Building Acoustics

Equations to calculate the uncertainty of the EN12354-1 estimate of the flanking sound reduction index due to the uncertainty of the input data are derived using the method of the ISO Guide to the Expression of Uncertainty in Measurement (GUM). The uncertainty equations have been validated using Monte Carlo simulations. It is shown that the magnitude of the uncertainty depends on the uncertainty of the resonant sound reduction indices of the elements, the uncertainty of the vibration reduction index and the uncertainty of the equivalent absorption lengths and areas of the elements. However, equations could not be derived to calculate the uncertainty of the EN12354 estimate of the apparent sound reduction index which has a lognormal probability density function and is therefore outside of the scope of the method of GUM. Monte Carlo simulations must be used to calculate the uncertainty of the apparent sound reduction index. It is recommended that guidance for calculating and declaring the uncertainty is included in future versions of EN12354, ISO10848 and ISO15712.

At present, new buildings are frequently built by thermally active concrete ceilings due to the requirements of thermal comfort and energy saving. Thus sound absorbers (sails, baffles) are installed below the ceiling in order not to deteriorate thermal efficiency. This frequently collides with architectural design asking for planar surfaces. A periodic arrangement of sound-absorbent strips in concrete ceilings shows higher sound absorption than is to be expected from surface-averaged sound absorption. This is well known, since the ceiling behaves like an acoustic phase grating. The challenge for acoustic consultants is to design periodic ceiling structures in a way that the sound absorption is adequate to keep thermal losses as low as possible. Based on the so-called Rayleigh expansion, frequently published so far, a simple model was derived to predict sound absorption at normal, oblique and field incidence. The model was validated by means of measurements and calculations for two types of absorbers: a micro-perforated resonator and a porous absorber made of recycled glass. The measured and computed results comply at normal incidence, whereas discrepancy between measurement and calculation is clearly higher at field incidence. It is assumed that this behavior is mainly due to the edge effect. The findings of a demonstration project in situ are: Sound absorption achieved by means of periodic strips in concrete ceilings is adequate for usual office use, thermal losses are very low, but implementation at the construction site requires further improvement. Part of this text was previously published in the proceedings of INTER-NOISE 2008.

This paper reports two experiments on the effects of degraded speech signals on memory for spoken lectures. Experiment 1 showed that broadband noise impairs university students' memory for a spoken lecture, even though the participants heard what was said. Experiment 2 showed that reverberation has detrimental effects to school adolescents' memory for spoken lectures, similar to broadband noise. The results suggest that poor listening conditions (resulting from background noise and/or long reverberation time) impair memory and learning, even if the conditions allow the listeners to hear what is said. Since the goal for students and pupils attending to lectures is to remember the lecture rather than just hearing what is said, the results presented here indicate that standards for acceptable signal-to-noise ratios and reverberation times in buildings designed for learning should consider the distinction between speech intelligibility and memory. Standards should be based on memory criteria instead of intelligibility criteria.

A prediction method for the sound insulation of walls by vibro-acoustical numerical analysis using the finite-difference time-domain (FDTD) method is described. In order to accurately predict the sound insulation performance of walls, numerical modeling of the vibration energy loss of walls in the vibration analysis is necessary. In this study, the energy loss at the boundary part of the plates and the internal damping of the plates are modeled and the sound transmission loss of glass plates and plasterboard walls is calculated. A reasonable agreement is found between the calculation and measurement results and the applicability of the numerical analysis is confirmed.

The results presented here are based on field measurements carried out in six Catholic churches (Goa, India). Rapid Speech Transmission Index (RASTI) and Subjective Speech Intelligibility (SSI) measurements were made in four different seating zones of a church for three speech source locations (altar, pulpit, high altar), two languages (English, Konkani) and two postures (standing, sitting). The effects of language and postures were not significant. The altar location and the sanctuary of the church were preferable for speech intelligibility. Although the altar location showed better averages than the other speech source locations, the best predictive relationship between RASTI and SSI was the exponential growth of RASTI for the high altar source location with SSI for the English language.