Page 1

Baltimore, Maryland

NOISE-CON 2010

2010 April 19-21

High resolution absorption mapping with a pu surface

impedance method

Emiel Tijsa

Microflown Technologies

Zevenaar, The Netherlands

Hans-Elias de Breeb

Microflown Technologies / HAN University

Arnhem, The Netherlands

Eric Brandãoc

LVA/UFSC (Federal University of Santa Catarina) CTC-UFSC

Florianopolis-SC, Brazil

ABSTRACT

The in-situ surface impedance method with pressure-velocity probes is documented in many

publications (see e.g. 1-6). The method is based on the measurement of sound pressure (p) and

particle velocity (u) close to an acoustic absorbing material. A loudspeaker at a defined distance

is used to generate a sound field with a known radiation impedance.

The impedance of a small area (a few square centimeters) with a known impedance is

scanned with an ultra miniature pu probe very close to the surface. The area is made of steel with

a cut-out, and behind this a material with a known impedance is placed.

In this paper the method is explained, the spatial accuracy of the measurement is examined

and a visualization technique is presented with a display of the spatial distribution (2D picture)

of the damping properties as function of frequency.

a tijs@microflown.com

b deBree@microflown.com

c eric@lva.ufsc.br

1. INTRODUCTION

The properties of many sound absorbing samples are not uniform. Often the material package

itself is inconsistent, and adjacent structures influence the overall acoustic behavior. To get a

better understanding of the mechanisms that create the absorption it is useful to have information

about small material sections instead of one general absorption value of the entire structure.

The pu surface impedance method consists of measuring the impedance in situ close to the

surface with a combined sound pressure and particle velocity sensor. From the impedance above

the material also the material’s surface impedance, reflection and absorption can be obtained.

In 2007 this method was used to measure a small area of a quarter lambda resonator sample

(4). From multiple measurements close to the surface it could be shown that the resolution of the

method is in the order of millimeters. By using spatial summation the average absorption of the

whole sample could be obtained. Since then this method has been applied to measure small

samples or areas (5, 6).

Page 2

with cavities is used. The cavities are filled with damping foam and thus the local absorption

should depend on their shape.

In this paper the resolution of this method is studied even further. A sample of sheet metal

2. DESCRIPTION OF THE MEASUREMENTS

The sample that is measured consists of a layer of sheet metal with a sophisticated µ-logo.

Behind the logo there is a layer of acoustically damping foam. This sample is moved with an X-

Y plotter, figure 1 left. At a distance of ~1.5mm from the sample surface is measured with a

miniature probe. A point like sound source is positioned at 23cm from the probe.

The X-Y plotter is moved in small steps (0.67mm). In an area of 12.3cm2 (1.9 square inch)

the sound pressure and particle velocity is measured on 2754 locations.

Figure 1: Left: the sample is moved with a X-Y plotter. Right: The sample is measured in 0.67mm steps.

3. MEASUREMENT RESULTS

In this section the measurement values at each location are displayed. First the pressure, velocity,

impedance and absorption values are shown of 4 typical locations. Then various 2D and 3D

contour and surface plots are shown to demonstrate the resolution of the measurements.

A. Spatial variation

The frequency responses of four locations are shown in figure 2-4. In the legend the

measurement number is shown. Also their position above the sample is plotted with red dots in

figure 1, right.

In figure 2 can be seen that the spatial variation of particle velocity (right) is much higher

than the pressure (left). Up to 2kHz there is little difference between the situation where the

pressure microphone is above the damping material or above the steel plate. It is therefore

expected that the spatial variation of the impedance, reflection and absorption coefficients is

more depending on the measured particle velocity than on the sound pressure. (figure 3-4).

Page 3

10

2

10

3

10

4

40

45

50

55

60

65

70

Frequency [Hz]

SPL [dB re. 20E-6Pa]

[7 18]

[7 27]

[7 53]

[25 31]

10

2

10

3

10

4

30

40

50

60

70

Frequency [Hz]

PVL [dB re. 50E-9m/s]

[7 18]

[7 27]

[7 53]

[25 31]

Figure 2: Measurement result on 4 positions. Left: Sound pressure level. Right: Particle velocity level

10

2

10

3

10

4

10

-1

10

0

10

1

Frequency [Hz]

abs Z [-]

[7 18]

[7 27]

[7 53]

[25 31]

10

2

10

3

10

4

-180

-135

-90

-45

0

45

90

135

180

Frequency [Hz]

phase Z [degrees]

[7 18]

[7 27]

[7 53]

[25 31]

Figure 3: Impedance normalized to air on 4 positions. Left: absolute value. Right: phase

10

2

10

3

10

4

0

0.2

0.4

0.6

0.8

1

Frequency [Hz]

absorption [-]

[7 18]

[7 27]

[7 53]

[25 31]

Figure 4: Local absorption at 4 positions

B. Contour plot of velocity compared with logo image

In order to get an understanding of the measurements resolution the velocity level at each

location is displayed as a contour plot. This image is that plotted on top of the actual image of the

µ-logo, see figure 5. It can be seen that the logo contours are quite nicely followed.

