Research Article Sound Quality Evaluation and ... | ResearchGate

Hindawi Publishing Corporation Advances in Mechanical Engineering Volume 2014, Article ID 820875, 10 pages http://dx.doi.org/10.1155/2014/820875

Research Article Sound Quality Evaluation and Optimization for Interior Noise of Rail Vehicle Kai Hu, Yansong Wang, Hui Guo, and Hao Chen Automotive Engineering College, Shanghai University of Engineering Science, No. 333 Longteng Road, Songjiang District, Shanghai 201620, China Correspondence should be addressed to Yansong Wang; [email protected] Received 26 March 2014; Accepted 18 July 2014; Published 13 August 2014 Academic Editor: M. G. Prasad Copyright © 2014 Kai Hu et al. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. A procedure for sound filed simulation, sound quality (SQ) evaluation, and optimization of interior noise of a rail vehicle is investigated in this paper. Firstly, some interior noises are measured on site when the subway is running in tunnel at a speed of 60 km/h. The sound pressure levels (SPLs), loudness, sharpness, and roughness of the measured noise are analyzed. A finite element model for acoustical simulation of the carriage is established by using the Actran software. The accuracy and feasibility of the finite model are verified by comparing the psychoacoustical parameters from the simulations and measurements. By using orthogonal experimental design, finally, the best optimization scheme is put forward, which obtained a sound quality improvement with a 4.81 dB decrease in SPL and a 1.07 sone reduction in loudness. The proposed optimization scheme may be extended to other vehicles for improving interior acoustic environment.

1. Introduction Nowadays, rail vehicle has become the preferred way to solve urban transportation problems, due to its advantages of high speed, large carrying capacity, efficiency, low energy consumption, and less pollution [1]. The total rail length in Shanghai is 570 kilometers and the average daily passenger capacity reaches up to 700 million persons. However, it has been found that the interior noises from the subways seriously effect on both the ride comfort of the subways and the mental and physical health of passengers [2]. Thus, how to control the interior noise of the carriage has become a common concern of scholars. In the previous studies, the objective evaluations of rail-vehicle interior noise usually focused on the physical characteristics of a sound, such as sound pressure levels (SPLs), and intensity and power [3]. The vehicle industry has succeeded in controlling the A-weighted SPL of interior noise at 70 dB below. However, the SPL of interior noise still does not meet the requirements of ride comfort. Even when the A-weighted SPL is only about 40 dB, people may feel annoyed. This is mainly due to that people’s subjective feeling of the noise is less considered [4]. Therefore, sound

quality evaluation (SQE), which considered the subjective feeling, is more reasonable. The concept of sound quality is first put forward by Bodden et al. [5]. Nowadays, new noise control conception is put forward that noise control should be not only to reduce acoustic physical parameters, but also to improve the sound quality. In the interior sound optimization field, two kinds of methods are adopted. One is active noise control, and the other one is passive noise control [6]. Austria AVL LIST Company proposed 48 physical parameters for describing acoustic properties. The objective evaluation of interior noise is investigated based on loudness and smoothness in Honda Company [7, 8]. Some scholars proposed several SQ parameters, such as the A-weighted SPLs, loudness, sharpness, roughness, and fluctuation strength [9]. Recently, the finite element method (FEM), infinite element method, and statistical energy analysis (SEA) methods were introduced and widely used in acoustic simulation industries [10]. In this paper, the Shanghai subway Line 9 is taken as an example. To conduct the objective evaluation of SQ, the interior noises of the carriage are measured on site by using the Bruel & Kjaer acoustical measurement system. Objective evaluations of the measured interior noises were

2

Advances in Mechanical Engineering Table 1: Sound quality analysis of two measuring points.

Measuring points Measuring point 1 Measuring point 2

SPL/dB(A) 76.13 76.47

Loudness/sone 42.19 42.37

Specific loudness Weighted Specific sharpness Integral spectrum spectrum

Sharpness/acum 1.310 1.302

1

Sharpness

2

3

Roughness/asper 0.571 0.565

Measuring point 1

Measuring point 2

Figure 1: The calculation flow of sharpness.

performed by using the parameters of A-weighted SPL, loudness, sharpness, and roughness, respectively. Based on the three-dimensional geometric parameters of the carriage, a finite element model is established for interior sound field simulation by using the Actran software and the simulated results are verified by experiments. Finally, an optimization scheme for interior noise reduction is put forward by the orthogonal experimental design. The noise reduction results suggest that the proposed optimization scheme is effective and feasible for SQ improvement of subway interior noise.

