Abstract

A popular method to study the acoustical behaviour of perforated plates is by examining its effective reflection and transmission amplitudes on an incident wave, by constructing an effective (homogenised) boundary (or interface) condition.

However, since these homogenisation problems are numerical (or at least semi-analytic) in nature, there is a question of the accuracy of the solution compared with the actual nature of the perforated plate. Some effects appear to be smoothed over or absent entirely, such as the existence of ’’band gaps’’ and other frequency ranges where near-total transmission occurs. This talk aims to determine a classification of diffraction grating problems; those which are amenable to homogenisation methods and those which are not, and why. The effects of a non-zero mean flow are also discussed.