Experimental evaluation of the diffracting performances of multipurpose noise barrier profiles

Size: px

Start display at page:

Download "Experimental evaluation of the diffracting performances of multipurpose noise barrier profiles"

Joseph King
5 years ago
Views:

$Experimental evaluation of the diffracting performances of multipurpose noise barrier profiles Francesco Department of Industrial Engineering, University of Perugia, via G.$ $it The use of diffracting devices to be installed on the top of noise barriers is becoming more and more widespread, since such devices allow to limit barrier height and at the same time to decrease$

1 Experimental evaluation of the diffracting performances of multipurpose noise barrier profiles Francesco Department of Industrial Engineering, University of Perugia, via G. Duranti 67, Perugia, Italy, The use of diffracting devices to be installed on the top of noise barriers is becoming more and more widespread, since such devices allow to limit barrier height and at the same time to decrease the diffracted sound in the shadow zone. The design of these devices is however not supported enough by experimental results and theoretical analyses. A European technical specification (CEN/TS ) was recently issued to evaluate the intrinsic characteristics of diffracting devices; the standard describes a measurement method based on the Maximum Length Sequence technique. This method allows the qualification of a device in terms of two purposely defined indexes: DI (Diffraction Index) and DI (Diffraction Index Difference). The paper illustrates the measurement procedures and the related problems for diffracting devices. An experimental free-field test facility was realized for this purpose at the Acoustics Laboratory of the University of Perugia. Preliminary tests were carried out to set up the measurement methodology; a graphical user interface was then suitably developed for the elaboration and presentation of the results. Various prototypes of diffracting devices were tested, and in particular devices integrated with photovoltaic modules. Results show that the experimental facilities and the measuring methodology are both efficient. However, the installation of this kind of devices can sometimes worsen the noise reduction performances in comparison with the simple noise barrier, due to the presence of gaps between the barrier and the cap and to the geometry of the element itself. The effect can even be amplified if the surface of the device is highly reflective, as the ones considered in the present experimental campaign. 1 Introduction Noise barriers are the most widely spread road traffic noise reducing devices, as they represent in most cases the only economically-feasible solution. Among the various non-acoustical criteria guiding the choice of a noise barrier (e.g. cost, safety, durability, structural strength, weight, fire-resistance etc.), the visual intrusion of such devices has progressively gained more and more importance, at least in Italy, both for road users and for receivers. Provided that sound attenuation is not worsened, a reduction in the barrier height is therefore always welcomed, not just from an economical point of view. Great efforts have been dedicated in recent years to improve the performance of top-edge noise-barrier profiles: several added devices (also called caps or crownings ) of different shapes and materials have been tested theoretically and experimentally (on scale models or at full-scale). This papers reports the results of full-scale tests performed at the University of Perugia on novel barrier caps having curved-t profiles by means of a MLSbased method, in accordance with the recent European Technical Specification CEN/TS [1]. An experimental test facility has been recently designed and employed for intrinsic characterisations of noise barriers samples [2] at the Acoustics Lab of the University of Perugia. Unlike sound absorption and insulation properties of a noise barrier, which can be characterized by both laboratory and in situ measurements, in fact, the diffracting properties of a noise barrier cap can be measured only by in situ techniques. Data analysis has been performed in order to evaluate the acoustical performance of different caps purposely developed to be integrated with photovoltaic modules, with respect to geometrical (e.g. width) and surface material features (reflective vs. absorptive). 2 Brief review Great efforts have been dedicated in the last decades to improve the attenuation performance of noise barriers, without increasing the height but just modifying the top barrier profiles. In 1980, May and Osman [3, 4] investigated new barrier shapes by means of full-scale and scale model insertion loss testing: in particular, they observed the highest attenuation for T-profile, relatively wide, absorptive top barriers. In the same year, Seznec [5] suggested to use the boundary element method (BEM) for numerical predictions of the diffracted sound field behind a barrier. The BEM approach allows making comparative studies in a relatively short time: for these reasons, it has been extensively employed to predict the performance of different barrier shapes (e.g. T, Y, 1255

Forum Acusticum 2005 Budapest Two samples of barrier caps (Figure 2) have been sequentially placed upon the barrier and tested according to the measurement procedure detailed in Section 3.