2. Main SQ Evaluation Parameters

𝑓 2 0.76𝑓 ) + 3.5 arctan ( ), 1000 7500

(1)

where 𝑧 denotes critical band rate and 𝑓 denotes frequency. The critical band rate (1–24 bark) is corresponding to the frequency range (20–20000 Hz) in human auditory system. The loudness calculation method of Zwicker is used in this paper, which has been applied in the ISO532standard. The calculation formula can be expressed as follows: 24 Bark

𝑁=∫

0

𝑁󸀠 (𝑧) 𝑑𝑧,

(2)

where 𝑁 is total loudness, 𝑁󸀠 (𝑧) denotes specific loudness of critical band, and 𝑧 denotes critical band rate. 2.2. Sharpness. Sharpness (in acum) is used for describing the high frequency component proportion of sound spectrum, which reflects harsh degree of sound signals [12]. The sharpness can be calculated by Zwicker method. The calculation formula can be expressed as follows: 24 Bark

𝑆=𝑘

∫0

𝑁󸀠 (𝑧) 𝑧𝑔 (𝑧) 𝑑𝑧 𝑁

where 𝑆 is sharpness, 𝑔(𝑧) denotes weighting coefficients in different critical bands, and 𝑘 = 0.11. The calculation flow of sharpness is shown in Figure 1. 2.3. Roughness. Roughness (in asper) describes the fluctuation of modulated sound loudness when the modulated frequency is at 15–300 Hz [13]. The roughness can be calculated by 24 Bark

2.1. Loudness. Loudness (in phon) is one of main psychoacoustic parameters, which reflects the sound intensity degree of ear feeling [11]. Considering the frequency masking effect, selective characters of the human auditory system are described by critical band and critical band rate in loudness model. The relation between critical band rate and center frequency can be expressed as 𝑧 = 13 arctan (

Figure 2: Layout of the measuring points.

,

(3)

𝑅 = 0.3𝑓 mod ∫

0

Δ𝐿 𝐸 (𝑧) 𝑑𝑧 ,

(4)

where 𝑅 is roughness, 𝑓 mod denotes modulated frequency, and Δ𝐿 𝐸 (𝑧) denotes masking depth.

3. Noise Acquisition and Sound Quality Analysis 3.1. Noise Data Collection. In this paper, the AC06 type carriage of Shanghai Metro Line 9 is considered, whose design speed is 80 km/h. According to the standardGBT3449 called “measurement of noise inside rail bound vehicles” [14], the measurement procedures are conducted under the conditions of the subway operating at straight road and the background noises are 10 dB lower than the interior noises. The Pulse multichannel test system is adopted, together with some data acquisition equipment, such as a laptop, and a NA28 Sound Pressure Meter. Then the collected noise signals are first filtered, and thus a database of interior noise is established. The layout of the selected measuring points is shown in Figure 2, in which the carriage joint, door, and seat are defined as 1, 2, and 3. In each measurement, the noise signals with length of 10 s are measured 5 times. The height of the binaural microphone is set to 1.1 m, approximately corresponding to ears position of a sitting Chinese people. The measured subway is assumed operating in tunnel at a speed of 60 km/h. 3.2. The Objective Evaluation of Sound Quality. The selected test data are imported into the Pulse Labshop software to obtain the objective parameters of SQE. The results are listed in Table 1.

Advances in Mechanical Engineering

3 5

Vehicle Subway Light rail vehicle

Operating line On elevated In tunnel On elevated

Noise limits (dB) 75 83 75

80 70

4 3 2 1 0

60 SPL (dB)

Specific loudness (sone/bark)

Table 2: The maximum allowable limits of equivalent SPLs.

0

2

4

6

8

10

12

14

16

18

20

22

24

Critical band (bark)

50

Measuring point 1 Measuring point 2

40 30

Figure 4: Specific loudness of two measuring points.

