient sound absorption provided by the lining materials on the room, which can disturb the response by introducing undesired reflections. The band gaps can 6

7 be easily observed, with peak Insertion Loss results reaching almost 20dB, and a generalized sound attenuation is visible in the medium to high-analysed frequency range. IL (db) Insertion Loss Numerical Experimental Frequency (Hz) 10 4 IL (db) Insertion Loss Numerical Experimental Frequency (Hz) 10 4 Figure 6: Experimental and numerical Insertion Loss results in 1:10 scale sonic-crystal tests (top: rectangular lattice; and bottom: triangular lattice configurations 3.2 Sound diffusion/dispersion capacity An additional acoustical feature that was also studied in the laboratory was related to the sound diffusion/dispersion capacity of the sonic crystals. In practice, for traffic noise applications, this parameter is important to understand the effect of the noise barrier on the road side (next to the vehicles), where it is important to spread evenly the sound energy and not to contribute to an increased amplitude of the sound pressure field due to the presence of the obstacle. In this respect, high dispersion coefficients would be desirable. Tests were performed in a simplified semi-anechoic room, making use of the setup shown in Figure 7, and evaluating the spatial uniformity of the scattered energy by the sonic crystal and computing the diffusion coefficient of several configurations. The adopted measurement procedure followed the ISO standard [16], in a simplified semi-anechoic room so as to meet the free field required condition. The experimental analysis of the diffusive behaviour of sonic crystal structures was performed, in this work, and indicated a substantial increase of the diffusion coefficient when compared with a plane flat surface (see Figure 8). Several configurations were tried, essentially consisting of variants to the traditional rectangular and triangular lattices, with different numbers of rows and spacing between the small-sized cylindrical scatterers (hollow PVC tubes). The measured data was processed in order to obtain, for each specimen, the diffusion coefficient along a frequency range from 100 to 4000 Hz, for incidence angles of 0º, +35º and -35º. In the mid to high frequency range, there in an improved diffusive capacity of the sonic crystal, although this is not 7

8 observed at low frequencies. Additionally, there is not a clear trend regarding the evolution of the diffusive coefficient when comparing the periodic structures with 1, 2 or 3 rows, which may indicate that the lattice configurations can be further optimized. +90º R1,9-90º +60º -60º 3 +30º 0º -30º Figure 7: View of the semi-anechoic room and layout configuration for diffusion coefficient laboratorial measurement and computation Plain reference surface 0º SC Rectangular 2D - 1 row SC Rectangular 2D - 2 rows SC Rectangular 2D - 3 rows 0.7 Diffusion Coefficient d freq (Hz) Figure 8: Schematic representation of a sonic crystal analysed configuration and diffusion coefficient of this rectangular lattice configuration for normal sound incidence 4 Full-scale prototype and in situ evaluation A real-scale structure was designed, constructed and analysed in outdoor conditions. The structure was built using maritime pine timber logs, and it was evaluated in terms of its acoustic efficiency and in terms of its vibrational behaviour. Figure 9 presents some images of the involved raw materials and of the prototype structure. Figure 9: Images of the pine timber logs and the full-scale assembled prototype being tested 8

9 The implemented real-scale model corresponds to a barrier with four rows of circular scatterers, two rows with approximate diameter of 0.1m and the other two with about 0.2m, in both cases spaced by one diameter. The Insertion Loss of this structure has also been estimated using the implemented numerical models, and is presented in Figure 10, where the red line represents the experimental data and the blue one the numerically simulated results. This solution exhibits a peak of maximum attenuation around 1000Hz (reaching 15 db), and still provides good Insertion Loss from 600Hz to 1500Hz. It is also interesting to note that the global trend of both curves is quite similar, as previously observed from the laboratorial tests IL(dB) f(hz) 10 3 Figure 10: Experimental in situ and numerical insertion loss results in full-scale prototype test 5 Final remarks This paper describes a set of results obtained in the scope of a research project (SCLog) developed at the University of Coimbra. These results refer to a highly sustainable sonic crystal noise barrier that can be used for traffic noise attenuation. The Insertion Loss values registered on a series of reduced scale models (1:5 and 1:10) correspond to interesting levels of sound attenuation. Additionally, the obtained results for the diffusion coefficients of small-sized sonic crystal barriers show an increased interest regarding the spreading of sound energy by this kind of periodic structures, which may be used on the road side. An in situ test of a full-scale barrier made of pine timber logs has been prepared and tested in outdoor conditions, allowing to verify the main conclusions from the laboratorial tests and numerical simulations. Regardless of the positive results obtained so far, there seems to be space for further improvement on the consolidation of this innovative and sustainable concept and on the optimization of the periodic arrangements and use of different log diameters. Acknowledgments The support of FCT (Fundação para a Ciência e a Tecnologia) and COMPETE, through research project PTDC/ECM-COM/1438/2012, is acknowledged. This work was also partly financed by FEDER funds through the Competitivity Factors Operational Programme - COMPETE and by national funds through FCT Foundation for Science and Technology within the scope of the project POCI FEDER