Indeed, polycrystalline silicate PV modules were integrated in the first sample (Figure 2c). arrow, circular profiles [6, 7] and multiple edge barriers [8]).

Cylindrical, pear-shape and curved top barriers have shown lower benefits, unless an absorptive treatment was incorporated into the barrier tops.

attenuation due to the added device [9, 10, 11]. 3 Experimental methodology b a 3.

2 Forum Acusticum 2005 Budapest Two samples of barrier caps (Figure 2) have been sequentially placed upon the barrier and tested according to the measurement procedure detailed in Section 3.2, based upon CEN/TS [1]. The specimens have a curved-t non-symmetrical profile (bending radius: 500 mm), purposely chosen to optimize the performance of integrated photovoltaic (PV) modules. Indeed, polycrystalline silicate PV modules were integrated in the first sample (Figure 2c). arrow, circular profiles [6, 7] and multiple edge barriers [8]). In many cases, the T-profile top barrier has been shown to provide the highest insertion loss improvement. Cylindrical, pear-shape and curved top barriers have shown lower benefits, unless an absorptive treatment was incorporated into the barrier tops. More recently, attempts have been made to determine, both numerically and experimentally, the intrinsic efficiency of noise barrier caps or, that is equivalent, the achievable improvement in sound attenuation due to the added device [9, 10, 11]. 3 Experimental methodology b a 3.1 Test field and specimens In the framework of an agreement between the University of Perugia (CIRIAF) and the Italian Ministry for the Environment, an experimental outdoor test facility has been built at the Acoustics Laboratory of the University of Perugia to determine the in-situ intrinsic acoustic performance (reflection, insulation, diffraction) of prototypes and samples of noise barriers and additional devices [2]. A reinforced concrete basement has been designed to sustain a 4 m-high barrier with a bending stress equal to 1100 Pa; it is possible to install 6m-long samples. The test facility has been built in an open place, in order to avoid any parasitic reflection apart from that of the ground. c Figure 2: Tested barrier caps: (a) sample 1, (b) sample 2, (c) PV modules integrated in sample 1. A thick lightweight concrete barrier (6 m in length and 4.1 m in height) has been mounted in the mentioned test field, having a casketed absorptive surface (thickness range: mm) at the source side and a plane reflective one at the receiver side (Figure 1). The absorptive elements have been coated by fibrous layers (thickness: 80 mm) and covered by profiled perforated steel sheets. The second sample showed a similar profile, wider on the source side and was not integrated with PV modules; its bottom surface was coated by a 40 mm layer of polyurethane foam (Figure 2b and 3). The latter sample was tested twice, modifying its surface impedance: with a reflective top surface (steel sheet) and covered with the abovementioned porous foam. Tables 1 and 2 report the main characteristics (geometries and surface materials) of the examined barrier caps samples. reference plane reference height curved-t profile cap polyurethane foam test construction Figure 1: Concrete noise barrier employed for the test. View of the absorbing panels from the source side. Figure 3: Vertical cross section of the test construction and of the second tested sample. 1256