20 10 0

31.5

125

500

2000

8000

One-third-octave center frequency (Hz) Measuring point 1 Measuring point 2

Figure 3: One-third-octave A-weighted SPL of two measured points.

It can be seen from Table 1 that the total SPLs at measuring points 1 and 2 are 76.13 dB and 76.47 dB, respectively; the difference is 0.34 dB. The difference can be attributed to different background noises and random errors. The maximum allowable limits of the equivalent SPLs in subway and light rail vehicle are shown in Table 2 [15]. It can be seen that the maximum allowable SPL is 83 dB when the subway is running in tunnel. Although the result does not exceed the limit, it also affects the ride comfort of the subway seriously. The loudness at measuring points 1 and 2 is 42.19 sone and 42.37 sone, respectively, and the difference is 0.18 sone. It can be known from auditory characteristics of the human ear that when the total loudness is below 25 sone, people’s subjective feeling is moderate. When the total loudness reaches 32 sone, people’s subjective feeling is less comfortable. When the total loudness reached 50 sone, people’s subjective feeling is very bad. When the total loudness is above 50 sone, the noise may damage the auditory system if people stay for a long time in this environment [16]. It can be seen from the measurement results that the total loudness of interior noise reaches up to 42 sone, which damages mental and physical health of passengers. The passengers who stayed in this excessive noise environment for a long time may feel nausea, dizziness, and malaise. To further determine the spectrum characteristics of interior noise, the collected data are imported into Pulse Sound Quality for analyzing one-third-octave A-weighted SPL and specific loudness. The results are shown in Figures 3 and 4.

It can be seen from Figure 3 that A-weighted SPLs of two measuring points reach maximum value at 500 Hz. The energy of interior noise is concentrated in low and middle frequency range of 20–1000 Hz. Figure 4 shows that specific loudness value is big in the range of 1–5 bark. The maximum values of the two measuring points appear at 1 bark. Therefore, how to reduce the low-frequency noise below 1000 Hz has become the key point to improving the SQ of interior noise.

4. Model Establishment and Simulation Analysis 4.1. Model Establishment. Firstly, the three-dimensional geometric parameters of the carriage are measured by laser range finder, tape, and vernier caliper. Based on these geometric parameters of the carriage body and its internal components, a CAD model is established in UG software which is shown in Figure 5. After simplifying some unnecessary components, the HyperMesh software is used for meshing according to acoustic simulation requirements [17, 18]. The size of these elements is defined as 70 mm. Finally, the finite element model includes 469391 structural meshes, 2095690 cavity meshes, and 480332 nodes, as shown in Figure 6. 4.2. Finite Element Method in Acoustics. In the vehicle engineering, the finite element method is mainly used for analyzing the interior noise generated from structural vibration. This method assumes that the sound propagation space is dispersed by using a lot of finite elements, and acoustic characteristics can be solved in spatial propagation process. Acoustic-structure coupling system is usually considered in this process [19, 20]. Under certain conditions, the aerodynamic equation and air continuous equation can be changed into acoustic wave equation. In acoustics, based on the

4


y

z

z

y y

z

z

x

x

z y x

z y x

x

(a)

(b)

y

x

(c)

z

y

x z x y

(d)

Figure 5: The established UG model of the carriage: (a) top view, (b) isometric view, (c) front view, and (d) left view.

(a)

(b)

(c)

Figure 6: The established finite element model of the carriage: (a) front view, (b) left view of structural meshes, and (c) isometric view of cavity meshes.

acoustic wave equation and the air unit function, the finite element state equation of cavity acoustic can be expressed as [𝑀𝑒𝑝 ] {𝑝𝑒̈ }

+

[𝐶𝑒𝑝 ] {𝑝𝑒̇ }

+

[𝐾𝑒𝑝 ] {𝑝𝑒 }

𝑇

+ 𝜌[𝑅𝑒 ] {𝑈̈ 𝑒 } = 0, (5)

where [𝐶𝑒𝑝 ], [𝑀𝑒𝑝 ], 𝜌[𝑅𝑒 ], and [𝐾𝑒𝑝 ] denote air damping matrix, air mass matrix, structure-acoustic coupling mass matrix, and air stiffness matrix, respectively, {𝑝𝑒 } denotes air nodal sound pressure vector, and 𝑈𝑒 denotes nodal

displacement vector. Similarly, the vibration state equation of air structure can be expressed as [𝑀𝑒 ] {𝑈̈ 𝑒 } + [𝐶𝑒 ] {𝑈𝑒̇ } + [𝐾𝑒 ] {𝑈𝑒 } = {𝐹𝑒 } + {𝐹𝑒Pr } ,