Page 4

Figure 5: An overlap of a velocity contour plot at 2253Hz on top of the logo image

C. Spatial distribution visualized by surface plots

In figure 6 to 22 the results are visualized as 2D and 3D surface plots. In section 3A was already

mentioned that the sound pressure variation is small compared to the velocity variation. In figure

6 can be seen that also the logo is less clearly visible from the sound pressure representation. Left

the frequency with the best result is shown (2733 Hz). At higher frequencies the shape of the

logo is almost not recognizable.

Figure 6: 2D surface plots of the sound pressure level (dB re. 20E-6Pa). Left 2733Hz. Right 7195Hz

At certain positions the stepper engine of the Y axis of the XY robot vibrated. Therefore at lower

frequencies the measurement is much affected, see figure 7. Despite the vibrations the µ logo

shape is already visible at 220Hz (1.5 meter wavelength). Above 900Hz the pattern of the µ logo

is clearly visible and the influence of the vibration of the scanner is negligible. This is because at

higher frequencies the vibration of the XY plotter are weaker and the signals from the speaker

are higher.

Page 5

Figure 7: 2D velocity surface plot of 220Hz. The vibration of the XY scanner affects certain columns.

Figure 8: Velocity surface plot of 1857Hz. Left 2D, right 3D representation

Figure 9: Velocity surface plot of 7371Hz. Left 2D, right 3D representation

Page 6

Figure 10: Impedance surface plot of 3457Hz (absolute value). Left 2D, right 3D representation

Figure 11: Impedance surface plot of 2631Hz (phase). Left 2D, right 3D representation

Figure 12: Impedance surface plot of 3457Hz (real part). Left 2D, right 3D representation

Page 7

Figure 13: Impedance surface plot of 3727Hz (imaginary part). Left 2D, right 3D representation

Figure 14: Impedance surface plot of 9999Hz (imaginary part). Left 2D, right 3D representation

Figure 15: Reflection surface plot of 1878Hz (absolute value). Left 2D, right 3D representation

Page 8

Figure 16: Reflection surface plot of 938Hz (real part). Left 2D, right 3D representation

Figure 17: Reflection surface plot of 4740Hz (real part). Left 2D, right 3D representation

Figure 18: Reflection surface plot of 2921Hz (imaginary part). Left 2D, right 3D representation

The resolution appears to be best in the mid frequency range (1.5kHz – 5kHz), and is worse at

lower at higher frequencies (figure 19 to 22).

Page 9

Figure 19: Absorption 1110Hz

Figure 20: Absorption 2133Hz

Figure 21: Absorption 2367Hz

Page 10

Figure 22: Absorption 5490Hz, Absorption 9895Hz

4. CONCLUSIONS

The surface of a sample is scanned with a miniature pressure-velocity probe with high spatial

resolution. The velocity and pressure are measured at 2754 locations in a 12.3cm2 area. The

contour of the foam behind the cut-out in the sample is recognizable in a broad band (220Hz –

9.9kHz), but is most clearly visible in the mid-frequency range (1.5 to 5kHz). The spatial

variation of sound pressure, which has an omni directional sensitivity, is lower than the spatial

variation of particle velocity, which has a figure of eight shape sensitivity.

Because many of the fine details on the company logo are visible in the impedance and

absorption plots it can be concluded that the spatial resolution is in the order of a millimeter.

REFERENCES

1 Teruo Iwase, Koichi Yoshihisa, “Measuring Method of Sound Reflection and Absorption Characteristics Based

on the Particle Velocity and its Applications to Measurements on Such as Drainage Pavement of Road Surface”,

Internoise, 2003.

2 R. Lanoye, G. Vermeir, W. Lauriks R. Kruse and V. Mellert, “Measuring the free field acoustic impedance and

absorption coefficient of sound absorbing materials with a combined particle velocity-pressure sensor”, JASA,

2006.

3 Hans-Elias de Bree, Tom Basten, Emiel Tijs, “Two complementary Microflown based methods to determine the

reflection coefficient in situ”, ISMA, 2006.

4 Emiel Tijs, Hans-Elias de Bree, Tom Basten, Martin Nosko, “Non destructive and in situ acoustic testing of

inhomogeneous materials”, ERF33, 2007.

5 Hans-Elias de Bree, Martin Nosko, Emiel Tijs, “A handheld device to measure the acoustic absorption in situ”,

SNVH, 2008.

6 Eric Brandão, Emiel Tijs, Hans-Elias de Bree, “PU probe based in situ impedance measurements of a slotted panel

absorber”, ICSV, 2009