Virtual round robin tests for uncertainty quantification of single-number ratings for the sound reduction index

As transmission loss assessment is subject to a significant variability across a range of possible facilities, virtual round robin testing is applied to assess the inherent uncertainty of a given measurement procedure on the weighted sound reduction index. The joint probability distribution of the uncertain parameters is quantified by means of a maximum entropy approach. A Monte Carlo simulation is then performed to predict the uncertainty of the sound insulation. At high frequencies, an efficient method is needed to calculate the sound transmission. A nonparametric probabilistic method is therefore developed based on the Guassian Orthogonal Ensemble (GOE), where the wall and the rooms are coupled using the spatial correlation of room modes. With the presented approach, the sound reduction index can efficiently be predicted up to high frequencies. This allows quantifying the uncertainty of the weighted sound reduction index and its spectrum adaptation terms according to ISO 717-1:2013. Application of a calcium silicate block wall shows good correspondence with indicative values for uncertainty of the weighted sound reduction index given in ISO 12999-1.