(6)

where [𝐶𝑒 ], [𝑀𝑒 ], [𝐾𝑒 ], and {𝐹𝑒 } denote structure damping matrix, structure mass matrix, structure stiffness matrix, and structure external excitation, respectively, and {𝐹𝑒Pr } denotes interface sound pressure vector. The state equation of acoustic-structure coupling system can be obtained by combining the above two equations.


5 Table 3: Material parameters of each component.

Component Body Seat Window

Material Aluminum alloy HDPE Tempered glass

Young modulus (GPa) 72 3.5 60

Poisson ratio 0.33 0.6 0.22

Density (kg/m) 2800 950 2600

Table 4: The data comparisons between simulation and experiment. Loudness (sone) 42.19 40.51

4.3. Sound Field Simulation. The Actran, as professional software in acoustical and vibration filed, can solve the problems, such as closed and/or open fields, acoustic coupling simulation, and sound wave radiation and scattering. In this present work, the Actran is used for acoustic simulation. The direct frequency response analysis method is used in the process. The analysis range is 20–500 Hz and the step length is 1 Hz. To obtain the simulated data, two points corresponding to the measuring points are defined in the Actran. It has been known from previous research that the main noise source is the wheel-rail noise when the vehicle is operating at the speed of 100 km/h or less [21]. The maximum speed of Shanghai Metro Line 9 is 80 km/h. Therefore, the wheel-rail noise is taken as the main noise source in the Actran. In the simulation process, the location, amplitude, and direction of wheel-rail noise and electrical equipment noise are defined as two main sources. The body, seats, windows, and so forth, are defined by different materials. The properties of these materials are listed in Table 3. A comparison of the simulated and tested data of measuring point 1 is listed in Table 4. It can be seen that the simulation results are consistent with experimental data in overall. The relative error of SPLs, loudness, and sharpness is less than five percent. However, the relative error of roughness is almost eight percent. There exist small differences between the simulated and tested results. This is mainly due to the fact that only electrical equipment noise and the wheel-rail noise are considered in the simulation, but the generation mechanism of the interior noise is complicated. The comparisons of SPLs and specific loudness between simulation and experiment are shown in Figures 7 and 8. It can be seen from Figures 7 and 8 that the variation tendency and total value of SPLs and specific loudness are similar. Therefore, it can prove that the finite element model and the simulation results are basically correct.

5. Orthogonal Experiment for Quality Optimization 5.1. Analysis of Carriage Sound Pressure Map. In the simulation process, the interior sound field should be calculated for targeted noise reduction. Part of the SPL distribution maps is shown in Figure 9, in which (a), (b), (c), and (d) represent the carriage sound pressure distributions of 20 Hz,

Sharpness (acum) 1.310 1.367

Roughness (asper) 0.571 0.529

120 110 100 90 SPL (dB)

SPL (dB(A)) 76.13 74.10

80 70 60 50 40 30 20 20

60 100 140 180 220 260 300 340 380 420 460 500 Frequency (Hz) Experimental value Simulation value

Figure 7: SPLs comparison between simulation and test.

5 Specific loudness (sone/bark)

Data Experimental data Simulation data

4 3 2 1 0

0

2

4

6

8

10

12

14

16

18

20

22

24

Critical band (bark) Experimental value Simulation value

Figure 8: Specific loudness comparison between simulation and test.

100 Hz, 150 Hz, and 250 Hz, respectively. It can be seen from Figure 9 that the SPLs at the both ends and the middle of the carriage are higher at 20 Hz, 100 Hz, and 150 Hz. At 250 Hz, the internal SPL distribution of carriage is complicated, but both ends are higher than other positions.

