Reproducibility of the Present and the Proposed Single-Number Quantities of Airborne Sound Insulation

Background. A proposal for new single-number quantities (SNQ) for airborne sound insulation was recently published by Scholl et al. The proposed SNQs, R _traffic and R _living, are determined in the frequency range of 50–5000 Hz. Their present counterparts, R_w + C_tr and R_w , are determined in the smaller frequency range of 100– 3150 Hz. There is concern that the high reproducibility values of ISO 140-3 measurements below 100 Hz could result in an increase in the reproducibility values of the proposed SNQs.

Aim. The aim was to compare the reproducibility values of the proposed and the present SNQs based on the results of a round robin test. The second aim was to compare the SNQs and their reproducibility values based on the pressure (ISO 140-3, 50–5000 Hz) and the intensity methods (ISO 15186-3, 50–160 Hz).

Methods. The laboratory measurements used for this study were obtained from a previously published interlaboratory measurement of a window. The sound reduction indices were measured at each laboratory using both the pressure method (ISO 140-3) and the intensity method (ISO 15186-3). The present and the proposed SNQs and their reproducibility values were determined from the interlaboratory data.

Results. The reproducibility values of the proposed SNQs were found to be larger (i.e. larger uncertainty) than the reproducibility values of the present SNQs when the pressure method was used. The reproducibility value of R _traffic was unacceptably high, 3.6 dB, while it was 3.1 dB for R_W + C_tr . Correspondingly, the reproducibility value of R _living was 2.1 dB while it was 1.5 dB for R_w . The increase was due to the inclusion of the measurement data in the 1/3 octave bands below 100 Hz which had significantly larger reproducibility values than the higher 1/3 octave bands included in the calculation of the SNQs. The use of the intensity method also resulted in slightly larger reproducibility values for the proposed SNQs. However, the increase was small and the reproducibility values of both R _traffic and R _living were approximately 2 dB.

Implications. It is suggested that scientific evidence for including the frequency range 50–80 Hz should be significantly improved before deciding that the low frequency measurements should be included in the calculation of the SNQs. The practical disadvantages of the change may be larger than the expected benefits.