Civil Engineering Reference
In-Depth Information
intensity is measured in the reception side using an automated arm and an intensity probe
with a 12 mm spacer between two
1
/
2
-in microphones. This allows measurements to be
carried out from 100 to 5000 Hz. The measured transmission loss and predictions with
both the FTMM and the presented approach (FEM/BEM) are given in Figures 13.13
and 13.14. In the latter method, a mesh of 45
×
34
×
9 linear brick poroelastic elements
was used for the foam. This mesh was compatible with the plate's mesh (45
×
34
Quad4 shell elements). Since the foam was only attached along its edges to the plate,
the coupling boundary condition was modelled as an air gap inserted between the
two components for both the FTMM approach and the FEM/BEM approach. The air
gap was modelled with linear acoustic 8-node brick elements. The diffuse field was
modelled as a superposition of plane waves using a GAUSS integration scheme of
6
6 points (6 plane waves along
θ
and 6 plane waves along
φ
). It should be noted
at this stage that the mounted panel damping was not measured, and that a nominal
modal damping ratio of 3% was assumed in the analysis (this value is justified by
edge damping). Since the measurements were done in 1/3 octave bands, the FTMM
results were calculated at the band centre frequencies while the FEM/BEM results were
calculated at 1000 frequency points using a logarithmic frequency step (Figure 13.13)
and converted to 1/3 octave bands (Figure 13.14). It is seen that the FTMM leads to a
very good agreement throughout the frequency range of the test (up to 5000 Hz), apart
from a slight overestimation at low frequencies (below 500 Hz). This corroborates the
effectiveness of this method for predicting transmission loss of multilayer systems over
the whole frequency range (see Chapter 12). Equally, the FEM method shows excellent
comparison, except at higher frequencies; where it diverges as a consequence of the
mesh used. However, at these high frequencies there is no need for this deterministic
×
60
50
40
30
20
10
Test
Prediction (poroelastic model)
Prediction (limp model)
0
10
2
10
3
10
4
Frequency (Hz)
Figure 13.13
Transmission loss of a plate-foam system. Tests versus FEM predictions:
narrow band comparison.