6


20.7

10.0

31.4

Map pressure 42.1 52.9

63.6

74.3

85.0

6.93

17.4

27.8


(a)

30.0

20.0

40.0


59.1

69.6

80.0

64.3

72.1

80.0

(b)

70.0

80.0

90.0

25.0

32.9

40.7


(c)

(d)

Figure 9: The SPL maps of the carriage at (a) 20 Hz, (b) 100 Hz, (c) 150 Hz, and (d) 250 Hz.

Part 1 The top surface

Part 2 Part 3 Part 4 Part 5

The left surface

The right surface

Part 6 Part 7 y

z

Part 8

z

x

Part 9

(a)

x

y

(b)

Figure 10: Each part of the carriage: (a) top view and (b) front view.

5.2. Measurement of Interior Sound Pressure Level. In order to verify the SPL distribution maps of the carriage, furthermore, a measurement experiment is conducted. In the measurement, the NA-28 type Sound Pressure Level Meter produced by Rion Company is used to measure SPLs of interior noise. The carriage is evenly divided into nine parts and each part includes three positions: the left, right surfaces and the top, as

is shown in Figure 10. The noise of each position is measured 5 times and then the five data are averaged. The measurement is conducted under the same conditions as those in the previous noise data collection. The measured total SPL of each position of the carriage is listed in Table 5. It can be seen from Table 5 that the total SPLs of position 1, 5, and 9 are higher at left and top surface, and the total SPLs of


7 4

Table 5: Measured total SPL of each position of the carriage.

3

2.5

The right side

Factor 3 The top

Factor 2 The left side

2

Factor 1

The right side

76.8 75.7 75.6 77.1 76.9 74.9 74.4 74.9 76.7

The top

76.3 73.6 72.4 73.8 76.3 73.6 73.8 73.9 75.1

The left side

76.5 76.7 74.5 75.4 77.0 76.0 75.5 75.1 77.6

3.5

The right side

The right surface

The top

Surfaces The top surface

The left side

Positions 1 2 3 4 5 6 7 8 9

The left surface

SPL reduction (dB)

SPL/dB

Figure 12: The tendency chart of index and factor of SPL.

1 0.9 0.8 0.7 0.6

The right side

The top

The left side

Factor 3 The right side

The top

The left side

Factor 2 The right side

The top

0.5

5.3. Orthogonal Experiment Design. According to the simulated SPL distribution maps and the measured data, some sound absorbing materials should be pasted on the middle and both ends of the carriage in the simulation process. However, sound absorbing material is expensive. If the material is used in all high SPL positions, the optimized scheme has no practical significance in engineering. In order to solve the above problems, a scheme of orthogonal experiment including three factors and three levels is designed for searching a good noise reduction scheme. Orthogonal experiment which is an efficient and economical testing method can greatly reduce the number of tests and the test results can provide effective information. In the orthogonal experiment, factors 1, 2, and 3 are defined as the carriage’s front end, middle portion, and back end. The left, right, and top sides corresponding to three factors are defined as the three levels, respectively. An orthogonal experiment table is designed, as presented in Table 6. It can be known from previous studies that the sharpness and roughness only have small effects on SQ [22]. Thus, only SPLs and loudness are taken as optimized index in the orthogonal experiment. The calculated results of each index are listed in Table 7.

Factor 1 The left side

position 1, 4, and 9 are higher at right surface. For the same surface, the SPLs of the middle part of the carriage are 0.5 ∼ 1 dB higher than those at both ends and the SPLs of both ends and middle position are 1 ∼ 3 dB higher than those at any other positions. The accuracy of the carriage SPL maps is verified by comparing the experimental data in Table 5 and the simulated SPL distribution maps. The highest SPL positions of the interior sound field are confirmed. Therefore, some effective measures should be taken for noise reduction.

Loudness reduction (sone)

Figure 11: Extra-fine glass wool.

Figure 13: The tendency chart of index and factor of loudness.

