In situ impulse response method of oblique incidence sound absorption coefficient with microphone array

Transcription

1 doi: / In situ impulse response method of oblique incidence sound absorption coefficient with microphone array Jin Hua 1, Tianhu Wang 2 1 Engineering Training Center, Nanjing Forestry University, No.159 Lonpan Road, Nanjing , China; 2 School of Electrical and Information Engineering, Jiangsu University of Technology, No.1801 ZhongWu Avenue, Changzhou , China Abstract The paper reports the development of the methods for measuring oblique incidence sound absorption coefficient in situ. The standing wave tube method and reverberation room method are not suitable for the test in situ. And the transfer function method presented by Allard fails to consider scattering effect of unsmooth surfaces. Based on the use of microphone array, the paper presents an impulse response method to study oblique incidence sound absorption coefficient. The impulse signal generator is designed to simulate pseudo-random noise including many different frequencies. The microphone array can eliminate the effect of reflection and scattering. The paper compares the impulse response methods of in situ measurement in reverberation room and free field, with another method which used in standing wave tube. The specimen is made of polyurethane foam. The results show that the absorption coefficient obtained by the standing wave tube is similar to reverberation room, but smaller than free field. Although the results obtained by three methods are not the same, the three curves reflect the same varying tendency. This in situ method can satisfy the technical requirement of measuring. Keywords: impulse response, oblique incidence, sound absorption coefficient, microphone array, in situ. 1. INTRODUCTION There are three methods of measuring sound absorption coefficient known respectively as standing wave tube, reverberation room and transfer function (ISO, 2001; ISO, 2003; ISO, 2001). The standing wave tube method can only test small specimens and the reverberation room method is complex, which lead to disability of using the methods to measuring sound absorption coefficient in situ. The transfer function method is presented by Allard (Allard and Champoux, 1989). It can be used to measure the surface impedance of grounds in situ. The influence of measurement errors on the predicted surface impedance is investigated numerically (Kruse and Mellert, 2008). A free-field transfer function method is used to measure acoustic impedance (Nocke, 2000). The method for measuring in situ the sound absorption coefficient of road surfaces has the 69

2 one-third-octave-band frequencies ranging from 250Hz to 1600Hz under normal incidence conditions (ISO, 2010). A modified method using phase conjugation of ultrasound to measure sound absorption coefficient is proposed by Smagin (Smagin et al., 2013). A simple method using flow resistivity and porosity to calculate sound absorption coefficient of the sound absorbing materials of the automobile cab is proposed (Liu and Ji, 2014). Some experimental methods and equipment have been put forward. The reflection method using periodic pseudorandom sequences of maximum length is proposed (Massimo, 1993). Ducourneau s researches focus on measuring sound absorption coefficients in an industrial hall (Ducourneau et al., 2008). The experimental device equipped with an acoustic array is used to measure the sound absorption coefficient of flat panels subject to small angle sound incidence. The directivity of this array has been optimized so that the major part of the received acoustic energy would come from one portion (Ducourneau et al., 2009). The micro-machined device measuring acoustic flows is called the micro-flown or μ-flown that replaces the two microphones with dynamic flow sensors (Bree et al., 1996). Another method uses a parametric array as a source of sound within a test vessel capable of simulating ocean depths. The use of the parametric array enables wideband measurements to be undertaken with short-duration pulses and reduces the effects of diffraction from the panel edges (Humphrey et al., 2008). The method using spatial Fourier transform has also been developed to measure reflection coefficients at oblique incidence (Tamura et al., 1995). Rathsam introduces a method of measuring acoustic absorption in the field using a spherical microphone array (Rathsam and Rafaely, 2015). A finite surface method combining microphone array measurements over a finite sample with the sound field model in an inverse manner is proposed and validated (Marco, et al., 2016). A microphone arrangement is used to measure sound pressure distribution on a plane, with and without the presence of the sample to be measured. By using a spatial Fourier transform, diffusion and absorption of the analyzed sample are obtained (Alfio and Michael, 2015). For the irregular surfaces, the sound absorption combines with sound scattering that restricts the use of the transfer function method. The paper presents a method to measure the oblique incidence absorption coefficient in situ. From the collected references, the method is almost not seen. The microphone array takes the place of two microphones to avoid the effect of sound scattering. The impulse generator is designed to generate the pseudo-random noise including many different frequencies, which can improve its efficiency to measure the sound absorption coefficient. The test system consists of noise generator, microphone array, baffle, computer and power supply. They are all movable so that the test can be carried out of doors in situ. 2. MEASURING SYSTEM 70

