The sound insulation and directivity of the sound radiation from double glazed windows.

The sound insulation and directivity of the radiated sound from double glazed windows have been measured by different researchers. Previously, airborne sound insulation models have been used to predict the associated measurement results with limited success. In this paper, the importance of accounting for the structure borne sound transmission between two glazing elements via the window frame on the prediction results is demonstrated. The decreased stiffness of the wall cavity as the depth is increased is the reason why sound transmission via the window frame needs to be considered. The reciprocity argument provided by Davy for the prediction of the directivity of sound radiating into a room is validated and it is shown that once the structure borne transmission is considered, an additional weighting term is not needed to compensate for the extra wall collisions which the sound experiences when radiated at grazing incidence.

The influence of finite and infinite wall cavities on the sound insulation of double-leaf walls.

Theories used to predict the sound insulation of double-leaf cavity wall systems are usually based on the assumption that the wall is of an infinite extent. To account for the effect of the finite extent of the wall, limiting the angle of incidence, a finite radiation efficiency model or the spatial windowing method is used in order to obtain realistic predictions. However, the effects of the finite extent of the cavity are often not included. This paper presents an extension of a finite two-dimensional cavity theory to include limp panels on each side of the cavity. It is shown that the oblique incidence mass-air-mass resonance can only occur for certain frequencies and certain angles of incidence. This is the reason why the infinite extent theories under-predict the sound insulation. The results of the predicted sound insulation agree with measurements when the wall cavity is empty. To obtain agreement when the cavity is full of a porous sound absorbing material, a flow resistivity of about one-fifth of the measured value has to be used. Use of the actual flow resistivity gives sound insulation values that are 10 dB too high.