In the multiobjective orthogonal experiment, the surfaces are defined as the extra-fine glass wool, a kind of porous sound absorbing material, as is shown in Figure 11. The material has many advantages such as light mass, noncombustible, and antiaging [23, 24]. Thus, it is suitable for sound absorption in the carriage. It has been known from the sound absorption properties of porous materials that, as the thickness is increasing, acoustical absorption coefficients will significantly increase in low and medium frequencies and keep a large value in high frequency. If the density of sound-absorbing material is increasing and the thickness remains unchanged, the absorption coefficient can be improved slightly in low frequency [25]. The interior noise energy in subway concentrates in the low frequency below

8

Advances in Mechanical Engineering 120

70

110 100

60 SPL (dB (A))

SPL (dB)

90 80 70 60

50 40

50 40

30

30 20 20 60 100 140 180 220 260 300 340 380 420 460 500

20

20

25 31.5 40

50

80 100 125 160 200 250 315 400 500

63

1/3 octave center frequency (Hz)

Frequency (Hz) Before noise reduction After noise reduction

Before noise reduction After noise reduction

(a)

(b)

5

Specific loudness (sone/bark)

4

3

2

1

0

0

2

4

6

8

10

12

14

16

18

20

22

24

Critical band (bark) Before noise reduction After noise reduction (c)

Figure 14: Comparisons of three parameters before and after the sound absorption treatment.

Table 6: Orthogonal experiment table. Levels Times 1 2 3 4 5 6 7 8 9

The front end

The middle part

The left side The left side The left side The top The top The top The right side The right side The right side

The left side The top The right side The left side The top The right side The left side The top The right side

Factors The rear end The left side The top The right side The top The right side The left side The right side The left side The top

SPL reduction (dB)

Loudness reduction (sone)

2.34 4.23 2.47 4.35 4.17 3.01 2.39 3.36 2.97

0.63 0.91 0.67 1.03 0.86 0.71 0.59 0.73 0.71


9 Table 7: The average value and range of two indexes.

Indexes

SPL

Loudness

Factors — Average value of level 1 Average value of level 2 Average value of level 3 Range Average value of level 1 Average value of level 2 Average value of level 3 Range

The front end 3.013 3.843 2.907 0.936 0.737 0.867 0.677 0.190

Absorption coefficient

125 250 500 1000

0.05 0.24 0.72 0.83

500 Hz. In order to obtain a better sound absorption effect, the absorption material with porosity 0.95 and thickness 6 cm is selected and applied in the Actran. The corresponding absorption coefficients in each band are shown in Table 8. It can be seen from Table 6 that the forth experiment works best which obtains a SPL reduction with 4.35 dB and a loudness reduction with 1.03 sone. Table 7 shows that, in terms of the SPLs, the range of the three factors from low to high is the front, rear ends, and the middle part of the carriage. The middle part is the main factor. In terms of the loudness, the range of the three factors from low to high is the middle part, the front, and rear ends of the carriage. The rear end of the carriage is the main factor. According to orthogonal test results, the tendency chart of index and factor of SPL and loudness are shown in Figures 12 and 13, respectively. It can be seen that the scheme of noise reduction achieves the best results when the top of the three positions is pasted by sound-absorption material. To verify the optimized effects, the best scheme is simulated in the Actran. The comparisons between before and after acoustic treatment are shown in Figure 14, in which (a), (b), and (c) represent the SPLs, A-weighted 1/3-octave SPLs, and specific loudness, respectively. It can be seen that the SPL of interior noise is reduced. The noise reduction effect is obvious, especially in the range above 300 Hz. This result corresponds to the sound absorption property of the material. Figure 14(c) shows that the specific loudness is reduced below 10 Bark and has little change at high critical bands. After treatments, the sound absorption scheme achieves a good effect on SQ improvement with a 4.81 dB decrease in SPL and a 1.07 sone reduction in loudness. The results show that the optimization scheme improved both the interior noise and the ride comfort of the sample subway. In order to present the just noticeable level difference, the SPLs comparison below 250 Hz is shown in Figure 15.

100

SPL (dB)

Extra-fine glass wool (radius of 4 microns)

Frequency (Hz)

The rear end 2.903 3.850 3.010 0.947 0.690 0.883 0.707 0.193

120

Table 8: Acoustical absorption coefficient of each frequency. Material

The middle part 3.027 3.920 2.817 1.103 0.750 0.833 0.697 0.136

80 60 40 20 20

40

60

80 100 120 140 160 180 200 220 240 Frequency (Hz)

Before noise reduction After noise reduction

Figure 15: Comparisons of SPL before and after the sound absorption treatment.

