Page 10 - base

Version HTML de base

Journée SFA / Renault / SNCF
9
Acoustique
&
Techniques n° 44
Absorption Coefficients: normal and random
incidence
The problem of the introduction of absorbents in SEA or
Ray-Tracing models is not recent and the solution known
for a long time: one has to use large reverberation rooms
(about 200 m
3
) and large surfaces of materials (12 m
2
)
to determine the diffuse field absorption coefficient of flat
materials or parts using the reverberation time technique
(cf. Norm ISO 354).
For practical reasons, small reverberant rooms have been
developed, the most famous one being the Alpha Cabin
(cf. figure 10) widespread in the automotive industry (a
common measurement procedure has been written by
Renault and PSA-Peugeot Citroën. : D 49-1977). We can
find many articles in the literature
dealing with the now well-known overestimation of the
Alpha Cabin (or small reverberant rooms), particularly in
the low and middle frequency range compared to large
reverberant rooms [10,11,12,13].
In fact, these small reverberant rooms are used below
the «Schroeder frequency» defining the frequency f
c
under
which the field is not diffuse anymore [6] :
f
c
=
2000
T
V
(4)
For the Alpha Cabin, with the maximum reverberation
time being T=2.5 s and the volume being V= 6,
44 m
3
, equation 4 yields f
c
= 1 246 Hz. The Alpha
Cabin absorption coefficient measurements are
nevertheless usually considered as valid between
400 Hz and 10 000 Hz...
As mentioned in [11,12] another important factor is the
diffraction effect due to the finite sizes of the samples.
The perimeter to surface ratio E of the sample has to
be low in order to minimize the diffraction. It is the case
for large reverberant rooms (E=1, 2 m
-1
for a 12 m
2
sample) but not for the Alpha Cabin (E=3 ,7 m
-1
} for a
1,2 m
2
sample).
Normal Incidence Absorption Coefficient
The Biot parameters of the absorbents and insulators
of the car are determined using an inverse technique
based on impedance tube measurements, direct airflow
resistivity measurements and direct Complex Young
modulus measurements.
Figure 11 shows the
very
good
correlation
between the impedance
tube measurement of a 13
mm thick PES felt and the
simulation using Maine 3A
V1.3 [9].
Thesameworkhasbeendone
for a 4 mm PES carpet and
a 20 mm Cotton felt leading
to very good correlation for
normal incidence as well.
RandomIncidenceAbsorption
Coefficient
The
three
materials
mentioned above have been
successively measured in
Alpha Cabin with a 1,2 m
2
flat sample (cf. figure 10) and
in the large reverberant emitting room 2 of the Faurecia’s
Center of Acoustic Technology with a 12 m
2
flat sample
(cf. figure 12).
The measurements are compared with the simulations
obtained with Maine 3A V1.3 by integrating on the solid
angle (the Biot parameters being unchanged compared to
Fig. 10 : small reverberant room : Alpha Cabin
Fig. 11 : Normal incidence absorption coefficient :
measurement versus simulation
Fig. 12 : Large reverberant rooms : coupled reverberating rooms
Vehicle Acoustic Synthesis Method 2nd Generation: an effective hybrid simulation tool to implement acoustic lightweight strategies