The influence of finite sample size on surface impedance determination of materials with low sound absorption at low frequencies
Regular paper
KTH-Royal Institure of Technology
Wednesday 3 june, 2015, 09:20 - 09:40
0.8 Rome (118)
Abstract:
The most common method to reduce noise is to add sound reducing material on
the domain
boundaries. One way to model the noise reducing effect of the material is to
set its surface
impedance as the boundary condition of the domain. The surface impedance can
be determined
experimentally, either from the standardized methods or by free-field
methods. The standardized
methods suffer from some drawbacks and are limited to normal or random sound
incidence. To
determine the surface impedance for specific sound incidence, measurement
are performed in free
field or in-situ. Several methods have been developed for this purpose and
these type of
measurements are of interest in this paper. The measurements are performed
close to the material
surface and based on the knowledge of the sound field from a monopole above
an impedance
surface, the surface impedance of the material can be derived. In earlier
papers, the errors in these
methods due to the ideal field description has been investigated both for
locally and bulk reacting
materials. The effect of signal conditioning, sensor types and unwanted
reflections have also been
studied. One major concern of the methods are the finite size of the sample,
and the influence of the
edge diffraction has previously been studied for high absorbing materials at
normal incidence. This
paper focus on errors associated with the free-field measurements in semi-
anechoic rooms for normal
and oblique incidence at medium and low frequencies for materials with low
absorption. This paper
provides an extension to previous studies: theoretical background to why the
edge diffracted waves
have large influence on the results and means to avoid them are presented.
This is studied both
theoretically and numerically. Results of parameter studies give guidelines
of how to avoid errors due
to finite samples when designing the measurement setup.