10 References [1] Deymier, P.A. Ed., Acoustic Metamaterials and Phononic Crystals, Springer, [2] Martínez-Sala, R.; Sancho, J.; Sánchez, J.V.; Gómez, V.; Llinares, J.; Meseguer, F. Sound attenuation by sculpture. Nature, 378, 1995, pp [3] Kushwaha, M.S. Stop-bands for periodic metallic rods: Sculptures that can filter the noise. Appl. Phys. Lett., 70, 1997, pp [4] Sanchez-Perez, J.V.; Caballero, D.; Martinez-Sala, R.; Rubio, C.; Sanchez-Dehesa, J.; Meseguer, F.; Llinares, J.; Galvez, F. Sound attenuation by a two-dimensional array of rigid cylinders. Phys. Rev. Lett., 80(24), 1998, pp [5] Rubio, C.; Caballero, D.; Sanchez-Perez, J.V.; Martinez-Sala, R.; Sanchez-Dehesa, J.; Meseguer, F.; Cervera, F. The existence of full gaps and deaf bands in two-dimensional sonic crystals. Journal of Lightwave Technology, 17 (11), 1999, pp [6] Sánchez-Pérez, J.V.; Rubio, C.; Martínez-Sala, R.; Sánchez-Grandia, R.; Gómez, V. Acoustic barriers based on periodic arrays of scatterers. Appl. Phys. Lett., 81, 2002, pp [7] Castiñeira-Ibáñez, S.; Rubio, C.; Romero-García, V.; Sánchez-Pérez, J.V.; García-Raffi, L.M. Design, Manufacture and Characterization of an Acoustic Barrier Made of Multi-Phenomena Cylindrical Scatterers Arranged in a Fractal-Based Geometry. Archives of Acoustics, 37 (4), 2012, pp [8] Umnova, O.; Attenborough, K.; Linton, C.M. Effects of porous covering on sound attenuation by periodic arrays of cylinders. J. Acoust. Soc. Am., 119 (1), 2006, pp [9] Martinez-Sala, R.; Rubio, C.; Garcia-Raffi, L.M.; Sanchez-Perez, J.V.; Sanchez-Perez, E.A.; Llinares, J. Control of noise by trees arranged like sonic crystals. Jour. Sound Vib., 291(100), [10] García-Chocano, V.M.; Sánchez-Dehesa, J. Optimum attenuation of broadband noise by sonic crystals made of recycled materials. Appl. Acoustics, 74 (1), 2013, pp [11] Koussa, F.; Defrance, J.; Jean, P.; Blanc-Benon, P. Acoustical efficiency of a sonic crystal assisted noise barrier. Acta Acustica united with Acustica, 99, 2013, pp [12] Martins, M.; Godinho, L.; Picado-Santos, L. Numerical Evaluation of Sound Attenuation Provided by Periodic Structures. Archives of Acoustics, 38 (4), 2013, pp [13] Santos, P.G.; Carbajo, J.; Godinho L.; Ramis, J. Sound Propagation Analysis on Sonic Crystal Elastic Structures using the Method of Fundamental Solutions (MFS). Computers, Materials and Continua, 43(2), 2014, pp [14] Forest Products Laboratory, Research Program on Small-Diameter Material - General Technical Report FPL-GTR-110 (Rev.), [15] Morgado, T.; Rodrigues, J.; Machado, J.S.; Dias, A.G.; Cruz, H. Bending and compression strength of Portuguese Maritime pine small-diameter poles. Forest Products Journal, 59(4), 2009, pp [16] ISO Standard. Acoustics. Sound-scattering properties of surfaces. Part 2: Measurement of the directional diffusion coefficient in a free field,