3 Table 1: Dimensions of the barrier top-edge and of the test specimens. Sample Maximum height [mm] Maximum width [mm] 0 (no cap) (PV-integrated curved T-profile) 2 (wider curved T-profile) Measurement procedure An impulse-response method was chosen to evaluate the diffracted sound attenuation performed by the barrier caps. It essentially consists in determining the system impulse response h, by means of a crosscorrelation technique employing a maximum-length signal with and without the test specimen mounted over the barrier [9, 10, 11]. Table 2: Surface materials employed for the performed tests. Test Code Barrier cap sample Surface Material Bottom Top P 1 reflective reflective R 2 absorptive reflective A 2 absorptive absorptive C 0 - reflective The method, recently prescribed by a European Technical Specification [1], defines a diffraction index DI to quantify in db the top-edge attenuation performance in terms of the ratio of diffracted and incident sound intensities I, in the third-octave bands between 100 and 5000 Hz: I d DI = 10log10 = Ii 2 n 2 d k F[ hdk () t wdk () t ] df k = 1 d f i = 10log10 2 n F[ hi() t wi() t ] df f where the suffixes d and i stands respectively for the diffracted and incident components, F indicates the Fourier transform and f is the bandwidth. System impulse responses h dk and h i have to be respectively (1) recorded, according to [1], in several sourcemicrophone positions around the noise reducing device (at least n = 4, see Table 3) and in a reference free-field position (being the ground the only reflective surface in the surroundings). Two measurement sessions have to be performed at different angles of incidence: one at normal (θ = 90, free-field reference positions: S1-M2) and another at oblique (θ = 45, free-field reference positions: S4- M8) incidence with respect to the barrier surface. The diffraction index is given by the sum of the results obtained by the two measurement sessions: DI = DI + DI (2) Table 3: Source and microphone positions and relative heights with respect to the top edge reference height [1] (see Figure 3). θ = 90 θ = 45 Code Height [m] Code Height [m] S1 (source) S4 (source) M1 (mic.) 0.75 M7 (mic.) 0.75 M2 (mic.) 0.50 M8 (mic.) 0.50 M3 (mic.) 0 M9 (mic.) 0 M4 (mic.) M10 (mic.) The time windowing procedure in Equation (1) (through a purposely defined Adrienne window w [1]) is needed to isolate the useful components from the parasitic ones (e.g. the top-edge diffracted response from the side-edge one). A geometrical wave spreading correction (through the factors d, based on a spherical spreading assumption [2]) has to be performed in order to take into account the different path lengths from the source to the measurement points. Indeed, the impulse response amplitude is inevitably dependent on the source-microphone distance. Measurements of the diffraction index have to be carried out in sequence with and without the test specimen installed upon the barrier and positioning the source and the microphone in order to keep unchanged their relative height with respect to the top edge. These two sessions lead respectively to evaluate the indexes DI ad and DI 0 ; their difference is called Diffraction Index Difference and provides an indication of the improvement of performance achievable installing the added device under test: DI = DI DI (3) It is clear that, to a certain extent, the measured diffracting performance depends on the characteristics ad