3 Figure 1. Principle diagram of the test in situ. The system of measuring the sound absorption coefficient is shown in Figure 1. The material specimen is laid on the floor in the center of a room. There is a sound absorption baffle placed on the top of the specimen. The baffle can keep the impulse generated by the speaker from being received by the microphones directly. In testing, impulse signals from the generator are sent to the power amplifier and the speaker. The gap between the baffle and the specimen can ensure that the impulse will be reflected by the specimen surface from the side with the speaker to another side with the microphones. Considering that the microphones can be installed conveniently, the linear array with four microphones is used. After the sound signals are received by the microphones synchronously, they are sent to the data acquisition card and analyzed by the second development procedure of Labview software. The testing procedure will be repeated after the rigid plate with same shape replaces the material specimen. When the signals are reflected by the material specimen, the total sound energy can be calculated from the signals received by the microphone array: E ' E' 1 E' 2 E' 3... E' n (1) Where E i (i=1,2,3,,n) is the sound energy received by one microphone. When the signals are reflected by the rigid plate, the total sound energy can be calculated from the signals received by the microphone array: E E1 E2 E3... E n (2) Where E i (i=1,2,3,,n) is the sound energy received by one microphone. 71

4 From Equation(1) and Equation (2), the sound absorption coefficient can be computed as follows: E' 1 R 1 E (3) Where R is the sound reflection coefficient. 3. IMPULSE SIGNAL GENERATOR Adaptive impulse signals must be selected to measure the sound absorption coefficient by the method in situ. The ideal signals must distribute evenly in different frequencies to satisfy the sufficient signal-to-noise ratio. Otherwise, the impulse signals can be generated repeatedly, so the test results will be compared with each other. For getting better impulse signals, various instrument suppliers and producers begin to research impulse signal generator on the basis of digital circuit. The multichannel signal generator is made with the theory of pulse width modulation. The impulse signals are adjusted and filtered to get necessary linear waveform which can be used as not only the signal source, but also the input signals of a power amplifier to generate impulse signals. The impulse signal generator takes one chip (DSPIC33EP256MU806)as the core in Figure 2. Three methods can be done to generate impulse signals: 1) The digital signals output by the MCU chip can be converted to corresponding simulated voltage value with the A/D Convert Chip (MCP4911) in Figure 3. Then the pseudo-random impulse signals are exported through operational amplifier of second stage. Five kinds of time switches are designed to control the impulse time such as 0.5ms, 0.8ms, 1ms, 2ms, 4ms. 2) The impulse signals are generated by changing the output values of the I/O pins periodically with the timer, such as exporting 0 or 1 by turns. F1, F2, F3, F4, F5, F6, F7, F8 can output the digital signals that correspond the single frequency impulse waves of 100Hz, 160Hz, 250Hz, 400Hz, 630Hz, 1000Hz, 1600Hz, 2500Hz. 3) The own function of the MCU chip can generate impulse signals through configuring the registers. RE1, RE3, RE5 and RE7 can generate the single frequency impulse waves of 4000Hz, 6300Hz, 10kHz, 20kHz separately. The impulse signal generator is taken as the core of the impulse signal system that also includes the power amplifier (ZT-1035U) and the speaker (BS-633A). The system can generate the single frequency impulse wave, mixing frequency impulse wave and pseudo-random impulse wave that has many kinds of frequency. Figure 4 shows the characteristics of single frequency impulse of 4ms. 72

5 Figure 2. Diagram of DSPIC33EP256MU806 Pins Figure 3. Pseudo-random Signal DA. 73

6 (a) (b) Figure 4. Characteristics of single frequency impulse of 4ms. (a) (b) Figure 5. Characteristics of mixing single frequency impulse of 4ms. 74

7 The sound impulse lasts 4ms. The single frequency of the signals is 1000Hz, and its sound pressure level is above 75dB. There is an environmental noise near 1400Hz, and its sound pressure level is 15dB or so. Because of the measurement in situ, environmental noise is difficult to avoid. In the actual test, a quiet environment is very important to satisfy the accuracy of the results. Figure 5 shows the characteristics of mixing single frequency impulse of 4ms. The sound impulse lasts 4ms. The mixing frequency of the signals includes 400Hz and 1000Hz with correspond sound pressure level of 75dB and 16dB. There is an environmental noise near 2000Hz, and its sound pressure level is 5dB or so. (a) (b) Figure 6. Characteristics of pseudo-random impulse. Figure 6 shows the characteristics of pseudo-random impulse. The pseudo-random impulse signals drive the speaker by the power amplifier. Sound impulse lasts for 30ms. The high frequency components of the signals are relatively uniform. The sound pressure level of the frequencies is mostly more than 15dB. There is a maximum sound pressure level of 70dB near 500Hz. In the frequency range from 200Hz to 1600Hz, the sound pressure level of every center frequency of the one third octave frequency is less than 15dB. Because the dynamic range of the acoustic instruments is usually more than 60dB, the impulse signals can be effective. The impulse period is about 0.1s. The time domain and frequency domain of each impulse signal are very consistent. For the diagram of frequency and sound pressure of pseudo-random impulse signal, if the energy in each frequency band distribute uniformly, 75

