The influence of structure-borne modes on the sound insulation

It was shown by scaled model measurements that different frequencies of the structure-borne sound modes of a test specimen lead to an uncertainty of about 2 dB in the measured sound reduction index at low frequencies.

Measurements in building acoustics serve different purposes. The results of laboratory measurements are used to characterise the acoustic properties of building elements. These values are used by architects and civil engineers for the design process and especially for the prediction of the sound insulation in buildings. In-situ measurements are carried out with the aim to quantify the level of sound insulation actually reached and - if necessary - to find the reasons for discrepancies between measurement results and regulatory requirements or values stipulated under private law. In this context, the corresponding measurement uncertainties play an important role in assessing measurement results.

On this account, investigations into the uncertainties in building acoustics are a major task of the working group "building acoustics". A project finished in 2004 led thereby to the assumption that the frequencies of the structure-borne modes of the test specimen influence its sound insulation properties significantly. Within the scope of a diploma thesis, model measurements in a special test suite (fig. 1) have been carried out to quantify this effect. The homogeneous test specimens were identical with respect to their material and thickness but the frequencies of the structure-borne modes were varied by a variation of the size and aspect ratio of the specimens. Therefore, a special model test suite has been used which is a conventional test suite scaled down by a factor of 8. For the first time, this setup enabled a separate assessment of the influence of the structure-borne modes.

Figure 1: Test suite scaled down by a factor of 8 with test specimen and the multichannel measurement system used

The measured standard deviation increased significantly in comparison to the standard deviation of repeatability for bending resistant panels with only few modes in the lower third-octave bands (fig. 2). The uncertainty of the sound reduction index can therefore be estimated to be in the order of 2 dB due to different frequencies of the structure-borne modes.

At higher frequencies, this effect is negligible since both standard deviations are nearly identical (fig. 2). The latter is only valid for the corrected sound insulation index where the increase of the outer damping of the test specimen with increasing size has been taken into account.

Figure 2: Deviations ?R_k (E) and 95%-confidence intervals derived from the standard deviation of repeatability (W) and from the total standard deviation (G) for the corrected sound reduction index of bending resistant panels of different sizes and aspect ratios. The model frequency was scaled up by a factor of 8 corresponding to the down scaled test suite.

For the future, it is planned to quantify further uncertainty contributions like e.g. the damping of the test rooms or the dimensions of the test suites.

Contact person:

Volker Wittstock, FB 1.7, AG 1.71, E-mail: volker.wittstock@ptb.de