6. Conclusions In this paper, objective evaluation of interior noise is carried out in respect of the SQ of A-weighted sound pressure level, loudness, sharpness, and roughness. Based on the psychoacoustical parameters, an optimized scheme for interior noise reduction of rail vehicle is performed by using sound field simulation. The Shanghai Metro Line 9 is chosen as an example. The interior noise of the carriage is measured on site and a finite element model is established in the Actran. The accuracy of the simulated results is verified by experiment, in which the highest SPL positions in the sound field are confirmed by the SPL distributions. The optimization is investigated by the orthogonal experiment design which obtained a SQ improvement in both the SPL and loudness. The results indicated that the optimization scheme proposed in this paper is effective for SQ improvement of the sample vehicle and may be extended to other types of subway vehicle and thus it is beneficial for vehicle acoustical design in engineering.

10

Conflict of Interests The authors declare that there is no conflict of interests regarding the publication of this paper.

Acknowledgments This work has been supported by the NSFC (Grant no. 51175320) and partly supported by the Program for Special Appointment Professor (Eastern Scholar) at the Shanghai Institutions of Higher Learning and the Fund for Talents Development by the Shanghai Municipality, China.

References [1] M. R. Prascevic, D. I. Mihajlov, and D. S. Cvetkovic, “Measurement and evaluation of the environmental noise levels in the urban areas of the city of Nis(Serbia),” Environmental Monitoring and Assessment, vol. 186, no. 2, pp. 1157–1165, 2014. [2] Y. Soeta and R. Shimokura, “The impact of external environments on noise inside a train car,” Noise Control Engineering Journal, vol. 59, no. 6, pp. 581–590, 2011. [3] M. Yamaguchi, T. Sekiguchi, W. Cho, and T. Toi, “Objective sound quality evaluation for home electrical appliances by using physiological information,” in Proceedings of the 18th International Congress on Sound and Vibration (ICSV ’11), pp. 3006–3013, Rio de Janeiro, Brazil, July 2011. [4] W. Zhang, G. Chen, X. Xiao, and X. Jin, “Objective evaluation of sound quality of noises inside high speed train,” Journal of the China Railway Society, vol. 33, no. 2, pp. 13–19, 2011. [5] M. Bodden, R. Heinrichs, A. Linow et al., “Sound quality evaluation of interior vehicle noise using an efficient psychoacoustic method,” in Proceedings of the 3rd European Conference on Noise Control-Euronoise, vol. 98, pp. 609–614, 1998. [6] Z. S. Liu, H. P. Lee, and C. Lu, “Passive and active interior noise control of box structures using the structural intensity method,” Applied Acoustics, vol. 67, no. 2, pp. 112–134, 2006. [7] K. Noumura and J. Yoshida, “Perception modeling and quantification of sound quality in cabin,” SAE Transactions, vol. 112, no. 6, pp. 1790–1799, 2003. [8] D. Mao, Y. Gao, W. Yu, and Z. Wang, “Grouped pair-wise comparison for subjective sound quality evaluation,” Acta Acustica, vol. 30, no. 6, pp. 515–520, 2005. [9] E. Zwicker, H. Fastl, and U. Widmann, “Program for calculating loudness according to DIN 45631 (ISO 532B),” Journalof Acoustic Society of Japan E, vol. 12, no. 1, pp. 39–42, 1991. [10] K. Ueno and H. Tachibana, “Experimental study on musician’s evaluation of stage acoustics using a 3-D sound field simulation,” in Proceedings of the 6th International Conference on Auditorium Acoustics, pp. 353–360, Institute of Acoustics, 2006. [11] E. Zwicker and H. Fastl, Psychoacoustics, Facts and Models, Springer, Berlin, Germany, 1999. [12] H. Fastl, “The psychoacoustics of sound-quality evaluation,” Acustica, vol. 83, no. 5, pp. 754–764, 1997. [13] Y. S. Wang, G. Q. Shen, H. Guo, X. L. Tang, and T. Hamade, “Roughness modelling based on human auditory perception for sound quality evaluation of vehicle interior noise,” Journal of Sound and Vibration, vol. 332, no. 16, pp. 3893–3904, 2013. [14] GBT 3449-2011 Measurement of Noise Inside Rail Bound Vehicles, Standard Press of China, Beijing, China, 2012.

