Investigation of the predictability of porous ceiling absorbers with large cavities

Abstract

Modern construction often accommodates large installations, leading to unwanted large air cavities between the acoustic ceiling and floor structure. This raises concerns about the absorption coefficient of the porous absorbers and its impact on room acoustics. Porous ceiling absorber suppliers typically do not provide data for cavities larger than 400 mm, leading to potential inaccuracies in predicted reverberation times and challenges in achieving optimal acoustic environments. This study investigates the impact of large cavities above porous ceiling absorbers by creating a model based on the transfer matrix model and Delany-Bazley’s impedance prediction method. The model is developed to predict the absorption coefficient of acoustic ceilings, and its accuracy is validated through measurements and software simulations. Even if the model is limited to calculating infinite areas and only using Daleny-Bazley’s impedance model, which is not the most accurate, the results show problems for the absorption coefficient with variation in the cavities and the airflow resistivity. The results show that porous absorbers have a higher absorption at higher frequencies and that an air gap increases the overall absorption but has a minimal influence at high frequencies. However, increasing the absorber’s thickness or airflow resistivity does not always result in better absorption, especially with large air cavities. In cases with high airflow resistivity, the absorption at low frequencies can even decrease. The thesis also discusses the edge effect that occurs in reverberation room measurement, positive and negative influences on the classification of absorbers, and the challenges that building acousticians face. Further measurements and research on mounted ceiling systems are necessary for future work. The model needs to be developed for finite areas and with a more complex impedance prediction method for it to be adapted to real-world scenarios.