8 the ideal distribution curve should be a straight line. There the pseudo-random impulse signal of the system is not flat with uneven spectrum that low frequency amplitude is high and high frequency amplitude is low. The sound pressure has a declined trend. But it can meet the requirement of the measurement of sound absorption coefficient on the whole. 4. COMPARISON OF RESULTS Figure 7 presents the test equipment for the absorption coefficient in situ in a room. The impulse signals of the generator are sent to the power amplifier and transferred to the speaker. The type of the speaker is BS-633A that has strong sound directivity. The height of the speaker with 45 inclined degrees is 1m. The gap between the sound absorption baffle and the ground is 40cm, which ensures the fine reflection of the sound. The linear microphone array made of four DR-238 microphones is 1m from the ground. There is 94mm distance between each two microphones. Figure 7. Measuring system in situ. The distance from the microphone array to the baffle is 1m, the same as the distance from the speaker to the baffle. The test equipment provides two methods to collect and analyze sound signals at the same time, such as NI-USB9215 data acquisition card with three channels and the sound card with one channel of Lenovo computer V470c. The values of five impulse signals every time are averaged to eliminate the effect of random disturbance. The procedure must be repeated twice considering that the test specimen and the rigid plate will be put under the baffle one after another. The standing wave tube method is used to compare with the transfer function method by microphone array indoors or outdoors. The specimen made of polyurethane foam is put into the standing wave tube. If the sample surface is uneven or irregular, the position of microphone should be far enough that the transfer function is in the plane wave region. Otherwise, it will lead to one air layer unexpected. Then one layer of the material, such as putty should be put in between the sample and the rear bottom plate of the impedance tube 76

9 to fill the back of the sample and increasing the thickness. So the sample surface can be parallel to the bottom plate. The test frequencies include 200Hz, 250Hz, 315Hz, 400Hz, 500Hz, 630Hz, 800Hz, 1000Hz, 1250Hz, 1600Hz. These impulse signals of different frequencies are confused to emit the pseudo-random noise, which can save time of dealing with the signals. In the test system in situ, the sound bandwidth is between 200Hz and 1600Hz. This system uses the sampling rate of the radio wave, that is, 22050Hz. It is wider than the sound bandwidth to satisfy the test requirement. The sampling number designed in the software is 1024, so the frequency resolution of the system is 21.5Hz. Figure 8. Comparison of sound absorption coefficient of three methods. The tests have been done in the standing wave tube, indoors like reverberation room and outdoors like free field separately. Figure 8 shows that the absorption coefficient obtained by the standing wave tube is similar to reverberation room, but smaller than free field. Although the results obtained by three methods are not the same, but the three curves reflect the same varying tendency of the sound absorption coefficient of the polyurethane foam. External ambient noise of the test outdoors causes the sound absorption coefficient to be larger than the other two methods. The sound absorption coefficient curve measured by the three methods is increasing from low frequency to high frequency. According to the theory of transfer function method, the main error lays in that the hardware device effects impulse signals when the signals go through a series of hardware devices. This error can not be avoided, and it belongs to the systematic errors. In particular, the microphones can not be synchronized with receiving the signals. It will interfere with the operation of the test system. In addition, the test system is placed in a certain environment in which the ambient noise has an impact on the test system. So it can be as much as possible to reduce the ambient noise, such as the choice of quiet environment and appropriate time to do the test. 77