Advances in Mechanical Engineering [15] “GB14892-2006 Noise limit and measurement for train of urban rail transit,” Standards Press of China, Beijing, China, 2006. [16] T. Uhl and A. Chudzikiewicz, “Noise, vibration and durability problems in rail vehicle design,” in Proceedings of the 25th International Conference on Noise and Vibration Engineering (ISMA ’00), pp. 793–798, September 2000. [17] T. Yokota, S. Sakamoto, and H. Tachibana, “Sound field simulation method by combining finite difference time domain calculation and multi-channel reproduction technique,” Acoustical Science and Technology, vol. 25, no. 1, pp. 15–23, 2004. [18] X. Chen, C. F. Gao, X. H. Geng, and C. Xie, “Simulation and optimization to vehicle interior noise in mid and high frequency using VAone,” Applied Mechanics and Materials, vol. 268, no. 1, pp. 851–855, 2013. [19] M. Rokutanda, T. Kanamori, K. Ueno, and H. Tachibana, “A sound field simulation system for the study of ensemble performance on a concert hall stage,” Acoustical Science and Technology, vol. 25, no. 5, pp. 373–378, 2004. [20] M. T. Fei, L. Yi, M. H. Tao et al., “A simulation study on NVH preference of a light-bus with of the combustion rigidelastic coupling and fluid structure interaction,” Automotive Engineering, vol. 27, no. 1, pp. 72–76, 2005. [21] Y. Soeta and R. Shimokura, “Survey of interior noise characteristics in various types of trains,” Applied Acoustics, vol. 74, no. 10, pp. 1160–1166, 2013. [22] D. Wang, Z. Liu, J. Liang, S. Wang, and J. Jiang, “Subjective evaluation test and objective quantificational description of vehicle interior noise quality,” Journal of Jilin University: Engineering and Technology Edition, vol. 36, no. 2, pp. 41–45, 2006. [23] L. Haitao, Z. Xi, and S. Yong, “Developments of porous sound-absorbing materials,” Journal of Materials Science and Engineering, vol. 22, no. 6, pp. 934–939, 2004. [24] K. Sakagami, Y. Fukutani, M. Yairi, and M. Morimoto, “Sound absorption characteristics of a double-leaf structure with an MPP and a permeable membrane,” Applied Acoustics, vol. 76, pp. 28–34, 2014. [25] J. Caillet, F. Marrot, Y. Unia, and P. Aubourg, “Comprehensive approach for noise reduction in helicopter cabins,” Aerospace Science and Technology, vol. 23, no. 1, pp. 17–25, 2012.

International Journal of

Rotating Machinery

The Scientific World Journal Hindawi Publishing Corporation http://www.hindawi.com

Volume 2014

Engineering Journal of

Hindawi Publishing Corporation http://www.hindawi.com

Volume 2014


Volume 2014

Advances in

Mechanical Engineering

Journal of

Sensors Hindawi Publishing Corporation http://www.hindawi.com

Volume 2014


Distributed Sensor Networks Hindawi Publishing Corporation http://www.hindawi.com


Volume 2014

Advances in

Civil Engineering Hindawi Publishing Corporation http://www.hindawi.com

Volume 2014

Volume 2014

Submit your manuscripts at http://www.hindawi.com

Advances in OptoElectronics

Journal of

Robotics Hindawi Publishing Corporation http://www.hindawi.com


Volume 2014

Volume 2014

VLSI Design

Modelling & Simulation in Engineering


Navigation and Observation


Chemical Engineering Hindawi Publishing Corporation http://www.hindawi.com

Volume 2014



Volume 2014


Advances in

Acoustics and Vibration Volume 2014


Volume 2014

Volume 2014

Journal of

Control Science and Engineering

Active and Passive Electronic Components Hindawi Publishing Corporation http://www.hindawi.com

Volume 2014


Journal of

Antennas and Propagation Hindawi Publishing Corporation http://www.hindawi.com

Shock and Vibration Volume 2014


Volume 2014


Volume 2014

Electrical and Computer Engineering Hindawi Publishing Corporation http://www.hindawi.com

Volume 2014