Sound insulation in buildings: linking theory and practice

The unavailability of theoretical models to predict direct transmission across all types of building element explains why laboratory measurements remain important at the design stage. However, sound insulation in-situ is determined by both direct and flanking transmission; hence prediction models are essential tools. To indicate the limitations of laboratory measurements, transient and steady-state SEA are used to illustrate how the transmission suite affects structural reverberation times of solid test elements and the inherent errors in structural coupling measurements on isolated junctions of walls/floors. In the field, the 'mid-frequency prediction problem' is considered in heavyweight buildings where structural coupling data from isolated junctions is incorporated in models. Despite decades focussing on steady-state sound pressure levels in buildings, transient sources in buildings cause significant disturbance to building occupants, with regulatory requirements based on maximum sound pressure levels to protect against sleep disturbance. Recent work using transient SEA illustrates the potential for the prediction of maximum levels. Concerning the revision of field sound insulation measurement Standards, new approaches are described to improve repeatability below 100Hz, particularly with lightweight constructions where low-frequency performance can be problematic and to allow testers to use manual scanning to reduce the amount of equipment needed on site.