10 5. CONCLUSIONS The impulse signal generator is designed in this paper, which is used to simulate the noise of various frequency bands. In the test, the sound absorption coefficient of polyurethane foam is tested by using pseudo-random signal, including the transfer function method of indoors simulation like reverberation room and outdoor simulation like free field. The results of the two methods are compared with the standing wave tube method, which verifies the feasibility of using the impulse signal method to test the sound absorption coefficient in situ. For the roughness of the surface, as a result of sound scattering, there may be some errors in the method of standing wave tube. If a large area of the sample is used, the directivity of the sound will be more uniform, which can reflect the average value of the sound absorption coefficient of the test material. The transfer function method has certain requirements for the distance between the test piece, the baffle, the sound source and the test point. Different parameters may lead to different test results. Ideal test results may not be obtained for the low frequency band of the simulation signal due to the longer sound wavelength. For the high frequency band, the test results may be relatively good. For the impulse signals of the generator, the frequency and sound pressure level curve is not an ideal straight line, and the low frequency part accounts for relatively large. But changes in the magnitude of the SPL values can meet the sensitivity test requirements of the microphones. In this paper, the method with the microphone array is used to replace the traditional transfer function method. When this method is used to test the oblique incident absorption coefficient, it can reduce the testing error of the system. Linear microphone array has been proved to have good effect. In this paper, a single speaker with strong directivity is used to make the sound signal. The directivity of loudspeaker array is stronger, and the other microphone array may be more suitable. It is worthy of further research. The test method of sound absorption coefficient in this paper can be used in the field to test the pavement, the lawn, the sound barrier, as well as other sound absorption materials with irregular surfaces. 6. Acknowledgements The authors would like to thank the funding support from the Natural Science Foundation of Jiangsu Province (Grants No. BK ) and Ministry of housing and Urban-Rural Development of China (No R1-021). 7. REFERENCES Alfio Y., Michael M. (2015). A measurement method for the sound absorption coefficient for arbitrary sound fields and surfaces, Acta Acustica united with Acustica, 101(4),

11 Allard J.F., Champoux Y. (1989). In situ two-microphone technique for the measurement of the acoustic surface impedance of materials. Noise Control Engineering Journal. 32, Batarseh L.R., Duwairi H.M. (2011). Isentropic sound propagation analysis and optimization over flat plate of saturated porous media with variable permeability, International Journal of Heat and Technology, 29(1), Batarseh L.R., Duwairi H.M. (2011). Permeability effect on sound wave propagation in saturated porous layer lied over flat plate, International Journal of Heat and Technology, 29(2), Bree de H-E., Leussink P., Korthorst T., Jansen H., Lammerink T. S. J., Elwenspoek M. (1996). The μ-flown: a novel device for measuring acoustic flows, Sensors and Actuators A: Physical, 54(1), Ducourneau J., Planeau V., Chatillon J., Nejade A. (2008). New method for measuring sound absorption coefficients in an industrial hall, Journal of the Acoustical Society of America, 123(5), Ducourneau J., Planeau V., Chatillon J., Nejade A. (2009). Measurement of sound absorption coefficients of flat surfaces in a workshop, Applied Acoustics, 70(5), Humphrey V.F., Robinson S.P., Smith J.D., Martin M.J., Beamiss G A, Hayman G, Ca.rroll N.L. (2008). Acoustic characterization of panel materials under simulated ocean conditions using a parametric array source, Journal of the Acoustical Society of America, 124(2), ISO. (2001). ISO : Acoustics-determination of sound absorption coefficient and impedance in impedance tube: Part 1. Method use standing wave ratio. ISO. (2001). ISO : Acoustics-determination of sound absorption coefficient and impedance in impedance tube: Part 2. Transfer-function method. ISO. (2003). ISO 354: Acoustics-measurement of sound absorption in a reverberation room. ISO. (2010). ISO : Acoustics-Measurement of sound absorption properties of road surfaces in situ: Part 2. Spot method for reflective surfaces. Kruse R., Mellert V. (2008). Effect and minimization of errors in in situ ground impedance measurements, Applied Acoustics, 69(10), Liu G.T., Ji X.M. (2014). Method for calculating sound absorption coefficient of the sound absorbing materials in the automobile cab, Journal of Mechanical Engineering, 50(12), Marco O., Jonas B., Jeong C.H., Efren F.G., Per T., Elisabet T.R. (2016). In situ measurements of the oblique incidence sound absorption coefficient for finite sized absorbers, Journal of the Acoustical Society of America, 139(1), Massimo G. (1993). Measurement of the sound-absorption coefficient in situ: the reflection method using periodic pseudorandom sequences of maximum length, Applied Acoustics, 39(1-2), Nocke C. (2000). In-situ acoustic impedance measurement using a free-field transfer function method, Applied Acoustics, 59(3), Rathsam J., Rafaely B. (2015). Analysis of absorption in situ with a spherical microphone 79

12 array, Applied Acoustics, 89, Smagin N.V., Krutyansky L.M., Brysev, A.P. (2013). Sound absorption coefficient measurements by phase-conjugate ultrasonic waves, Acoustical Physics, 59(2), Tamura M., Allard J. F., Lafarge D. (1995). Spatial Fourier-transform method for measuring reflection coefficients at oblique incidence. II. Experimental results, Journal of the Acoustical Society of America, 97(4),