Sound diffraction by a partially inclined noise barrier

A new analytic method is presented for predicting the performance of a partially inclined barrier. The insertion loss of the barrier is predicted by summation of multiple diffractions occurring at the top and corner points, at which convex and concave edges are formed. Kouyoumjian and Pathakʹs diffraction theory (Kouyoumjian RG, Pathak PH. A uniform geometrical theory of diffraction for an edge in a perfectly conducting surface. Proceedings of the IEEE 1974;62:1448–61) is employed to construct diffraction coefficients for both convex and concave edges, whereas the barrier thickness is neglected and an omni-directional point source is assumed. A scale model experiment is performed in an anechoic room. Using measured speaker directivity and theoretical directivity formula does not show any significant changes in prediction. However, including additional diffraction at edge points due to finite thickness leads to improvement by removing many small oscillations in the frequency domain. The comparisons of theoretical and experimental insertion losses show reasonable agreements. For the two-dimensional case, predictions with and without thickness effect are compared to numerical result obtained by boundary element method (BEM), in which the smoothing phenomenon is more clearly observed and good agreement is found.