4 of the noise barrier on which the sample is placed (test construction). The overall performance of the device under test with respect to road traffic noise can finally be evaluated (in db) by a single-number rating, weighting the diffraction index calculated in third-octave bands with the normalized traffic noise spectrum L i defined in EN [12]: DL DI = 10log 3.3 Instrumentation 18 i= Li 0.1 DIi i= Li (4) A two-channel PC card front-end (01dB-Metravib model Symphonie) has been used to generate the MLS signal (clock frequency: Hz, order 16, 32 averages), drive the source (two-way coaxial speaker, Bouyer model CP2050), record the microphone signal (sampling frequency: Hz, passband: Hz, ½ free-field condenser microphone G.R.A.S. model 40AR) and finally perform the input-output crosscorrelation procedure in order to determine the system impulse response. Post-processing of the responses has been carried out by means of a code and a graphical user interface (GUI), developed on purpose (Mathworks Matlab release 13) to control all the numerical operations needed to get the final results: construction and positioning of the time windows, power spectra calculation, wave spreading corrections, graphical output. In Figure 4 an example of the GUI output is shown. Considering the length of the employed test construction and, consequently, of the specimen under test (6 m), the time window width has inevitably to be shorter than that prescribed in [1] (6.5 ms vs ms), in order to get rid of the parasitic reflections; a lowfrequency band limit of 250 Hz has then been estimated for the results reported in the following Section. 4 Results Figure 5 shows the diffraction index data obtained, through the measurement procedure described in Section 3.2, for the four tests performed according to Table 2. It is apparent that the configuration used for test A is the only able to provide diffraction index values higher than that without the cap (test C), in the frequency bands above 1 khz. Such result is not totally surprising, being achieved by other Authors in the past [11]. Nonetheless, the fairly low values observed for sample 1 (test P) can be justified by the presence of gaps between the cap and the top-edge of the concrete noise barrier due to non-perfect mounting conditions. A mean improvement of 2 db was then achieved employing a wider cap (sample 2) and sealing all the gaps (test R). The curved T-profile, however, although optimized for the efficiency of integrated PV modules at medium latitudes, shows a rather poor intrinsic diffraction performance, at least with respect to that of a relatively thick noise barrier: its shape (Figure 2) may indeed have the effect of guiding the sound waves in the so-called shadow zone. The diffraction index differences DI, calculated by Equation (3), are shown in Figure 6 for sample 2, with a reflective top surface (test R) and an absorptive one (test A), demonstrating the strong improvement achievable by a finiteimpedance cap surface, for frequencies above 600 Hz. The constructive interference phenomena that occur at 2500 Hz with a reflective surface are then moved to 1600 Hz when the absorptive layer is added, leading to an increase of DI of 6 db in that frequency band. It is worth noting that no substantial differences can be observed below 600 Hz (the Hz third-octave bands should be disregarded): indeed, the only mean to get better results at lower frequencies seems the use of Helmoltz s resonators, as investigated by other Authors [13]. The effect of modifying the top surface impedance of the barrier cap can even be investigated from a local point of view: indeed, the diffraction index difference and the corresponding single-number rating can be evaluated for each measurement point, through Equations (3) and (4). In Figure 7, results obtained at normal (S1M1 S1M4) and at oblique incidence (S4M7 S4M10) are reported for tests R and A (see Table 2). It is apparent that, both for normal and oblique incidence, the attenuation performances of caps with the profile under test decrease going from the transition to the shadow zone of the noise barrier with respect to a relatively thick plane top-edge. The addition of an absorptive layer does not change the observed trend, while is able to provide significant improvements, in particular at oblique incidence (mean difference among test A and R: 2.5 db), exceeding the performance of the reference test construction, at least above the top-edge level. The corresponding values of the single number index DI (Diffraction Index) are negative for the first two tests P and R, while for test A, which presents a satisfying performance, is only equal to 0 db. It is clear that the index is of scarce significance, since diffraction measurements are highly influenced by the spatial configuration and the range of frequencies considered. 1258

5 Figure 4: Graphical user interface developed for post-processing the results. DI [db] 10,0 9,0 8,0 7,0 6,0 5,0 4,0 3,0 2,0 1,0 DIad - test P DIad - test R DIad - test A DI0 - test C 0, f [Hz] DL DI [db] 1,5 1,0 0,5 0,0-0,5-1,0-1,5-2,0-2,5-3,0 S1M1 S1M2 S1M3 S1M4 S4M7 S4M8 source-microphone position test R S4M9 test A S4M10 Figure 5: Results of the diffraction index measurement sessions (see Table 2). DI [db] 5,0 4,0 3,0 2,0 1,0 0,0-1,0-2,0-3,0-4,0 test R test A f [Hz] Figure 6: Effect of an absorptive surface on the diffraction index difference (see Table 2). Figure 7: Effect of an absorptive surface on the singlenumber ratings for single source-microphone positions. 5 Conclusions The study and optimization of the diffracting performances of caps and devices to be installed at the top of noise barriers is a topic of increasing scientific and technical interest. These devices are in fact becoming more and more common even though their design is not supported enough by experimental results and theoretical analyses. According to the recent CEN/TS , the evaluation of the performances of diffracting caps can be carried out only through in situ methods; to this extent, an experimental free-field test facility was realized at the Acoustics Laboratory of the University 1259

6 of Perugia, within an agreement with the Italian Ministry for the Environment. The paper describes the measurement methodology, and discusses the significance of the indexes proposed by the CEN/TS 1793/4. Measurements were carried out on different diffracting caps samples, integrated with photovoltaic modules. The results show that the methodology is accurate and reliable, even though the characterization of diffracting caps can not leave out of consideration a spatial analysis of the sound field; a detailed frequency analysis is also important, especially considering that the values of the single number index suggested by the CEN/TS (Diffraction Index DI) can be misleading. The geometry and the absorbing properties of the caps appear to be the fundamental parameters influencing the phenomenon; in particular curved profiles, as the ones examined in the paper show a tendency to accompany the sound wave into the lower part of the shadow zone, thus compromising the performance of the diffracting cap. Since good acoustic diffracting performances and high efficiencies of the PV modules appear to be in contrast, future research activities will focus on the optimization of the geometry and the inclination of the diffracting multipurpose caps. Acknowledgements The Author whishes to thank the Italian Ministry for the Environment for having supported the research and the realization of the free-field noise barriers test facility. The Author is also indebted with Mr. Giulio Pispola and Mr. Francesco D Alessandro, for their support during the experimental campaigns, and with Mr. Giorgio Mannelli (Saico Energiambiente S.p.A., Italy) for having provided the barrier and the test specimens. References [1] European Committee for Standardisation, Road traffic noise reducing devices - Test method for determining the acoustic performance Part 4: Intrinsic characteristics In situ values of sound diffraction, Technical Specification CEN/TS , December [2] F., G. Pispola and F. D Alessandro, Acoustic intrinsic performances of noise barriers: accuracy of in situ measurement techniques, to be published on Proc. 12 th Int. Cong. on Sound and Vibration, Lisbon (2005). [3] D. N. May and M. M. Osman, The performance of sound absorptive, reflective, and T-profile noise barriers in Toronto, Journal of Sound and Vibration, Vol. 71, No. 1, pp (1980). [4] D. N. May and N. M. Osman, Highway noise barriers: new shapes, Journal of Sound and Vibration, Vol. 71, No. 1, pp (1980). [5] R. Seznec, Diffraction of sound around barriers: Use of the boundary elements technique, Journal of Sound and Vibration, Vol. 73, No. 2, pp (1980). [6] D. C. Hothersall, D. H. Crombie and S. N. Chandler-Wilde, The performance of t-profile and associated noise barriers, Applied Acoustics, Vol. 32, No. 4, pp (1991). [7] R. J. Alfredson and X. Du, Special shapes and treatment for noise barriers, Proc. Inter-Noise 95, pp , Newport Beach, CA, USA (1995). [8] D. H. Crombie, D. C. Hothersall and S. N. Chandler-Wilde, Multiple-edge noise barriers, Applied Acoustics, Vol. 44, No. 4, pp (1995). [9] J. Defrance and P. Jean, Integration of the efficiency of noise barrier caps in a 3D ray tracing method. Case of a T-shaped diffracting device, Applied Acoustics, Vol. 64, No. 8, pp (2003). [10] P. Demizieux and G. Dutilleux, Experimental evaluation of CEN/TS on two noise barriers with caps, Proc. Inter-Noise 2004, Prague, Czech Republic (2004). [11] G. R. Watts, P. A. Morgan and M. Surgand, Assessment of the diffraction efficiency of novel barrier profiles using an MLS-based approach, Journal of Sound and Vibration, Vol. 274, No. 3-5, pp (2004). [12] European Committee for Standardisation, Road traffic noise reducing devices - Test method for determining the acoustic performance - Part 3: Normalized traffic noise spectrum, Technical Specification EN , [13] M. Moser, R. Volz, Improvement of sound barriers using headpieces with finite acoustic impedance, J. Acoust. Soc. Am, Vol. 106, No. 6, pp (1999) 1260

Performance of Roadside Sound Barriers with Sound Absorbing Edges

Performance of Roadside Sound Barriers with Sound Absorbing Edges Diffracted Path Transmitted Path Interference Source Luc Mongeau, Sanghoon Suh, and J. Stuart Bolton School of Mechanical Engineering,